Prevalence of Salmonella spp. in environmental samples from table egg barns in Alberta

ABSTRACT

Some Salmonella spp. are zoonotic, a frequent cause of foodborne illness in Canada, and known to infect humans through contaminated poultry and poultry products. Certain serotypes of Salmonella spp. have been demonstrated to be vertically transmitted from hen to egg. The incidence of Salmonella spp. isolation in the flock has been correlated to its isolation from the environment. Twenty-one producers were enrolled in this study to examine the occurrence of Salmonella spp. in 48 table egg layer flocks housed in 35 barns in Alberta. The purpose of this study was to: (i) identify Salmonella serotypes isolated from the environment of table egg layer facilities in Alberta and (ii) record the prevalence of Salmonella spp. across eight defined environmental sampling points. Salmonella spp. were isolated from the environment of 20/35 barns representing 29/48 flocks. The most common serotypes isolated were S. Heidelberg, S. Kentucky and S. Mbandaka. The order of most to least contaminated sample location was manure belts (54.1%), feeders (47.9%), feed motors (45.8%), egg belts and walls (41.7%), fans (35.0%), cage bottoms (31.3%) and lobbies (27.1%). Salmonella spp. were isolated from 7/7 barns post cleaning and disinfection, demonstrating the persistence of this organism in the environment and the need for effective eradication protocols.

KEYWORDS:

• Prevalence
• Salmonella
• environment
• serotypes
• Alberta

Introduction

Foodborne illness is believed to be underreported and it is estimated by the Public Health Agency of Canada (PHAC) that 11,000,000 people in Canada acquire this condition every year (PHAC, 2008). Improper food handling and preparation by the consumer is considered to be the primary cause of foodborne illnesses (Scott, 2003). In Canada, non-typhoidal Salmonella are second only to Campylobacter as the most commonly identified bacterial foodborne pathogen (PHAC, 2011). In Alberta, during 2012, 846 (21.68 cases/100,000) confirmed human Salmonella spp. isolates were reported (PHAC, 2012). One source of human non-typhoidal Salmonella infection is poultry and/or poultry products (Lee & Middleton, 2003; Hennessy et al., 2004; MacDougall et al., 2004). As such, control and prevention of Salmonella contamination in poultry and poultry products is paramount for this industry.

Isolating Salmonella from the environment is reported to be correlated to Salmonella infection in the flock (Riemann et al., 1998; Carrique-Mas et al., 2008; Mahé et al., 2008; Arnold et al., 2010). Sampling of the environment has been shown to be a reliable indicator of the Salmonella status of the flock as long as sampling locations are carefully selected (Davies & Wray, 1996).

It has been demonstrated that dust from poultry houses is an excellent environmental sample as Salmonella spp. outcompete rival bacteria (Davies & Breslin, 2001). Gole et al. (2014) found the highest prevalence of Salmonella spp. was isolated from dust samples, followed by samples taken from the egg belt and faeces. In addition to the dust associated with the feeders and feed motors, the feeding system may increase access to the feeders by vermin, such as mice, which are a known source of Salmonella infection (Henzler & Opitz, 1992; Davies & Wray, 1995a; Guard-Petter et al., 1997; Liebana et al., 2003). A study involving commercial broilers demonstrated that the faeces were the sample most often to be contaminated with Salmonella spp. (Marin et al., 2011).

In this study, eight specific locations in the barn were chosen based on the theoretical likelihood of isolating Salmonella. Four of the locations mirrored the locations sampled by the Egg Farmers of Alberta (EFA) in their Salmonella Enteritidis program: walls, egg belts, cage bottoms and fans. In addition, three more locations: feed motors, feeders and manure belts were selected to sample. Finally, to assess the potential for contamination outside the bird handling area itself, the lobby was sampled.

In Canada, there are several commercial systems utilized for poultry housing, including caged (conventional and enriched) and cage-free (free-run, aviary and free-range) systems. All 35 barns in this study housed birds in cages; 34 barns with conventional cages and one with enriched cages.

