J APPL POULT RES 2006. 15:326-332
© 2006 Poultry Science Association
A Microbiological Assessment of On-Farm Food Safety Cleaning Methods in Broiler Barns1
P. J. Ward,
G. M. Fasenko2,
S. Gibson and
L. M. McMullen
Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Alberta, Canada T6G 2P5
2 Corresponding author: gaylene.fasenko{at}ualberta.ca
 |
SUMMARY
|
|---|
Little is known about the effectiveness of current poultry industry cleaning and disinfection procedures in broiler barns. Surface swabs were taken from a commercial broiler barn and a research broiler barn. The swabs were analyzed to determine if cleaning and disinfectant processes used would reduce the numbers of total aerobic bacteria and the total Enterobacteriaceae on the walls of the commercial barn (Experiment 1) and on the walls and floor of the research barn (Experiment 2). Microbiological analyses were used to assess bacterial loads within each barn. In Experiment 1 the mean of total aerobic bacterial counts were not affected by the cleaning and disinfectant process. However, the means of total Enterobacteriaceae counts were significantly reduced by the washing process. Disinfection with Virkon did not further reduce the numbers of Enterobacteriaceae. In general, the cleaning processes used in wood and metal barns were effective in reducing Enterobacteriaceaecounts. In Experiment 2, neither the mean of total aerobic bacterial counts nor Enterobacteriaceae counts were affected by the washing process. However, disinfection with Virkon did result in a significant reduction of bacterial numbers. In summary, the disinfectant step performed in the research barn was effective in decreasing both total aerobic and Enterobacteriaceae counts.
Key Words: broiler barn • aerobic bacteria • Enterobacteriaceae • disinfection • cleaning • microbiology
|
DESCRIPTION OF PROBLEM
|
|---|
Over the past 30 yr food safety concerns have increased, and poultry products continue to contribute significantly to the incidence of foodborne disease [1]. Consumer pressure has increased the scrutinizing of production practices, especially in the poultry and beef industries. In response to consumer demands, Canadian poultry producers have initiated on-farm food safety programs with the goal of standardizing production practices used at primary production farms.
These programs are based on a food safety program (Hazard Analysis Critical Control Points; HACCP) developed by Pillsbury Company in conjunction with the National Aeronautics Space Agency and Natick Labs. Originally, HACCP was implemented to ensure that astronauts would not become ill during space missions because of consumption of food contaminated with pathogens [2].
Canadian poultry producers are using HACCP as a guide to develop standardized production practices at the farm level. The goal of the on-farm food safety program is to ease consumer concerns about food safety by producing safer foods of high quality that can compete within the global market. This may lead to an increase in average value for poultry products produced in Canada.
Research on the appropriate production practices to maximize on farm food safety is scarce. In the Canadian nonintegrated production system, individual producers have the freedom to follow their own production practices, and, therefore, there is large variation among farms with respect to the cleaning and disinfection methods used in broiler barns. Because of individual producer cleaning procedures, in some cases not all of the organic material is completely removed from the barn before the disinfectant is applied. If organic material is present during the disinfecting process then the efficacy of the disinfectant used can be significantly reduced. Previous research has demonstrated that the efficacy of a peroxygen compound (Virkon) is reduced when organic matter is present [3].
The first objective of this study was to determine if the cleaning and disinfecting processes used on one commercial farm are effective in reducing the microbial load. The hypothesis for this study was that lower total aerobic and Enterobacteriaceae counts would be present in the barn after each step of the cleaning process.
The second objective of this study was to determine the efficacy of a commercially available disinfectant in the presence of organic material in broiler research pens. The hypothesis for this study was that the microbial counts would be less in pens that had disinfectant applied after being washed down with cold water than in those pens that had the disinfectant applied after only sweeping out the litter.
