Influenza A virus (IAV) is an endemic pathogen in swine populations around the world (Vincent et al., 2008; Liang et al., 2014; Simon et al., 2014). Swine actively infected with IAV may display clinical signs characterized by loss of appetite, lethargy, dyspnoea, fever, nasal discharges and coughing (Vincent et al., 2008; Van Reeth et al., 2012). Additionally swine can be subclinically infected with IAV, which complicates detection and diagnosis (Bowman et al., 2012; Gray et al., 2012; Van Reeth et al., 2012).Transmission of IAV in swine has been documented to occur via multiple routes, such as direct contact between animals, through contact with IAV contaminated fomites and via airborne transmission. Direct contact is the most effective route of transmission for IAV; however, fomite and airborne transmission continue to provide an alternative route for IAV into swine populations (Allerson et al., 2013; Corzo et al., 2013a).

Swine are a potential source of novel IAV for the human population, as swine have demonstrated the capability to serve as ‘mixing vessels’ in which multiple IAVs can reassort (Scholtissek, 1990; Ma et al., 2009). When humans are infected with an IAV that typically circulates in the swine population, the infection is classified as ‘variant’ IAV (WHO, 2014). Zoonotic movement of IAV between swine and people has been documented globally and most is commonly associated with swine–human interfaces such as agricultural fairs, live animal markets, abattoirs and swine farms (Shinde et al., 2009; Van Reeth and Nicoll, 2009; Jhung et al., 2013; Perera et al., 2013; Choi et al., 2014).

The largest outbreaks of swine-to-human transmission of IAV have occurred at agricultural fairs and transmission events have been well documented in the United States (Myers et al., 2007; Jhung et al., 2013). In 1988, a woman contracted variant IAV and subsequently died after attending a county fair where ill swine had been reported (Wells et al., 1991). Subclinically infected swine can contribute to zoonotic IAV transmission, as documented in 2009 when a 12-year-old boy contracted variant H3N2 IAV after petting visually healthy swine at a county fair in Kansas (Cox et al., 2011). In 2011, multiple cases of variant IAV were associated with exposure to swine at an agricultural fair in Pennsylvania (Wong et al., 2012). The number of variant IAV cases in humans spiked in 2012, with 306 cases of variant H3N2 IAV (Jhung et al., 2013). With the advent of sequencing technology, many of these cases can be directly classified as swine lineage IAV through molecular epidemiology and linked to swine exposure at agricultural fairs (Bowman et al., 2014). These cases illustrate that infection with variant IAV can result in human hospitalizations and in some cases death. Every swine-to-human IAV transmission event is concerning from a pandemic preparedness standpoint; thus, controlling IAV at agricultural fairs is critical to protecting public health.

The North American exhibition swine industry is a diverse mixture of culture, education and business. Swine are frequently exhibited at agricultural exhibitions as part of educational projects to expand youth knowledge about agricultural practices. Additionally, swine can be shown in exhibitions open to any age competitor that occur throughout the year across the United States. For 2015, the National Swine Registry estimated that 1 million swine were involved in the United States exhibition swine industry (National Swine Registry, personal communication). Exhibition swine are a relatively small part of the national swine industry as a whole, representing an estimated 1.5% of the total swine population in the United States (USDA, 2012); however, exhibition swine represents the most common direct interface with a large number of humans, outside of the owners and caretakers, when commingled for the competition events.

Surveillance conducted from 2009 to 2011 identified one or more IAV-infected pig(s) at the end of approximately one quarter of the agricultural fairs tested in Ohio (Bowman et al., 2012). Of the fairs with infected swine, the average frequency of virus isolation was 62.9% (Bowman et al., 2012), demonstrating that a high proportion of the pigs at those exhibitions were actively shedding IAV at the end of the 5- to 7-day fairs. Overall, Bowman et al. (2012) found an IAV prevalence of 14.4% among exhibition swine at the end of fairs, higher than the ≤5% IAV prevalence typical of commercial swine herds (Corzo et al., 2013b). While it seems likely that IAV is amplifying through swine populations at livestock exhibitions, little is known about IAV prevalence in swine arriving at exhibitions before swine have commingled and IAV spread through the population. Previously, one study found a large variation in incoming prevalence estimates across two study sites, ranging from 0% to 19% via rRT-PCR testing (Gray et al., 2012). However, only a small proportion of the swine population was tested. A pilot study by Bowman et al. (in press) found the prevalence of IAV at one state fair to be 2.4% among the incoming swine. The objective of this study was to estimate the prevalence of IAV in exhibition swine as they enter fairs as a prelude to understanding the transmission of IAV at the fair with an ultimate goal of preventing intrafair transmission.