A joint investigation between Alberta Agriculture and Forestry (AAF) and the Egg Farmers of Alberta was undertaken to identify why S. Heidelberg, and possibly other Salmonella spp. were isolated with increased frequency between 2003 and 2009 (AAF, unpublished data). The first objective of this study was to identify Salmonella serotypes isolated from the environment of table egg layer facilities in Alberta. The second objective was to record the prevalence of Salmonella spp. across eight defined environmental sampling points. Within objective two, the efficacy of cleaning and disinfection was further evaluated for a subgroup of the facilities tested.

Materials and methods

Study design

The sampling component of the study commenced in September 2011 and continued until December 2012. In all, 21 table egg producers throughout the province of Alberta participated.

Environmental samples were collected in 35 barns from 44 layer flocks (333,596 birds). Of the 35 barns, 28 housed multiage flocks (14/21 producers). In addition, four of these flocks were sampled again at a different point in the production cycle during the collection period. The average flock size was 7144 hens.

Environmental sampling

The sampling protocol consisted of taking environmental samples from eight defined locations: egg belts, fans, walls, cage bottoms, manure belts, feeders, feed motors and lobby within the barn. The environmental sampling swabs consisted of four folded sterile 3 × 3 cotton swabs soaked in buffered peptone and placed in a sample vial using a clean pair of disposable gloves. Specimens were collected by wiping the swab along strips of the specific sampling area. Each of the eight defined locations had four swabs used per location (total of 32 swabs per barn). Swabs were then submitted to AAF Agri-Food Laboratories Section for culture for the presence of Salmonella. In some barns, due to inaccessibility of ceiling fans, only seven locations were sampled. To assess the efficacy of cleaning and disinfection by the producer, five barns were selected to be sampled again post cleaning and disinfection; with two of these barns sampled twice (separate flocks).

Microbiology

Salmonella culture was performed using Agri-Food Laboratories Section Standard Operating Procedure (SOP) BA-0266; an ISO 17025 accredited test. Briefly, sterile forceps were used to remove swabs from each vial and to place them into a Whirl-pak^® bag containing 50 ml of buffered peptone water. The four swabs were pooled per vial to constitute one sample. Following the 20–24 h. incubation at 35 ± 2°C, 0.1 ml and 1.0 ml of the incubated buffered peptone were transferred to 10 ml Rappaport Vassiliadis (RV) broth and 10 ml Tetrathionate (TT) broth with 0.2 ml iodine, respectively. The RV broth was incubated at 42 ± 1°C for 22–24 h. and the TT broth was incubated at 35 ± 2°C for 22–24 h. Real-Time Polymerase Chain Reaction (RPCR) was then used to screen pooled samples (0.15 ml each) of the TT and RV broths. In addition, 0.3 ml of the TT broth was transferred to a Modified Semi Solid Rappaport Vassiliadis (MSRV) agar plate, a selective motility media, and further incubated at 42 ± 1°C for 48 h.

RPCR positive samples and MSRV positive samples were inoculated onto selective/differential agar plates Xylose Lysine Terigitol 4 agar and Rambach agar. Colonies exhibiting characteristic morphologies were biochemically and serologically identified as Salmonella. Presumptive colonies were sent to the Alberta Provincial Laboratory for Public Health (ProvLab, Calgary, AB, Canada) for serotyping.

Serotyping

Isolates were serotyped by the ProvLab using conventional methods (Ewing, 1986), based on the White-Kauffmann-Le Minor scheme (Grimont & Weill, 2007), to determine the O (somatic) and H (flagellar) antigens. Slide agglutination was used with Statens Serum Institut antisera (SSI; Copenhagan, Denmark, UK) to determine O (somatic) antigens while tube agglutination was used with Difco™ antisera (BD, Sparks, MD, USA) to determine H (flagellar) antigens.

Real-time PCR for detection of Salmonella spp.

