Objective To evaluate whether pig farms interconnected within the same cooperative share similar Salmonella contamination patterns.
Setting Ten finishing pig farms within a 100 km radius of a common slaughterhouse were selected. Their inclusion was based on their association to the same cooperative and the sharing of common resources: piglets, feed, swine transporters, slaughterhouse, technicians and veterinarians.
Procedure Each farm was visited three times over a 10-month period. Pig faeces, the barn front door handle, the feed pipeline, mobile objects (shovel, balance and pig board), the landing stage, the concrete slab of the feed bins, the tire tracks left on the pathways by the animal feed truck, the pig delivery truck and the carcase knacker truck and the mudguards and cabin carpets of the veterinarian and technician vehicles on their arrival at the farm were all analysed for the presence of Salmonella.
Results All farms were not equally contaminated with Salmonella. Whereas some farms yielded up to 12 Salmonella isolates, other farms were Salmonella free. Some locations, most notably the landing stage, were more contaminated than others. Salmonella contamination was dynamic in time. Some contaminations seen on farms, on specific locations on the first visit, had disappeared on the second and third visits, but new contaminations were detected on different locations.
Conclusions Contamination with Salmonella was not disseminated through the network of the 10 pig farms interconnected within the same cooperative but was rather most often restricted in time to specific locations on specific farms.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, an indication of whether changes were made, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Salmonella is a genus of Gram-negative, rod-shaped bacteria and a member of the class γ-proteobacteria and the family Enterobacteriacae. It is the causal agent of salmonellosis, a zoonotic infectious disease characterised by abdominal cramps, diarrhoea, fever and vomiting. In Western countries, salmonellosis is the second most common foodborne disease. Whereas Salmonella enterica subsp. enterica serotype Enteritidis, referred to as S. Enteritidis, infection, is mostly associated to consumption of raw or undercooked eggs, S. enterica subspecies Typhimurium, S. Typhimurium, infections are more largely represented in pork, poultry, beef, dairy products and so on.1
Mitigation measures have been implemented at three levels: (1) at preharvest, the control of Salmonella in pig, (2) at harvest, through improved hygiene during slaughter and meat processing and (3) at postharvest, the preparation of food by the industry and the consumer.4
Several risk factors for the transmission of Salmonella to pig have been well studied,5 including the role of rodents as vectors for Salmonella,6 hygiene,7 8 lairage conditions,9 10 sources of animals,8 herd management,11 production system,12 feeds,13 antibiotics,14 parasite infestations15 and so on. However, none of these have focused on the comparative presence and dynamics of Salmonella on a network of farms interconnected within a common cooperative with shared common resources: piglets, feeds, swine transporters, a slaughterhouse, technicians and veterinarians. Here the presence of Salmonella is described in such a network of 10 pig farms, with three visits per farm over a 10-month period.
Materials and methods
A network of 10 finishing pig farms (A–J) was selected within a 100 km radius of their common slaughterhouse. Their inclusion was based on their association to the same cooperative that provided common resources: a breeding facility for the production of all piglets for its members, feeds, swine transporters, a slaughterhouse, and technicians and veterinarians.
Each farm was visited three times over a 10-month period, from August 2011 to May 2012. These three visits were done during two consecutive production batches and at the end of the growing period of the second batch, respectively. Inside the barn, five different pools of approximately 100 g of fresh faeces, each from three separate pens, were collected. Additional samples were obtained by swabbing four different locations or series of objects: the barn front door handle and a 20 cm × 20 cm area around it on both sides of the door, a 100 cm section of the feed pipeline, mobile objects (shovel, balance and pig board) and a 30 cm × 30 cm surface of the landing stage. Likewise, outside the farm, samples were collected by swabbing a 30 cm × 30 cm surface from four locations: the concrete slab of the feed bins and the tire tracks left on the pathways by the animal feed truck, the pig delivery truck and the carcase knacker truck. Thirteen samples were collected per farm per visit for a first subtotal of 390 samples (13 samples per visit × 10 farms × 3 visits).
Thirty additional samples (5 vehicles × 2 samples per visit × 3 visits) were collected by swabbing a 30 cm × 30 cm surface from mudguards and cabin carpets of the veterinarian and technician vehicles on their arrival at the farm.
