Assessment of soy‐based imports into the United States and associated foreign animal disease status

Abstract Soy‐based products are known to pose a viable risk to U.S. swine herds because of their ability to harbour and transmit virus. This publication aimed to evaluate soy imports into the United States as a whole and from foreign animal disease positive (FAD‐positive) countries to determine which products are being imported in the highest quantities and observe potential trends in imports from FAD‐positive countries. Import data were accessed through the United States International Trade Commission website (USITC DataWeb) and summarized using R (version 4.0.2, R core team, Vienna, Austria). Twenty‐one different Harmonized Tariff Schedule (HTS) codes were queried to determine quantities (metric tonnes, MT) and breakdown of different soy product types being imported into the United States from 2015 to 2020. A total of 78 different countries exported soy products to the United States in 2019 and 2020 with top contributors being Canada (546,467 and 481,497 MT, respectively), India (397,858 and 430,621 MT, respectively) and Argentina (122,116 and 79,471 MT, respectively). Soy oilcake (582,273 MT) was imported in the largest quantities, followed by organic soybeans (270,194 MT) and soy oil (134,436 MT) for 2020. Of the 78 countries, 46 had cases of FAD reported through the World Organization for Animal Health (OIE) World Animal Health Information Database (WAHIS). Top exporters of soy products to the United States from FAD‐positive countries in 2019 and 2020 were India (397,858 and 430,621 MT, respectively), Argentina (122,116 MT in 2019) and Ukraine (40,293 and 56,392 MT, respectively). The risk of FAD introduction to the United States through soy imports can fluctuate based on where FAD outbreaks are occurring, shipping methods and end usage of products. A system to monitor these factors could help make future decisions about trade and risk of FAD introduction to U.S. swine herds.


INTRODUCTION
The ability of feed and feedstuffs to serve as a vector of disease for swine can have significant impacts on the U.S. swine industry. Feed has been associated with the U.S. porcine epidemic diarrhoea virus outbreak, which caused large death loss of nursery pigs and impacted pork supplies (Scott et al., 2016). Studies have shown that African swine fever virus (ASFV)-contaminated feed has the ability to cause infection in pigs as well (Dee et al., 2014(Dee et al., , 2018Niederwerder et al., 2019;Scott et al., 2016). Several viruses have been shown to survive shipping models in a variety of feed ingredients including soy products, such as soybean meal and soy oilcake (Dee et al., 2018;Stoian et al., 2020). Several viruses included in this study such as ASFV, classical swine fever virus (CSFV), Ajueszky's disease (pseudorabies) and foot and mouth disease (FMDV) are foreign animal diseases (FADs) in the United States and of direct interest to the swine industry. Of these viruses, only pseudorabies and CSFV have been introduced to U.S. swine herds and eradicated (Brown & Bevins, 2018b;Haagmans et al., 1999). Vaccines exist for CSFV, FMDV and pseudorabies, but their introduction to the United States would still be detrimental to the industry (Beer et al., 2007;Haagmans et al, 1999;Knight-Jones & Rushton, 2013). Of particular concern is ASFV because it has never been reported in the United States and does not currently have a vaccine against it. With this knowledge, it is critical to understand what feed ingredients are being imported to the United States and where they originated, so the risk level of FAD introduction can be evaluated. Of particular interest are soybean meal and soy oil because of their likelihood of being added to swine diets. Soy products have an increased ability to harbour viable virus when compared to other feed ingredients and organic soy is of particular interest because of the claim that most soy imports into the United States are organic (Dee et al., 2018;National Pork Producers Council, 2020). Some work has been done to prove an analytical approach to quantify soy imports into the United States but was focused only on ASFV-positive countries and not on total imports of soy products regardless of FAD status (Patterson et al., 2021). Also, understanding the ports of entry (POEs) into the United States for these products is beneficial for understanding the shipping time and conditions that the majority of soy products experience. This information is useful when modelling shipping conditions or for product traceability. The objectives of this paper were (1) to evaluate annual soy imports into the United States by product type and determine the portion coming from countries with FAD, (2) evaluate POE into the United States and determine ports handling the largest volume of products and (3) track soy import trends with regard to imports from FADpositive countries.

METHODS
This work looked at import records from 2015 to 2020 with a particular focus on 2019 and 2020. Product classification, quantity, country of origin, POE and year were obtained through the International Trade Commission Harmonized Tariff Schedule website (DataWeb; https://dataweb.usitc.gov/). Product categories are identified by unique 10-digit Harmonized Tariff Schedule (HTS) codes (Table 1). Twenty-one HTS codes associated with soy products that have potential to be used in animal feed were used to query the database. Several products, such as lecithins or butter substitutes, were included that may end up in by-products fed to animals. Data were exported to R (version 4.0.2, R core team, Vienna, Austria) where they were refined to total imports from each country by year and product type. Each HTS code was assigned a shortened description to improve data manipulation and reporting. Because of the low import rate of organic soy flour and meal; soy flour and meal; and soy flour and meal, not elsewhere specified or indicated (NESOI), these three HTS product categories were combined into one group in this report (