DNA extraction

Aliquots of 150 µl each of the TT and RV broths from each sample were combined in a microfuge tube containing 700 µl of sterile, purified water. The samples were briefly mixed on a vortex mixer and centrifuged at 18,000×g for five minutes. The supernatants were discarded, and the pellets were re-suspended in 100 µl of sterile, purified water. The DNA was extracted with a semi-automated magnetic particle processor (KingFisher, Thermo Electron Corporation, Vantaa, Finland) and DNA extraction kit (Magnesil KF genomic system kit, Promega, Madison, WI, USA) according to the manufacturers’ instructions. A positive culture control sample was used as an extraction control with each assay.

Real-time PCR

Primers and dual-hybridization probes targeting the Salmonella invA gene were used as previously described (Bohaychuk et al., 2007). Briefly, RPCR was conducted in a 20 µL volume consisting of final concentrations of 4 mM MgCl₂, 0.5 µM of each primer (LC-Sal 1 5ʹ-AACTTCATCGCACCGTCA-3ʹ and LC-Sal 2 5ʹ-TATTGTCACCGTGGTCCAG-3ʹ), 0.15 µM of each probe (LC-Sal probe 1 5ʹ-CGCATCAATAATACCGGCCTTCA-FL-3ʹ and LC-Sal probe 2 5ʹ-Red610-TCGGCATCAATACTCATCTGTTTACCG-3ʹ and LC-Sal IC probe 1 5ʹ-CGTTGGCTACCCGTGATATTGC-FL-3ʹ and LC-Sal IC probe 2 5ʹ-Red670-GAAGAGCTTGGCGGCGAAT-3ʹ), 1X LC Master Hyb solution (Roche, Mannheim, Germany) and 0.8U Uracil DNA glycosylase (New England Biolabs, Mississauga, Ontario, Canada). Four microliters of extracted DNA was added and the reactions were amplified using an initial denaturation cycle at 95°C for 10 min followed by 50 cycles [95°C for 5 s; 54°C for 10 s; 72°C for 15 s] and a final cooling step of one cycle at 40°C for 30 s. Results were interpreted as positive or negative, depending on the shape of the curve and the crossing point provided by the LightCycler analysis software. A positive control of genomic DNA from Salmonella Typhimurium ATCC 14028 and a negative template control (water) were included in each assay.

Results

Salmonella serotypes

The number of Salmonella serotypes isolated from the environmental swabs was recorded (Table 1). Of the 48 flocks tested, 29 flocks had Salmonella spp. isolated from the environment. The majority of these (16/29) had only one serotype isolated (Table 1), but in 13/29 flocks there were two to four serotypes per flock identified. Salmonella spp. isolated and the frequency of the 49 isolates is shown in Table 2. Each isolate indicates that the serotype was isolated from one sampling visit (per flock) and not the frequency (eight sampling sites). As indicated in the table, the most common isolates were S. Heidelberg and S. Kentucky.

Table 1. Salmonella serotypes isolated from the environment of 48 flocks and the frequency of each serotype.

Number of flocks (n = 48)	Number of serotypes isolated	Serotypes isolated (frequency)
19	0	n/a
16	One serotype isolated	Kentucky (6) Heidelberg (4) Mbandaka (2) Agona (1) Braenderup (1) Schwarzengrund (1) Tennessee (1)
8	Two serotypes isolated	Kentucky, I 8,20:-:z6 (1) Kentucky, I 4,5,12:i:- (1) Kentucky, Heidelberg (1) Heidelberg, I OR:r:1,2 (1) Heidelberg, Braenderup (1) Seftenberg, Schwarzengrund (1) Alachua, Mbandaka (1) Mbandaka, Agona (1)
3	Three serotypes isolated	Heidelberg, Alachua, Agona (1) Alachua, Mbandaka, Agona (1) Braenderup, Mbandaka, Hadar (1)
2	Four serotypes isolated	Mbandaka, I O Rough:H undetermined, Agona, Hadar (1) Mbandaka, Agona, Alachua, Heidelberg (1)

Table 2. Frequency of isolation of various Salmonella serotypes from one sampling visit.