A total of 3547 samples were collected from the 5462 swine in attendance at the nine exhibitions. Of the samples collected, 188 (5.3%) were IAV positive using rRT-PCR and viable IAV was recovered from 53 (1.5%) (Table 1). Within exhibitions A through E, IAV prevalence, as determined by virus isolation, ranged from 0.2% to 10.3%. No IAV was detected in the samples collected from the swine at exhibitions F, G, H or I. Overall, IAV isolates were recovered from 28.2% of the samples identified as rRT-PCR positive but there was a wide range in isolation success from rRT-PCR-positive samples between fairs. For Exhibition E, 100% of the rRT-PCR-positive samples yielded an IAV isolate, whereas at Exhibition C, only one IAV isolate was recovered from the 16 rRT-PCR-positive samples (6.2%). Forty-seven (88.3%) of the 53 isolates were recovered during the first passage and the remaining 6 isolates were recovered through a second passage.

The 53 IAV isolates were subtyped as H1N1 (n = 23), H3N2 (n = 28) and mixed with both H1/H3 N1/N2 subtypes (n = 2) (Table 2). All IAV isolates contained the influenza A(H1N1)pdm09 lineage matrix gene. In Exhibition A, both H1N1 (n = 2) and H3N2 (n = 4) subtypes were recovered. Similarly, H1N1 (n = 18), H3N2 (n = 23) and mixed subtype isolates (n = 2) were found at Exhibition B. Only one IAV subtype per exhibition was detected among the swine entering exhibitions C-E (Table 2).

At the two fairs where swine were sampled in a chute and swine tested positive for IAV (exhibitions B and C), a pattern in rRT-PCR C_t values for IAV was observed at both fairs. As illustrated in Fig. 1a and b, samples with a low rRT-PCR C_t value were often followed by samples with similarly low C_t values that would gradually increase over subsequent samples (time) until the next low C_t spike. Principle component analysis found three distinct clusters of rRT-PCR-positive samples in Exhibition B and two clusters were observed at Exhibition C (Supporting Information). This pattern was not observed at exhibitions D and E, fairs where IAV-positive swine were detected but the swine were not sampled in a chute (Fig. 1c).

With viable IAV being recovered from only 1.5% of the swine entering the sampled fairs, findings of the present study highlight the potential to, and importance of, limiting the intraspecies spread of IAV during swine exhibitions. As swine are co-housed for several days during an exhibition, viruses arriving at the beginning of the fair have ample time to spread through the swine population. Surveillance for IAV in swine conducted at the end of exhibitions indicates that when IAV is among the swine at a given fair, viable IAV is typically recovered from 60% to 70% of the pigs (Bowman et al., 2012, 2014). This demonstrates a remarkable increase in IAV prevalence among swine between entry to the exhibitions and the end of exhibition. This rapid spread with in the population is expected as the basic reproductive rate for IAV in unvaccinated swine has been estimated at 10.66 (Romagosa et al., 2011). Certainly the risk to public health in these settings increases as the IAV prevalence among the exhibition swine increases during the fair, a notion supported by the timing of variant IAV detection in relation to the implicated fair (i.e. when variant IAV cases are associated with fairs, they are almost always detected at the end of the fair or in the days immediately following the conclusion of the fair). Strategies to maintain a very low prevalence of IAV within an exhibition site, thus preventing the apparent amplification of IAV during fairs, have the potential to reduce the threat to public and swine health. Strategies such as shortening the swine exhibition period and encouraging IAV vaccination have been suggested (Bowman et al., 2014; Officials & Veterinarians, 2014). However, vaccine effectiveness is challenged by the rapid evolution of IAV and strains circulating in the field strains may quickly differ from strains used in vaccine production. Additionally, vaccination has been shown to eliminate clinical signs of IAV in swine, without blocking infection and pathogen transmission (Loving et al., 2013). In these cases of subclinical infections and mismatched strains, vaccination of the pigs likely does not completely prevent swine–human transmission of IAV.