|
MATERIALS AND METHODS
|
|---|
Experiment 1
The microbiological sampling sites for the first experiment were located within several commercial broiler barns [4]. Samples were collected in 2 barns with wood walls [5] and 1 barn with metal walls [6]. The first samples were taken after the broilers had been removed from the barns but before the litter was removed. After the first samples were taken, manure within the barns was then pushed out using front loader equipment; this was completed for each flock. The condition of the barns at the time of the second sampling varied between the metal and wood barns. The second sample collection in the metal barn was done after the barn walls (from the ceiling down) and the floors had been washed with hot water (approximately 45°C). Extremely cold environmental temperatures outside the barns at the time of the experiment (about –40°C), and the lack of floor drains in the wood barns prevented washing these barns with hot water prior to application of disinfectant. Therefore, the second sampling in the wood barns was completed after the barn walls (from the ceiling down) and the floors were blown out using compressed air. The third set of samples in all 3 barns was collected after application of the disinfectant at the manufacturers recommended concentration of 1% [7] to the walls and the floors of the 3 barns. The sampling area at each site was 500 cm2 [8]. Sixteen samples were taken, the samples prepared for microbial analysis [9], and total aerobic bacteria and Enterobacteriaceae counts were determined. Counts were then converted to log colony-forming units per centimeter squared, and data were then analyzed using appropriate statistical procedures [10]. The presence of Enterobacteriaceae was determined because this family of bacteria includes many human enteropathogens, including Salmonella and Escherichia coli, and total aerobic bacteria were examined to provide an indication of the overall cleanliness of the barns.
Experiment 2
Because the data from Experiment 1 were collected in a commercial setting, we decided to conduct additional research in a setting in which the variables could be more easily controlled. All sampling for the second experiment was conducted in broiler pens at the Alberta Poultry Research Center at the University of Alberta [8, 9]. All 8 pens sampled were identical [12], and the samples were taken at 4 identical locations in all pens [13]. Half the pens (n = 4) were assigned to 1 of 2 cleaning methods. Four pens on the north side of the barn were allocated to be washed with cold water, and 4 pens on the south side of the barn were allocated to just have the litter shoveled out of the pens [14]. The first samples were taken from all 8 pens after the birds had been removed and the litter had been shoveled out (sample 1) [15]. Four of the 8 pens were then washed with cold water [16]. After approximately 24 h of air drying, samples were taken in all 8 pens (sample 2). The disinfectant was then applied to all 8 pens [17], and the pens were allowed to air-dry for approximately 24 h before the last samples were taken (sample 3). Overall, each of the 8 pens had 4 sample locations, and the samples were taken over 3 sample times, resulting in a total of 96 samples. Total aerobic and Enterobacteriaceae bacterial counts were performed on all samples [9], and all data were analyzed using proper statistical procedures [10].
|
RESULTS AND DISCUSSION
|
|---|
Experiment 1: Effect of Sampling Location
Wood Barns.
Mean log total counts of aerobic bacteria varied significantly (P = 0.0078) due to sampling location (Figure 1
). Total aerobic counts obtained from samples collected at locations 2 (corner) and 6 (wall near exhaust fan) were significantly higher compared with those collected at location 1 (wood loading door). Mean log of total aerobic bacteria counts from all other locations were not significantly different to those obtained at location 1, 2, or 6. In contrast, mean log total counts of Enterobacteriaceae were not influenced by sample location (P = 0.4912).
View larger version (21K):
[in this window]
[in a new window]
|
Figure 1. Effect of sampling location in 2 wood-walled barns on mean log total aerobic bacteria and Enterobacteriaceae counts. Location 1 = main wood loading door, location 2 = corner, location 3 = one-third of the length of the barn, location 4 = two-thirds of the length of the barn, location 5 = wood service door, location 6 = one-third of the length of opposing wall near a exhaust fan site, location 7 = two-thirds of the length of the opposing wall near a exhaust fan site, and location 8 = corner. a,bBars for each location that share a common letter are not significantly different (P  0.05); n = 12. Each mean is the average of 2 samples taken at each location on 3 separate days in 2 barns.
|
|
Metal Barn.
Mean log total counts of aerobic bacteria from the metal wall barn were significantly different (P = 0.0303) due to sampling location (Figure 2). Location 6 (corner) had significantly higher total aerobic bacteria counts compared with location 2 (metal service door). Mean log of total aerobic bacteria counts from all other locations were not significantly different to those obtained at locations 2 and 6. In contrast, mean log counts of Enterobacteriaceae were not significantly different based on sampling location (P = 0.7883).