In addition, on two separate occasions on all 10 farms, right before loading the pigs on the trucks for transportation to the slaughterhouse, samples were collected by swabbing a 30 cm × 30 cm surface of the mudguards and the empty truck box. Here, 40 more samples (10 farms × 2 visits × samples) were collected.
Finally, the efficiency of the pig transporter to eliminate the Salmonella contamination on selected locations on the trucks was measured twice, once in winter, once in summer, by swabbing a 30 cm × 30 cm surface from the mudguards, the empty box and the cabin carpet right after the cleaning, disinfection protocol. Here, 24 more samples (four transport trucks × two seasons × three samples) were taken.
In total, 484 samples were collected, chilled in ice, transported to the laboratory and analysed for the presence of Salmonella.
Isolation of Salmonella
Isolation of Salmonella was done according to The International Organization for Standardization ISO 6579:2002+Amd 1:2007.16 Briefly, each sample was pre-enriched in buffered peptone water and incubated at 37°C overnight (18–20 hours). Next, each culture from the pre-enrichment broth was inoculated in Rappaport-Vassiliadis medium with soya (RSV broth) and Muller-Kauffmann tetrathionate/novobiocin broth (MKTTn broth, pH 8.0). The RSV and MKTTn broths were incubated at 42°C and 37°C overnight, respectively. Cultures were plated on Xylose lysine deoxycholate agar (XLD agar) and incubated at 37°C overnight. In some cases, colonies on XLD agar were obtained from both RSV and MKTTn broths. In these cases, both colonies were further characterised. Identity of Salmonella isolates was confirmed by biochemical tests including Triple Sugar Iron agar reaction, Lysine Iron Agar and urease1 17 18 and API 20E (Biomérieux Canada, Montreal, Quebec, Canada) according to the manufacturer’s instructions. All confirmed Salmonella isolates were sent to ‘Laboratoire d’épidémiosurveillance animale du Québec’ (MAPAQ, Saint-Hyacinthe, Quebec, Canada) for serotyping.19S. Typhimurium isolates were further characterised by lysotyping20 at the National Microbiology Laboratory (Public Health Agency of Canada, Winnipeg, Manitoba, Canada). All comparisons of Salmonella isolate detection rates were done using Fisher’s exact test.21
Restriction fragment length polymorphism (RFLP)-Pulsed-field gel electrophoresis (PFGE) genotyping of Salmonella
RFLP-PFGE genotyping was carried on selected S. Derby strains according to the standardised CDC PulseNet protocol22–24 with the restriction enzymes XbaI and BnlI. DNA patterns were visually compared.
Distribution and identification of Salmonella on the 10 farms
Ten farms were visited three times over a 10-month period. The 390 samples collected on the farms were analysed for the presence of Salmonella. The bacterium was isolated on 9 of the 10 farms and in 39 of the 390 samples. Here, 42 Salmonella isolates were obtained (table 1). They encompassed seven different serotypes: Derby (14 isolates), Typhimurium,10 Infantis,8 Branderburg,6 Senftenberg,1 Agona1 and Mbandaka1 and one autoagglutinable isolate (a rough variant of Salmonella that did not express O antigen). Five different S. Typhimurium lysotypes, 12 (2 isolates), 104,1 193,4 208,2 and the monophasic S. typhimurium variant 1,4,,12:i-1 were detected.
Some farms yielded more Salmonella isolates than others. At one end of the spectrum, the three visits at farms F and E yielded 12 and 8 isolates, respectively, whereas at the other end (Fisher’s exact test; p<0.05), farm G was Salmonella free on all three visits for the samples tested.