Salmonella serotype	Number of isolates (n = 49)
Heidelberg	9
Kentucky	9
Mbandaka	8
Agona	6
Alachua	4
Braenderup	3
Hadar	2
Schwarzengrund	2
Senftenberg	1
Tennessee	1
I 4,5,12:i:-	1
I OR:r:1,2	1
I 8,20:-:z6	1
I O Rough:H undetermined	1

Sampling locations

Salmonella spp. isolation from each barn location is reported in Table 3. There were 48 samples taken for each location except for the fans, where due to inaccessibility, only 40 samples were collected. The order of most to least contaminated sample location was manure belts (54.1%), feeders (47.9%), feed motors (45.8%), egg belts and walls (41.7%), fans (35.0%), cage bottoms (31.3%) and lobbies (27.1%).

Table 3. Isolation of Salmonella spp. from barn sample locations before and after cleaning and disinfection (C & D).

	Number of samples positive over all barns sampled before C & D		Number of samples positive before and after C & D
Barn sample location	Positive samples/Total samples	Percent	Pre C & D Positive samples/Total samples	Percent	Post C & D Positive samples/Total samples	Percent
Manure belt	26/48	54.1	6/7	85.7	7/7	100.0
Feeders	23/48	47.9	7/7	100.0	6/7	85.7
Lobby	13/48	27.1	3/7	42.9	6/7	85.7
Egg belt	20/48	41.7	5/7	71.4	5/7	71.4
Feed motors	22/48	45.8	7/7	100.0	4/7	57.1
Cage bottoms	15/48	31.3	4/7	57.1	4/7	57.1
Walls	20/48	41.7	6/7	85.7	3/7	42.9
Fans	14/40	35.0	4/6	66.7	0/6	0.0

Isolation of Salmonella post cleaning and disinfection

Table 3 also includes Salmonella spp. isolation from each location following cleaning and disinfection of a subset of barns. As indicated, there was a higher percentage of Salmonella spp. positive samples isolated from the manure belt and the lobby after cleaning and disinfection of the barn. Isolation rate remained the same for egg belt and cage bottoms, and declined on feeders, feed motors, walls and fans.

Comparison of serotypes isolated before and after cleaning and disinfection

Salmonella serotypes isolated before and after cleaning and disinfection of barns are shown in Table 4. There were three producers that were visited once and two producers that were visited twice (two flock cycles). Salmonella spp. were isolated from all barns sampled following the cleaning and disinfection process. Looking at the distribution of Salmonella serotypes isolated before and after cleaning and disinfection, the identical serotype pattern was isolated in five of seven instances (Table 4). In two instances, there was a reduction in the number of serotypes.

Table 4. Salmonella serotypes isolated from environmental samples collected from table egg laying barns before and after cleaning and disinfection.

Sampling event	Visit	Serotypes isolated before cleaning and disinfection	Serotypes isolated after cleaning and disinfection
1	One	S. Heidelberg, S. Kentucky	S. Heidelberg, S. Kentucky
2	One	S. Agona, S. Mbandaka	S. Agona, S. Mbandaka
3	One	S. Braenderup, S. Heidelberg	S. Heidelberg
4	One	S. Heidelberg	S. Heidelberg
5	Two	S. Heidelberg	S. Heidelberg
6	One	S. Alachua, S. Mbandaka	S. Alachua, S. Mbandaka
7	Two	S. Agona, S. Heidelberg, S. Mbandaka, S. Alachua	S. Agona

Discussion

In 2011, as part of the Egg Farmers of Alberta Salmonella Enteritidis testing program, environmental samples were submitted to AAF for Salmonella culture. Of these, 132/195 were positive for Salmonella spp. with 23 different serotypes identified. The most frequently isolated serotypes were S. Kentucky (42/132), S. Mbandaka (13/132), S. Montevideo (7/132), S. Heidelberg (6/132) and S. Rissen (6/132) (AAF, unpublished data). With the exception of S. Heidelberg and S. I 4,5,12:i:- (1/132), these serotypes are identified very infrequently as causes of morbidity in humans in Canada (PHAC, 2011, 2012). Of particular importance are those that have been clearly demonstrated to be vertically transmitted, such as S. Heidelberg (Gast et al., 2004, 2005, 2007a, b) and have the potential to infect market table eggs.