DISCUSSION
The soy-based HTS codes used were selected because of the product's potential to be included in swine diets in their imported state with minimal processing or their potential to be used as a by-product from the food industry. Several product groups, such as salad dressings, are not very likely to end up in a large volume of swine diets and therefore pose less disease risk. Products such as lecithins and butter substitutes also are not likely to be used in swine diets in their current forms but their inclusion in by-products fed to livestock gives them a greater chance of being added to swine diets. Thermal processing of feedstuffs, such as bakery by-products, may kill or inactivate pathogens in the original ingredients, but recontamination of the by-product postprocessing may allow viable virus to survive. In addition, this paper identified soy imports from FAD-positive countries.  et al., 2020). The current evaluation found organic soybean meal to be a very small portion of the imports into the United States, with 6690 MT of organic and non-organic soybean meal being imported out of 1,234,513 MT of total soy in 2020. Less than a third of that soybean meal (2163 MT) is sourced from countries with FAD cases which could be viewed as a low probability of disease introduction, but the severity and economic impact of disease introduction is still high. With an infectious dose of 10 6.8 TCID 50 for ASFV, 300 TCID 50 for FMDV and 10 4.2 TCID 50 for CSFV, even a small amount of virus entering the country can lead to infection of U.S. swine herds (Alexandersen et al., 2002;Cowan et al., 2015;Niederwerder et al., 2019). Smaller amounts of virus in feed can still lead to infection due to pigs eating throughout the day and accumulating virus at or above that infectious level.
Although organic soybean meal may not be a large contributor to U.S. soy imports, soy oilcake is imported in large quantities. Soy oilcake is the by-product of compressing soybeans to extract the soy oil. This oilcake can then be ground into soybean meal and included in swine diets. Soy oilcake made up 47% of soy imports in 2020, with 67% of the soy oilcake being imported from reported FAD-positive countries. The largest contributor, overall and of reported FAD-positive countries, of soy oilcake being imported into the United States in 2020 was India (ASFV-positive) with 387,269 MT. The current HTS codes do not differentiate between organic and non-organic soy oilcake; therefore, it was not possible to quantify the amount of this product that is organic using the USITC DataWeb. Dee et al. (2018) have shown that soy oil-cake is able to harbour viable, infectious virus through an international shipping model. The need to grind the soy oilcake down to soybean meal before use adds another point of concern. Viral contamination of feed processing facilities has been shown to occur after the handling of contaminated feed ingredients and decontamination can be difficult (Elijah et al., 2020;Huss et al., 2017). Although these studies did not address the grinding step, the production of dust during this stage of processing does provide potential for one contaminated batch of soy oilcake to infect subsequent batches by contaminating the equipment and environment in the manufacturing facility. This could allow one shipment of contaminated oilcake to inoculate multiple shipments of formerly uncontaminated oilcake or other ingredients that are processed through the same facility. Organic soybeans also were a large portion of soy imports contributing 22% of the total soy imports from 2020. Being able to understand how these soybeans are being used, whether for human consumption, livestock feed or oil extraction, will be beneficial to understand the risk of these beans introducing disease to U.S. swine herds. Although they are imported at about 4% of the total volume of soy products, other by-products such as bran, middlings and residues are important to keep in mind when considering soy imports because of the possibility of these products being used as a fibre source in swine diets.
POE can also be an important factor in disease risk of feed ingredients due to the variation of time required for transport. Routes that require significant land or air transport prior to reaching a seaport  (Dee et al., 2018;Dee et al., 2016;Stoian et al., 2020). such as soybean meal and soy oilcake (Dee et al., 2018). Pseudorabies has been eradicated from U.S. domestic swine herds, but is currently endemic in feral swine populations (Brown et al., 2019). This eradication was possible, in part, because of the availability of a vaccine for pseudorabies, and domestic swine herds can be vaccinated if pseudorabies is re-introduced (Haagmans et al., 1999). Classical swine fever virus has been eradicated from the United States since the 1970s, but this virus still exists in portions of Central and South America, Asia and Africa (Brown & Bevins, 2018b). Vaccines against CSFV currently exist and can be utilized if CSFV is introduced into U.S. swine populations (Beer et al., 2007). FMDV introduction to U.S. herds is particularly detrimental because of the large number of species that can be affected   Organization for Animal Health, whether that is a quarterly, 6-month or annual report. Due to this fact, the FAD status of countries within this timeframe may change as countries that had no data for a year report positive cases or a negative status. As a result, numbers for 2020 will likely change with the release of more information.
Overall, imports from reported FAD-positive countries contributed about 53% of the total soy imports in 2019 with India, Argentina and Turkey being the largest individual contributors within this group. In 2020, approximately 46% of the total soy import was sourced from reported FAD-positive countries. This high percentage is primarily due to reports of FAD in India and Argentina. Both of these were among the top three exporters of soy products to the United States for each year.
The trend of imports from ASFV-positive countries is a prime example of the impact of even one major soy exporter experiencing a disease outbreak. India breaking with ASFV in 2020 lead to a drastic increase in imports sourced from ASFV-positive countries. It also should be noted that this information does not take into account that the products imported from FAD-negative countries may have been imported from somewhere else previously. Similar to a statement expressed by Patterson et al. (2021), the interconnectedness of the global economy makes it difficult to trace the original source of products in some from would be beneficial in understanding the disease risk of the product more objectively. The end use is also an important consideration because a product that is used exclusively for human or industrial consumption also has a low disease risk, even if it is contaminated, because of its removal from interaction with swine herds.

CONCLUSION
Understanding the sources and intended uses of products being imported to the United States is vital to determine the risk of FAD disease introduction. Quantifying the amount and country of origin for imports into the United States is beneficial to start digging deeper into the biosecurity of feed across the country. Although this quantification is beneficial, it should not be taken as a defining declaration of the risk of FAD introduction without a holistic view of the storage, transport and usage of imported soy products. Being able to monitor FAD disease outbreaks and imports from countries could be useful for evaluating the risk of FAD introduction into U.S. swine herds more readily.