Subclinical infections of IAV occur in pigs and have been detected at 83.3% of agricultural fairs with IAV-infected swine (Bowman et al., 2012). This presents an additional challenge as IAV-infected swine cannot be identified by clinical signs during entry. Attempts to use infrared and rectal thermometers have also been unsuccessful for screening pigs of IAV (Bowman et al., in press). Ultimately, the IAV status of the swine can only be determined by diagnostic testing. Given the extensive spread of IAV within the population of swine at an exhibition, a viable method to detect the small percentage of IAV-positive pigs at entry and prevent their entry to the exhibition site would be ideal. However, this option is currently an unrealistic proposition because the labour, cost and time required to perform diagnostic testing are beyond the capacities of most exhibitions. Mitigation strategies to avert swine-to-swine transmission of IAV during the fair will likely decrease the total IAV burden in swine barns at fairs and reduce the risk of zoonotic IAV transmission.

For the present study, the ideal time point for sampling was on the trailer prior to unloading, as was performed at Exhibition A, as the swine were not yet exposed to the exhibition’s animals or environment. Given logistic complications, this approach was not feasible at the other exhibitions. The maximum time that swine were at Exhibition B through I prior to sampling was 24 h. Swine at exhibitions D and E were sampled in their pens after unloading but prior to weighing. The swine at the remaining exhibitions were sampled in the chute during weighing. The clustering of IAV positive exhibition swine at exhibitions B and C observed in Fig. 1a and b are similar; however, the amount of viable IAV recovered differed greatly between these two exhibitions. At Exhibition B, 34.8% of the samples were rRT-PCR positive for IAV, and 43 IAV isolates were recovered. In the case of Exhibition C, 4.5% of the samples were rRT-PCR positive for IAV, and one isolate was recovered.

Temporal clustering of positive swine within fairs was expected because swine from the same farm, predicted to have similar IAV exposure, would arrive at the exhibition together, be placed in related pens, and moved together through the chute. However, the large proportion of samples testing IAV positive with rRT-PCR and the unusual trailing off of C_t values observed at exhibitions B and C (Fig. 1a and b) was not anticipated. The most likely explanation for this trend is swine snout contamination during corralling activities. In this scenario, an IAV-infected pig deposits virus via oral and/or nasal secretions onto swine contact surfaces (i.e. gating, hurdle boards, scale walls, etc.) as it moves through the chute. Subsequent swine moving through the chute likely contact the contaminated surface (s) and acquire IAV on their snouts, thus testing positive for IAV. Over time, each pig passing through the chute following the truly infected pig likely removes some IAV from the contaminated surface(s). Therefore, IAV concentration on the surfaces, and thus the snouts contacting the contaminated surface(s), would be expected to diminish over time. Although genomic sequencing is needed to fully assess IAV isolate identities, the temporal clustering of identically subtyped isolates, as observed at Exhibition B (Fig. 1a), provides additional support for this hypothesis. One would expect a more dispersed distribution of subtypes across the sampling if the 43 IAV-positive swine at Exhibition B were in fact infected prior to arrival at the exhibition. We hypothesize that moving swine through a chute has the potential to expedite pathogen transmission during the exhibition because this practice allows many swine to be rapidly exposed to a variety of pathogens within hours of arrival. This deposition of IAV onto the snouts of swine during corralling likely results in greater IAV exposure for the swine population than if the virus had to spread pig-to-pig via direct contact. IAV can be transmitted between swine via fomites (Allerson et al., 2013); thus, it is possible that common contact surfaces within the chute (i.e. gates, handing equipment, walls, scale, etc.) could facilitate IAV spread. Corralling activities of swine, and similar situations during the exhibition, may be critical points for limiting IAV transmission at fairs. Future work is needed to assess IAV contamination of chutes and similar surfaces at fairs. Supporting these findings is the recent study by Choi et al. (2015), where IAV was recovered from railings at live animal markets, demonstrating that IAV isolates can be recovered from gates and railing areas that have high contact with swine. Attention to reducing contamination of these surfaces (e.g. frequent disinfection) may be warranted.

