1. Introduction
Salmonella infections in poultry are a health risk to the birds and later on to humans as a foodborne pathogen. The consumption of turkey in the U.S. has increased 17-fold since 1909 [
1]. Scharff estimates
Salmonella species in poultry to cost approximately
$2.8 billion annually in the U. S. [
2]. Consumption of undercooked contaminated products can cause acute, severe gastroenteritis in humans [
3]. This straight, non-spore-forming Gram-negative rod can be quite challenging to control as the introduction to and maintenance within poultry operations is multifaceted. More than 2500 serovars exist within the subspecies
enterica with about 10% of those being found in poultry production [
3]. Groups are designated based on the somatic (O) and flagellar (H) antigens [
4].
Adult turkeys are often asymptomatic with
Salmonella infections but can transmit the bacteria through eggs to poults [
5]. Vaccination is available and often implemented in breeder turkey flocks as a means to control certain serovars that pose a higher risk to human health. A 2014 study utilizing data from another commercial production company in the US showed that 90% of turkey farms that sampled positive for
Salmonella via drag/bootie swabs tested positive for
Salmonella at the processing plant [
6], emphasizing the need to identify and mediate this pathogen prior to the production of a product that is destined for human consumption.
Incidences of
Salmonella in turkey flocks has been estimated at between 16% and 54% [
5]. The introduction of
Salmonella to the flock can occur from many sources since it is a bacterium transmitted both horizontally and vertically (internal and/or external contamination of eggs) [
5]. From the hatchery to the processing plant,
Salmonella is a constant threat to poultry production systems. The main modes of transmission include contaminated feed, biologic vectors (birds, rats, mice, mites, insects), contaminated water and mechanical vectors/fomites (trucks, personnel, equipment) [
5]. Transmission may occur between farms and between houses on a farm when on-farm biosecurity (use of personal protective equipment including bouffant, clean coveralls, clean boots or boot cover and gloves, disinfection of vehicle prior crossing established clean-dirty line of property, cleaning and disinfection of equipment used within multiple houses during visit, use of boot disinfection stations between entering houses, and following of all company protocols for personal visits to farms) is not being followed, as shown by several other studies that analyze the movement of other diseases such as avian influenza through the use of a network model [
7,
8]. Current literature specifically on the analysis of the movement networks and the dissemination of
Salmonella within turkey production is limited. In the present study, a contact network was created by following the movement of turkey flocks from their brooding location to their growout location. Following these pairs would enable the identification of brooder farm locations that were more or less likely to repeatedly test positive for
Salmonella at the surveillance sampling time point.
The transportation of live turkeys to the processing plant is the final opportunity to expose the birds to
Salmonella once they leave a growout facility [
9]. Proper cleaning and disinfection of transport coops is essential to ensure the minimal transmission of bacteria. The ability of
Salmonella to persist in products deemed for human consumption has forced processing plants to consider decontamination treatments as the easiest way to control this pathogen during processing [
3].
Multiple studies have shown the value of using contact networks paired with between-flock movement of personnel to implement improved surveillance and intervention to combat disease [
10,
11]. However, there is limited information on network analysis [
7,
10,
11,
12,
13,
14,
15] and disease transmission models in poultry systems [
7,
10]. Furthermore, no information is available in the literature on the transmission of
Salmonella among turkey flocks. In this study, we described the distribution of
Salmonella among the sites of one turkey producing company located in North Carolina and the spread along the movement of birds from the brooder to the growout farms’ network. This work is based on
Salmonella cases collected approximately two weeks prior to harvest from growout farms from 2017 to 2020. Additionally, we assessed the distribution by serotype and group.
3. Discussion
The purpose of this study was to identify contributing factors to the spread of Salmonella within an eastern U.S. turkey production system by analyzing the prevalence of positive samples collected during routine surveillance of growout flocks and between-farm bird movements in mediation of Salmonella dissemination. While the human-health impact of Salmonella is important, it was not evaluated in this study, as information on the impact of working in poultry production on each employee was not available. Rather, the focus was placed on evaluating the incidence and potential movements of this pathogen within this system. In this study, factors that were considered included rearing style (antibiotic-free vs. conventional) over time, impact of brooder location on growout location and age of the farm according to company documentation. Rearing style and seasonality had varying trends and the older a facility was the more likely it was to be positive for Salmonella on surveillance sampling in this system. Group C1 was the serogroup with the highest prevalence of Salmonella, with almost one third of all samples collected falling into this group, including the most frequently isolated serovar: Infantis. We have also demonstrated that the turkey farms in the evaluated production company are weakly connected through the movement of bird’s brooders directly into growouts. The Salmonella infected network showed a lower degree when compared with the entire network, additionally the cases of Salmonella did now show a clear association between movement of birds and positive growout sites. Despite the fact that, for Salmonella, the infected networks were similar to the non-infected network, we remark that the use of contact networks could become a useful tool to assess the movement-associated risk of other poultry diseases such as avian influenza and Mycoplasmosis. Infected networks were not more infected than farms in the whole contact network.
A 2010 study analyzing
Salmonella serotypes isolated from turkey cuts at different steps of processing at a processing plant identified 83% (101/122) of their isolates were
Salmonella Derby, rendering it the overwhelming majority [
16] from samples collected. In the present study, we demonstrated that the same serotypes were found in the live production stage, however at much lower rates, i.e.,
Salmonella Derby accounted for less than 1% (0.29%) of the positive samples collected. This suggests that certain serovars may be more likely to be harbored and introduced to incoming carcasses at the processing level rather than in live production. In comparison to broiler live production and processing, the most commonly isolated group is C3, especially
Salmonella Kentucky [
17]. The present study considering live production of turkeys did not have any positive samples from group C3. While there are many serovars of
Salmonella that affect both species, trends from these surveillance samples indicate that some may be more species-specific than others, including those that fall into group C3 being better adapted to chickens.