This study analysed data collected from table egg-laying facilities (21 different producers) in the province of Alberta from 2011 to 2012. In all, the environment of 48 flocks was sampled and Salmonella spp. was found in 20/35 barns (note: some barns raised separate flocks of two ages). None of the isolates were S. Enteritidis. Foley et al. (2011) found that with the elimination or reduction in the prevalence of S. Enteritidis, there was a corresponding increase in S. Kentucky and S. Heidelberg in poultry facilities, and this same phenomenon could be occurring in Alberta table egg-producing facilities. Along with S. Kentucky (9/49), S. Heidelberg was the most frequently isolated serotype (9/49), as shown in Table 2; the high frequency of S. Heidelberg isolation from poultry has been reported elsewhere (Guerin et al., 2005; Foley et al., 2011).

In 2011, S. Heidelberg comprised 4.6% of all positive Salmonella spp. environmental isolates from commercial poultry facilities in Alberta (AAF, unpublished data) and this number reflects the third year of a declining trend of this serotype (13.8% in 2008 and 2009; 6.1% in 2010). S. Heidelberg from broiler chicken carcasses in the US has also been declining in proportion (12.96% in 2008, 14.07% in 2009, 3.49% in 2010 and 2.76% in 2011) to all Salmonella spp. isolates (USDA, 2014).

The presence of S. Heidelberg as a single isolate (Table 1), or in combination with other serotypes, and its propensity to cause disease in humans suggests that this type of Salmonella spp. could have active genes for both survivability and for pathogenicity as shown by Gokulan et al. (2013). S. Heidelberg has been demonstrated to be vertically transmitted in poultry (Gast et al., 2004, 2005, 2007a, b), which provides a route of contamination in eggs. As such, poultry and poultry products are a source of S. Heidelberg for humans, and these have been reported as a cause of salmonellosis in humans (Hennessy et al., 2004; MacDougall et al., 2004). Salmonellosis in humans due to S. Heidelberg was reported to be 9.4% (PHAC, 2011) and 15.0% (PHAC, 2012) in Canada and 2.3% in 2012 in the US (Centers for Disease Control and Prevention [CDC], 2014).

In 2011, S. Kentucky was the most common serotype (29%) of all positive Salmonella spp. environmental samples isolated from commercial poultry facilities in Alberta (AAF, unpublished data) and one of the most common (24.9%) in Ontario (Sivaramalingam et al., 2013). In 2011, S. Kentucky was also isolated frequently (51.84%) from broiler chicken carcasses in the US (USDA, 2014).

No previous reports have indicated that S. Kentucky is vertically transmitted in chickens. Despite its high incidence in poultry, this serotype is not commonly isolated from humans. Salmonellosis in people in Canada was reported to be 0.26% in 2011 (PHAC, 2011), 0.45% in 2012 (PHAC, 2012) and 0.8% in the European Union in 2012 (European Food Safety Authority [EFSA], 2015), but co-morbidity associated with this serotype was not reported. Although this serotype has been reported to cause gastroenteritis in a few patients, S. Kentucky did not result in invasive disease or death (Jones et al., 2008). There is an association between high antimicrobial resistance and greater pathogenicity in Salmonella spp. (Rosengren et al., 2009). S. Kentucky has been shown to have a high degree of antimicrobial resistance (Fricke et al., 2009; Lestari et al., 2009; FDA, 2012), although its antimicrobial resistance profile from Alberta poultry has not yet been determined.

In this study, S. Kentucky was more often present as a single serotype (Table 1) and the reason why warrants further investigation. It is possible that this may be linked to pathogenicity factors either present in, or active in, S. Kentucky (Fricke et al., 2009).

Salmonella Mbandaka was isolated at a high frequency (8/49) as shown in Table 2. Similar to S. Heidelberg, this serotype was isolated as a single Salmonella or with one to three additional serotypes (Table 1). In 2011, S. Mbandaka comprised 2.7% of all positive Salmonella spp. environmental isolates from commercial poultry facilities in Alberta (AAF, unpublished data) and 0.92% in 2011 from broiler chicken carcasses in the US (USDA, 2014). This serotype has been reported to cause salmonellosis in humans at low levels of 0.31% and 0.46% respectively in Canada (PHAC, 2011, 2012).