Other studies investigating IAV prevalence among incoming exhibition swine found discrepancies similar to what was observed between fairs in the present study. During 2008–2009, Gray et al. (2012) conducted a surveillance study at fairs in Minnesota and South Dakota. No IAV was detected in the swine sampled at the Minnesota fair in 2008; however, in 2009, 19.3% of the swine at the same Minnesota fair and 2.2% of the swine at a South Dakota fair were PCR positive for IAV. The lack of IAV found in 2008, during the study by Gray, can be explained by the year-to-year variability of IAV within fairs as previously demonstrated (Bowman et al., 2012). This study provided an insightful preliminary look at IAV among swine at exhibitions; however, the sample size was relatively small and timing of sample collection relative to arrival was loosely defined making on-sight IAV transmission prior to sampling possible. In a 2013 pilot study, Bowman et al. (in press) recovered IAV isolates from 2.4% of samples collected from swine on trailers prior to unloading at Exhibition A.

There are some limitations with the prevalence established during the present study. It was necessary, due to time and funding restraints, to screen samples with rRT-PCR before inoculating them for virus isolation. Therefore, all samples were subject to two freeze-thaw cycles prior to a virus recovery attempt. Freeze-thaw cycles have been associated with decreased IAV infectivity, and thus, the frequency of virus isolation in the present study is likely lower than if samples had been directly inoculated without refreezing during rRT-PCR screening (Greiff et al., 1954). Additionally, the snout wipe method used in the current study is less sensitive than the gold standard nasal swabs (Edwards et al., 2014). Snout wipes were chosen because they are non-invasive, do not require restraint and take less time to administer than nasal swabs. In the present study, exhibitors were very concerned about stress on their animal, even momentary stress; therefore, snout wipes were chosen to maintain a level of acceptance from the exhibitor to the sampling process. It is important to note that snout wipes are taken from the surface of the pig’s snout and may represent more of a conglomerated sample of the pig and its environment when compared with a gold standard nasal swab that requires pig restraint. The possible cross-contamination observed at exhibitions B and C creates the potential that the individual animal IAV prevalence was overestimated in the present study. Based on the IAV subtypes identified at Exhibition C and the results of the cluster analyses, we suspect at least three swine had active IAV infections during the time of sampling at Exhibition C.

Other potential bias in our prevalence estimate include the following: fair type, proximity to other swine shows, differences in county exhibition swine industry, weather conditions and the intrinsic differences between fair management. Eight of the agricultural fairs sampled were local county fairs, and one exhibition was a larger regional fair. The majority of the pigs attending the regional fair would have attended a county fair prior to the sampling in this study and thus may have had a previous exposure to IAV. Similarly, pigs raised in counties with multiple swine shows could have attended exhibitions prior to this study’s sampling and experienced previous IAV exposure. Some counties are also known to have a rich tradition of exhibition swine, with many families travelling to multiple swine shows throughout the year. While the weather condition between the nine fairs was not identical, it was similar, as all fairs were sampled in a 2-month period of July and August.

In conclusion, the current project demonstrated that a small number of swine arrived at exhibition actively infected with IAV. However, the estimated prevalence of IAV at individual fairs ranged widely and may in part be due to the slight difference in sampling procedures used at each fair. In particular, sampling that occurred in chutes resulted in a far greater prevalence of IAV when compared with sampling in pen or on a trailer. We hypothesize this to be the result of IAV contamination in the chute. If true, fomite transmission of IAV may occur whenever swine are moved through the barn such as during weighing, washing, walking to and from the show ring and at times when animals encounter contaminated sites. These activities could be heightened areas for IAV transmission between swine, and thus, potential time points and locations to target for controlling swine-to-human zoonosis. Increasing our knowledge about IAV introductions into the agricultural fair settings is important for controlling animal-to-animal IAV spread, which will in turn limit swine-to-human IAV transmission. These data provide a critical first step towards mitigation strategies that will limit zoonotic transmission and improve public health.