Differential distribution of Salmonella over the three visits
In many cases, presence of Salmonella on objects and locations inside and outside the barn on a specific farm, A–J, was restricted to only one of the three visits. On farm A, during the first visit (A1), the feed pipeline yielded a Salmonella isolate, whereas Salmonella was not detected from this feed pipeline during visits 2 and 3 (A2 and A3). Here, however, Salmonella was isolated from the landing stage on farm A during visits 1 and 2 (A1 and A2), but not during visit 3 (A3). Interestingly, the first visit on farm E yielded two Salmonella isolates, S. Typhimurium lysotype 104 from the landing stage inside the barn and S. Agona from the tire tracks left on the pathway of the animal feed truck. However, the second visit on farm E yielded six isolates, four from the faeces, a fifth one from the mobile objects and a sixth one from the tire tracks left on the pathway of the pig delivery truck. The contamination seen on specific locations on the first visit had disappeared, but new contaminations were detected elsewhere. Here, five of the six isolates were S. Brandenburg, the sixth was S. Derby. No Salmonella was detected on the third visit on farm E. Somehow, S. Brandenburg was present on multiple objects and locations on farm E during visit two but was undetected during visits one and three. Interestingly, although all 10 farms were linked to the same cooperative and they all shared several common resources, S. Brandenburg was not detected on the other nine farms on each of the three visits, except on farm H on the first visit, and only on the landing stage inside the barn. Likewise, the first visit on farm F yielded two S. Infantis isolates – one from mobile objects, the other from the landing stage inside the barn. On the second visit on farm F, S. Infantis was again isolated but from faeces, pools 2–5 and again from the landing stage inside the barn. On the third visit, five Salmonella isolates were obtained from farm F, S. Typhimurium lysotypes 193 and 208 from the landing stage inside the barn, S. Infantis from tire tracks left on the pathway of the pig delivery truck and S. Typhimurium lysotypes 12 and 193 from tire tracks left on the pathway of the carcase knacker truck. Salmonella was not detected in the faeces and on the mobile objects. Interestingly, S. Infantis was not isolated from the other nine farms during all three visits possibly ruling out here systemic cross-contamination between farms interconnected within the same cooperative. However, S. Typhimurium lysotypes, 12, 193 or 208, were also isolated on the third visit on farms A and C but not on the other farms. Interestingly, S. Derby was detected on 6 of the 10 farms studied and on at least two successive visits on farms A, B and D. Although successive contaminations on one farm may question the efficiency in the local mitigation measures, the wider distribution of S. Derby isolates on this network of farms with 13 isolates/42 total isolates may suggest cross-contamination or contamination from a common origin within the network for this specific serotype.
Differential Salmonella contamination of locations and objects on the farms
Some locations and objects were more contaminated than others; they yielded more Salmonella isolates. Inside the barn, Salmonella was mostly isolated from the landing stage (14 isolates) (Fisher’s exact test; p<0.05) and to a lesser extent from the faeces (five samples on each farm yielded a total of 13 isolates), the mobile objects,5 the barn front door handle1 and the 100 cm section of the feed pipeline.1 The mobile objects are restricted to their respective farm and different Salmonella serotypes and lysotype were isolated indicating different origins. The mobile objects can also reveal the persistence of specific Salmonella on some farms, as exemplified by the presence of S. Derby on visits 1 and 2 on farm D. Outside the barn, the tire tracks left on the pathways by the carcase knacker truck (four isolates) and the pig delivery truck3 yielded the highest number of Salmonella isolates, followed by the tracks left on the pathways by the animal feed truck.1Salmonella contamination was not widespread but was rather restricted to some objects and locations, often during only one of the three visits, and on a small number of farms. Salmonella isolates covered different serovars, suggesting no or very limited cross-contamination between farms. No Salmonella was detected from the concrete slab of the feed bins.
Differential Salmonella contamination inside and outside the barn
Salmonella strains were isolated inside the barn (34 isolates) and outside.8 The identities of the Salmonella serotypes and S. Typhimurium lysotypes varied accordingly. Whereas most S. Brandenburg, S. Derby and S. Infantis were isolated inside the barn, S. Agona, S. Mbandaka and S. Typhimurium lysotype 12 were only isolated outside the farms.
Distribution and identification of Salmonella from the mudguards and cabin carpets of the veterinarian and technician vehicles
Of the 30 samples tested from truck mudguards and carpets from the driver’s cabin of the veterinarian and technician vehicles, a single mudguard sample tested positive and contained S. Derby.
Distribution and identification of Salmonella from selected locations on the pig transportation trucks
Of the 40 samples tested from the empty box and the mudguards of the transportation trucks, only three box samples were positive for Salmonella. They contained S. Derby, S. Agona and S. Brandenburg, respectively. The 20 truck mudguard swabs did not reveal the presence of Salmonella.
Distribution and identification of Salmonella from selected locations on the pig transportation trucks after the cleaning, disinfection protocol in winter and summer
The transporter’s efficiency in eliminating Salmonella from its truck with a cleaning, disinfection protocol was assessed. In winter, immediately after cleaning and disinfection, no Salmonella was detected in all parts sampled: the truck empty box, the cabin carpet and the mudguard. In summer, however, a positive sample was obtained from a truck box and identified as S. Brandenburg, and two more samples were obtained from a second truck, the cabin carpet and the mudguard and identified as S. Derby and S. Typhimurium, respectively.