Three other serotypes, S. Agona (6/49), S. Alachua (4/49) and S. Braenderup (3/49) were isolated (Table 2), often in combination with other Salmonella serotypes (Table 1). The incidence of human salmonellosis is relatively low for these and the remaining eight serotypes isolated with the exception of S. I 4,5,12:i:-. Salmonellosis due to the latter serotype was reported to be 3.2% and 4.0%, respectively, in Canada (PHAC, 2011, 2012), 7.3% in 2012 in the European Union (EFSA, 2015) and 4.3% in 2012 in the US (CDC, 2014).

The second objective of this study recorded the prevalence of Salmonella spp. across eight defined environmental sampling points. The most common location Salmonella spp. were isolated from was the manure belt (Table 3) where the level of contamination was 54.1% over all the barns tested. As Salmonella are shed in faeces, its isolation with high frequency from the manure belt was not surprising. Interestingly, in this study, none of the manure belts were cleaned and disinfected sufficiently to eliminate Salmonella. The location of manure belts under the cage systems and the material of the belt may have impeded the ability to effectively wash and disinfect the belt. This was one of two locations where the ability to isolate Salmonella spp. from those barns selected for sampling was higher following cleaning and disinfection (100%) versus prior to cleaning and disinfection (85.7%). A possible explanation may be that during the washing procedure debris from cages and other sites dropped onto the manure belts, thus contaminating an otherwise non-contaminated location. This research is consistent with others that have also reported an increase in Salmonella spp. following cleaning and disinfection (Kradel & Miller, 1991; Davies & Wray, 1995b).

The manure belts were made of material that is relatively smooth and non-porous (plastic or vinyl), thereby reducing surface area available for bacterial colonization (Stocki et al., 2007). Wear and tear over time, however, results in damage to the surface area of the belt and provides an opportunity for bacterial attachment. Although the belts were dry by the time of sampling, evidence of water pooling existed (water staining), which also may have contributed to the survival of Salmonella spp. in this location (Eriksson de Rezende et al., 2001; Kieboom et al., 2006; Payne et al., 2007). In addition, it has been demonstrated that hard surfaces such as stainless steel are easier to clean versus materials such as plastic (Tolvanen et al., 2007).

The accessibility of certain areas/equipment for effective cleaning and disinfection may explain the differences in re-isolation of Salmonella spp. (Table 3) and the reduced contamination rates on fans (66.7–0%) and walls (85.7–42.9%) following cleaning and disinfection (initially 35.0% and 41.7% over all the barns tested). Fans were typically located on the sides of the buildings, which, along with the walls of the buildings, make these locations conducive to thorough cleaning through debris removal. In addition, upright walls and fans are less likely to allow standing water to pool and therefore dry more rapidly than other equipment/surfaces in the barn. Reduced water activity has been demonstrated to reduce Salmonella spp. survivability (Eriksson de Rezende et al., 2001; Kieboom et al., 2006; Payne et al., 2007).

The ability to isolate Salmonella spp. from the cage bottoms over all the barns tested was lower (31.3%), compared to six of the seven other sampling locations (Table 3). Those barns that were selected for sampling before and after cleaning and disinfection had the same percentage of positive cage bottoms (57.1%) before and after cleaning and disinfection. A similar isolation rate from the cage bottoms to the manure belt was expected because the manure passes through the cage wire bottoms. The cage wire architecture, however, made thorough sampling of the cage bottoms challenging. This may explain the relatively low incidence of Salmonella spp. isolation compared to the manure belt and also the failure to eliminate contamination from this site.