We found clear patterns in the occurrence of
Salmonella in ABF production with peaks in the second half of each calendar year. In contrast, a previous study, also from North Carolina, determined no association between
Salmonella and seasonality [
18] while we found trends in the occurrence of
Salmonella with peaks from September to November in ABF production as seen in
Figure 1 and
Figure 2. Analysis of
Supplemental Figure S2 shows peaks of positive samples in the months of October and November in years 2018, 2019 and 2020. The entrance of biological vectors into the poultry house is more likely to occur during this time of the year as these animals attempt to evade the cooler temperatures by seeking refuge in the climate-controlled poultry house that also has a readily available supply of food and water. This system determines rearing style (conventional vs. ABF) of each flock at each location based on market demands so this study is unable to assess
Salmonella prevalence solely based on the rearing style of each location, as each location grows turkeys under both styles.
Figure 2 shows an overall increase in the collected positive samples over time from ABF farms over the course of the four years. This trend was consistent with the goals of this production system, as the national demand for ABF turkey at the level of processing has increased over the four years analyzed. In the years 2018–2020 during the months of September through December, many more positive samples were collected from ABF flocks compared to conventional flocks. Conventionally reared flocks appear to show a decline in negative samples collected over the course of the four years, with peaks of positive samples collected inconsistently dispersed throughout the year with occasional peaks in summer months, especially during the years 2018 and 2019. During the summer, naturally ventilated, curtain-sided facilities will often utilize a fogger system to aid in cooling of the birds. These cooling systems can increase relative humidity in the environment 4–5% [
19], facilitating the creation of the ideal environment for bacteria growth [
20]. In summating all four years of production, this system has a higher prevalence of
Salmonella in ABF production flock compared to conventional production flock of over 10%, indicating that the use of some antimicrobials may aid in the reduction in
Salmonella even though therapy may not be targeted specifically at this pathogen.
Since
Salmonella can be horizontally transmitted within and between flocks, a flock transmitting the bacteria at the time of transfer would continue to pass it from bird to bird after arrival at the growout facility [
5]. The ability of a location to harbor a pathogen, specifically
Salmonella, was considered in this study by identifying flocks sampled at multiple growout locations that all stemmed from the same brooder location. Through our analysis we determined that the presence of
Salmonella at each growout location is independent of the brooder from which the flock originated within this operation. The brooder farms do not appear to be a source of spread for
Salmonella. Similarly, the proximity of one location of a farm to another did not appear to be a risk factor for testing positive for
Salmonella.
Salmonella was widely distributed across eight counties in which this system grows birds with clusters of farms in the same geographical area not necessarily consistently testing positive for the same serovars confirming that other methods of transmission (such as biologic vectors, fomites, contaminated feed) are a more important source of transmission than geographic proximity with airborne transmission via contaminated dust.
A study in a laying hen system determined that the older the infrastructure of the housing system, the more likely they were to test positive for
Salmonella [
21]. Similar results were found in this study consisting of all curtain-sided, naturally ventilated houses. This result is likely due to better overall infrastructure of newer houses with cleaner and modern construction materials that are designed with good structural biosecurity in mind. It would be expected that all curtain-sided, naturally ventilated houses have more physical openings compared to solid-sided, tunnel-ventilated houses, and consequently a higher chance for the introduction of pathogens via biologic vectors. Additionally, older houses are more likely to have structural deterioration that allows the entrance of biological vectors of
Salmonella including birds and rodents. It can be concluded that farm age is a risk factor for
Salmonella presence.
For the North America poultry industry, our results from the between-farm movement suggested that even though there were insignificant associations between movements and
Salmonella positive cases, key network centrality metrics such as degree could be considered in order to implement risk-based disease surveillance. The utilization of the ranked farms by total degree could be used to disrupt the contact network between infected and uninfected movement flows in the case of
Salmonella and also in the case of a foreign animal disease (FAD) epidemic [
7,
8,
10]. On the other hand, the extremely low level of clustering of the infected network compared to a random network is suggestive that
Salmonella might not maintain itself and spread within the poultry network. Therefore, other unmeasured or unknown transmission routes such as proximity of mortality disposal (incineration, on-farm composting) to houses, feed truck time at each facility and movement of company personnel should be further investigated while targeting between-farm
Salmonella control.
In an ideal study of a poultry production system in which
Salmonella prevalence and risk factors were being assessed, the management method of rearing turkeys (ABF vs. conventional) would be set by location in order to allow for an analysis of each type of style’s influence on the likelihood of
Salmonella presence. Similarly, our study focused on surveillance data from growout locations only. One study found the bacteria persist in the dust of a depopulated poultry facility up to one year after the birds were removed [
22]. Further work could build on the present study by collecting samples at the brooder level in order to determine if there is true maintenance and/or movement of the bacteria from one location to the next when the flock transfers at five weeks of age within this particular system. The lack of data collection from the brooder location is a limitation of this study as we cannot determine the origin of the bacteria in each flock. Moreover, a consideration of sampling the growout houses multiple times, to investigate whether there is “replacement” of
Salmonella serotypes over time would be ideal. It would also be beneficial to consider data from breeder farms in an effort to identify any vertical transmission of
Salmonella within the system. Future work should also include feed truck movement data along with information on movement of service and maintenance personnel. Litter age and management should be included as well.
Salmonella is a complex and ever-present pathogen within poultry production. While many factors exist that may be contributing to the movement of
Salmonella within poultry flocks, an analysis of these factors to identify the major risk factors can be accomplished.