2

1 2 3 4 5 6

is highly concentrated. One example is the meatpacking industry in both the US and Canada. GIPSA (2007) shows that in 2005, the top four firms accounted for 67% of US animal slaughter. Canadian statistics indicate that the top four firms accounted for 79% of all animal slaughter and 62% of all poultry processing in 2001 (Harrison and Rule, 2004). In the UK, the top 10 slaughter plants account for approximately 37% of cattle slaughter, 44% of sheep and 60% of swine (Meat and Livestock Commission, 2006). Animal agriculture is also significantly concentrated geographically. Breeze (2004) estimates that 75% of the swine are raised in the Midwest, 90% of the broilers are in the Southeast and over 80% of the feedlot cattle are in the Midwest and Southwest. Canadian statistics show beef cattle are located predominantly in Alberta (48%) and Saskatchewan (21%), dairy cattle in Quebec (37%) and Ontario (36%), and poultry in Ontario (36%) (Beaulieu and Bédard, 2003). Geographic and industry concentration in the livestock industry increases the susceptibility to both intentional and unintentional biological threats (Cameron and Pate, 2001; Chalk, 2004; Cupp, Walker and Hillison, 2004; Crutchley et al., 2007). Second, livestock and meat products routinely move across large geographic areas, passing through many hands (Chalk, 2004; Cupp, Walker and Hillison, 2004; Crutchley et al., 2007). For example, cattle auction surveys show 20–30% of US cattle move to locations at least 48 km away from the auction site, often crossing several states within 3–4 days (Chalk, 2004). Cupp, Walker and Hillison (2004) argue that on average US meat travels approximately 1,600 km from farm to table. The degree of movement provides numerous potential entry points for animal pathogen introduction and makes it possible to have wide-spread disease outbreaks. USDA modelling suggests that food-and-mouth disease (FMD) could spread to as many as 25 states in as few as five days facilitated by animal movement (Chalk, 2004). Third, a lack of physical security and surveillance systems enhances vulnerability. Current control systems largely depend on farmer visual diagnosis and self-reporting (Kuchler and Hamm, 2000; NAO, 2002). This expands vulnerability in at least three major ways. First, since detection is based on visual observation of clinical signs, the disease could have been present and possibly spreading before discovery. Second, the clinical signs of some animal pathogens like FMD are indistinguishable from the signs of some other diseases (Bates et al., 2003; Elbakidze, 2007) and could be misidentified. Third, farmers often lack incentives to report infection. Kuchler and Hamm (2000) 4 Y. Jin, B.A. McCarl and L. Elbakidze demonstrate that farmers reduce their efforts to identify infected livestock within their flocks as the indemnity payments decrease. Furthermore, industry agents seldom prevent unauthorised access to animal holding areas, especially in the animal auctions, sale barns, ranches and smaller feedlots (Chalk, 2004; Crutchley et al., 2007). This is especially true in smaller operations. Pelzel, McCluskey and Scott (2006) suggest that the 2004 Texas avian influenza outbreak most likely was caused by unauthorised bird movement to and from a live bird market. Food processing and packing plants also tend to lack security and safety preparedness measures (Chalk, 2004; Crutchley et al., 2007). Overall, with the increasing concentration of animal agriculture and food supply chain along with globalisation, the likelihood of having unintentional introduction of animal pathogen is increasing. Deliberate attacks can exploit these vulnerabilities (Casagrande, 2001; Chalk, 2004; Crutchley et al., 2007). But in spite of this vulnerability, intentional cases of livestock (or crop) infections have been rare. Chalk (2004) lists 12 cases where pathogenic agents were used to infect livestock or contaminate food produce. Only two cases were categorised as terrorist attacks: the 1984 Rajneeshee salmonella food poisoning in Oregon and the 1952 Mau Mau plant toxin incident in Kenya. Japan is the only country known to have used anti-agricultural weapons (Madden and Wheelis, 2003). Wilson et al. (2000) indicates that anthrax was spread by Rhodesian Security Forces to destroy African cattle between 1978 and 1980 and glanders was used by the Soviet Army in Afghanistan against horses. However, Kumagai (2006) argues that a biological attack on an agricultural target should be expected to become one of the most likely terrorist modes. Tommy Thompson, former Health and Human Services Secretary expressing concern about the US food supply attack during his resignation speech, said “I, for the life of me, cannot understand why terrorists have not attacked our food supply because it is so easy to do!” (Washington Post, 2004). 2.2 Consequences of animal pathogen-related threats Animal pathogen introduction, whether intentional or unintentional, could cause substantial economic, social, political and public health disruptions. Given the historical scarcity of intentional events, researchers need to estimate the extent to which consequences of intentional animal pathogen use would exceed the consequences of unintentional introductions of pathogens to agricultural animal populations. It is most likely that intentional events, designed to cause maximum damage, will be more costly than random naturally occurring outbreaks. As shown in Pendell et al. (2007), simultaneous introduction in multiple production sites, a more likely scenario for an intentional attack, will be more costly than a single site introduction. In general, consequences can be divided into two broad groups: those associated with disease control and management and those associated with market and trade disruptions as discussed immediately below. 2.2.1 Costs associated with disease control and management Should an animal disease outbreak occur, extensive costs associated with disease control and management are expected. The 1997 Taiwan FMD outbreak cost the pork industry $US15 billion in total (Blancou and Pearson, 2003), among which the cost for surveillance, cleaning and disinfection amounted