RFLP-PFGE genotyping of Salmonella
The genetic diversity of a subset of our S. Derby isolates was characterised by RFLP-PFGE. Interestingly, all S. Derby isolates analysed shared similar XbaI or BnlI profiles (figure 1). At least 44 different S. Derby PFGE XbaI and BnlI profiles are known.25 Here, despite the different sources of our S. Derby isolates, different farms, different visits, their similar PFGE profiles for both enzymes strongly suggest a single origin. This origin is yet unknown. The distribution of this specific S. Derby pulsovar outside this network of 10 pig farms is also unknown.
We have shown here that Salmonella can occasionally be isolated from a network of farms linked to a same cooperative. Salmonella can be isolated inside the barn, on the landing stage, in faeces and on some mobile objects. The landing stage was the location with the highest number of swab samples positive for Salmonella across all farms. Regrettably, this location is barely mentioned in other studies on Salmonella contamination in swine farms. Yet, landing stages are the first contact of piglets when entering the farm and the fattening barn. Naïve pigs in contact with this contaminated environment could rapidly become Salmonella positive and shedders.26 27 Interestingly, S. Derby was mostly isolated from the landing stages across a number of different farms and visits, with exceptions, and it was always found inside the barn, not outside. We showed that these S. Derby isolates are genetically similar and are presumably of clonal origin. Clearly, our study indicates that a swine farm cleaning and disinfection protocol should include a thorough cleaning of the landing stages. Conversely, most other Salmonella serotypes and lysotypes isolated in our study were mostly restricted to a single farm. This is interesting given that the 10 farms under study consisted of a network interconnected within the same cooperative. Here, the presence of specific Salmonella restricted to a single farm on a single visit indicates the efficiency of the mitigation measures in cleaning and preventing the persistence of Salmonella.
Outside the barn, the tire tracks left on the pathways by the animal feed truck, the pig delivery truck and the carcase knacker truck can also carry Salmonella suggesting that the tires of either truck could be contaminated, thus potentially bringing contamination from the outside to the farm, or alternatively, the tires could become contaminated at the farm and carry Salmonella elsewhere, perhaps to another farm or to the slaughterhouse, both components of the network here. This is not what we saw, with an absence of common Salmonella serotypes or lysotypes on tire tracks on other farms. Salmonella dynamics associated to transport appeared limited.
At the farm, mudguards and carpets from the driver’s cabin of the veterinarian and technician vehicles did not appear as a source of Salmonella contamination. Likewise, of the items tested on the transportation trucks, the mudguards did not appear as a source of contamination, but the empty box may, at times, be one. Here also, a cleaning and disinfection protocol should prioritise a thorough cleaning of the truck box. Several other parts of the transportation truck were not tested for the presence of Salmonella: the carpet inside of the cabin among others. Given that these carpets are in contact with boots that have walked several pathways on the pig farm, outside and inside the barn, and elsewhere on the network, including the slaughterhouse, it is likely they come in contact with Salmonella and potentially become carrier. Other parts of the transportation trucks, yet to be characterised, could also be contaminated with Salmonella.
In conclusion, our work shows that good biosecurity standards28 were well implemented on the 10 farms studied here, interconnected within a common cooperative, and have proven effective, with the possible exception of the pig landing stage inside the barn on some farms where more attention should be paid.
We are planning to follow-up on this work by studying the presence of Salmonella at the pig slaughterhouse from this same cooperative. Unusual objects and locations, yards, trucks, mobile objects and so on will be sampled. Salmonella Derby isolates will be characterised by RFLP-PFGE and the profiles compared with the ones revealed here. One of our goals includes the study of the distribution of Salmonella in the Canadian swine industry to help control potential disease outbreaks.
Contributors The study was designed by AH, AL and PF. All authors made substantial contributions to the acquisition of data, their analysis and interpretation. The manuscript was drafted by J-CC, AH and PF. All authors revised the manuscript critically and gave final approval of the version to be published.
Funding This work was supported by the Canadian Swine Research and Development Cluster #1002 and by Complexe de diagnostic et d'épidémiosurveillance vétérinaires du Québec.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.