It is common to isolate Salmonella spp. from the egg belt in table egg barns (Stocki et al., 2007). The present study found that 41.7% of all the egg belts sampled (Table 3) were contaminated with Salmonella spp. Those barns that were selected for resampling had the same percentage of positive egg belts (71.4%) before and after cleaning and disinfection. As indicated elsewhere (Stocki et al., 2007), the type of material that is used to manufacture the egg belt can have an effect on the ability to remove Salmonella spp. All egg belts sampled in the present study were made of similar material and design.

The present study found that 47.9% of feeders and 45.8% of feed motors over all the barns sampled were contaminated with Salmonella (Table 3). In those barns selected for sampling before and after cleaning and disinfection, all the feeders and feed motors were positive for the isolation of Salmonella spp. before cleaning and disinfection. There was a reduction of Salmonella spp. isolated after cleaning and disinfection to 85.7% and 57.1% for the feeders and feed motors, respectively. These locations could be prone to Salmonella spp. contamination because the dust and the composition of the feed provide an environment for biofilm production (Donlan & Costerton, 2002). The configuration of the feeders with recessed areas, and in some cases chains, may increase the complexity of cleaning and allow for pooling of water. Conversely while cleaning of the feed motors requires care to prevent corrosion and avoid electrical shorts, it is less likely to allow pooling of water.

The ability to isolate Salmonella spp. from the lobby (27.1%) was lowest over all the locations sampled (Table 3). The lobby was the second location where the ability to isolate Salmonella spp. was higher following cleaning and disinfection (85.7%) versus prior to cleaning and disinfection (42.9%). Due to their office nature, the lobbies are not cleaned and disinfected in the same manner as the barns. It is possible that this area was re-contaminated following and/or during barn cleaning. This high incidence of lobby contamination could then be a source of contamination to the barn when flock production has resumed.

Table 4 shows that in 5/7 sampling events, the identical serotype pattern was isolated before and after the cleaning and disinfection process. There have been several other reported instances where Salmonella spp. were isolated from poultry facilities following cleaning and disinfection (Davies & Breslin, 2003; Wales et al., 2006; Carrique-Mas et al., 2009; Mueller-Doblies et al., 2010). The reason for continued isolation of Salmonella spp. could include inadequate or inappropriate application of detergent and/or disinfectants (Corcoran et al., 2014). Under these conditions, Salmonella spp. may develop chemical resistance (Randall et al., 2007; Condell et al., 2012; Corcoran et al., 2014) and form or contribute to the development of biofilms (Lapidot et al., 2006; White et al., 2006; Kim & Wei, 2007; Díez-García et al., 2012; Lianou & Koutsoumanis, 2012; Wang et al., 2013).

This study provides information on the prevalence of Salmonella spp. in table egg facilities in Alberta. Salmonella spp. were isolated from the environment of 29/48 flocks with 19/48 flocks Salmonella spp. negative. S. Heidelberg, S. Kentucky and S. Mbandaka were the most commonly detected serotypes; 14 serotypes were isolated in all. Further studies with respect to environmental survivability of Salmonella spp. and intra serotype competition would be of value.

Of the eight locations sampled, the manure belt was the most common location where Salmonella spp. were isolated from. In facilities where no Salmonella spp. were isolated, cleaning and disinfection by the producer is likely effective, or biosecurity to prevent the initial introduction of Salmonella spp. into the facility was successful. This study supports the premise that effective cleaning and disinfection of the barn must be part of a complete biosecurity program.

Acknowledgements

We acknowledge the contribution of the Egg Farmers of Alberta, particularly Christina Robinson, Susan Gal and the producers who took part in the study.

The authors would also like to thank Karen Wickerson, Karen Chrapko, Ludovic Silasi, Laura Elliott and Norma Pronteau for their technical support, Jagdish Patel for his constructive and helpful comments on this manuscript, Narine Singh for his dedication and expert advice, and the Enteric Department, at the Alberta Provincial Laboratory for Public Health, for Salmonella serotyping. Finally, thank you to the technical staff in the Agri-Food Laboratories for sample processing and the isolation of Salmonella.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was collaboratively funded by the Alberta Livestock and Meat Agency (ALMA) under grant number [2010R080] and Alberta Agriculture and Forestry.