View larger version (22K):
[in this window]
[in a new window]
|
Figure 2. Effect of sampling location in a metal-walled barn on the mean log total aerobic and Enterobacteriaceae counts. Location 1 = main overhead loading door (wood); location 2 = service door (metal); location 3 = one-third of the length of the barn; location 4 = two-thirds of the length of the barn; location 5 = rear wall, near exhaust fan; location 6 = corner; location 7 = one-third of the length of opposing wall; location 8 = two-thirds of the length of the opposing wall. a,bBars for each location that share a common letter are not significantly different (P 0.05); n = 6. Each mean is the average of 2 samples taken at each location on 3 separate days in on barn.
|
|
Sampling location within the barn types significantly influenced total aerobic bacterial counts. Corner locations in the wood barns had higher total aerobic bacterial counts than those found on the wood overhead loading doors. Lower bacterial counts by the loading doors could be due to cold air draft coming into the barn, thereby keeping chickens away from this location. A corner location in the metal barn had higher counts than the metal service door. In the barns with wooden walls, a location near the exhaust fans also had higher counts than the loading door. The corner locations and the area near the exhaust fans represent areas that are more difficult to clean and, therefore, have a greater potential for higher contamination. Location in the wood or metal barns had no effect on the Enterobacteriaceae counts. The numbers of Enterobacteriaceae present were often close to undetectable levels, which could have precluded our ability to detect any minor differences in the counts.
Effect of Cleaning Steps
Wood Barns.
log total counts of aerobic bacteria were not significantly affected by the various cleaning steps (P = 0.7892; Figure 3). Contrast, mean log counts of Enterobacteriaceae were significantly lower (P = 0.0029) after cleaning and disinfecting the barn than those found after just removing litter from the barns.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 3. Effects of cleaning and disinfecting procedures used in two wood walled barns on the mean log total aerobic bacteria and Enterobacteriaceae counts. Preclean samples were collected after flock shipment but before litter removal; dry clean samples were collected after litter removal and blowing out the interior of the barn with pressurized air; disinfect samples were collected after the barn was disinfected with Virkon [7]. a,bBars for each cleaning step that share a common letter are not significantly different (P 0.05); n = 32. Each mean is the average of 2 samples taken at all 8 locations in the 2 barns.
|
|
Metal Barns.
Mean log total aerobic bacteria counts were not significantly affected as a result of cleaning procedures (P = 0.0741; Figure 4). In contrast, mean log counts of Enterobacteriaceae were significantly affected (P < 0.0001) by the different cleaning steps. Washing with hot water significantly reduced mean log Enterobacteriaceae counts (P < 0.05), but subsequent application of disinfectant did not further reduce these counts.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 4. Effect of cleaning and disinfecting procedures in a metal walled barn on the mean log total aerobic bacteria and Enterobacteriaceae chounts. Preclean samples were collected after flock shipment but prior to litter removal; wet clean samples were collected after litter removal and washing the interior of the barn with pressurized hot water; disinfect samples were collected after the barn was disinfected with Virkon [7]. a,bBars for each cleaning step that share a common letter are not significantly different (P 0.05); n = 16. Each mean is the average of 2 samples taken at all 8 locations in the barn.
|
|
Counts of Enterobacteriaceae from samples collected from the wooden barns demonstrated that the dry cleaning process significantly reduced these counts, but subsequent use of disinfectant did not result in a further reduction of Enterobacteriaceae. The same trend was noted in the results from the metal barn. The presence of organic material in the barns at the time of disinfection in the wood barns may have reduced the antibacterial effectiveness of the Virkon. Previous research found that the presence of visible organic material on the bottom of boots interfered with the antibacterial efficacy of a 1% Virkon solution [3]. More research is needed to elucidate the impact of organic material on the efficacy of Virkon before definitive conclusions can be made about the disinfecting procedures. It was also determined after sampling was complete that there was a dilution error with the pressure washer that was used. Although a base solution of 1% Virkon was mixed, the pressure washer further mixed 1% of this solution into the waterline that was applied to the walls rather than simply applying the 1% solution directly. This meant that the concentration of disinfectant actually applied was about 0.1%. This could have been the major cause of why there was no further significant reduction in bacterial counts within the barns with the application of the disinfectant.
Experiment 2: Effect of Cleaning and Disinfecting Steps
Enterobacteriaceae.
Enterobacteriaceae counts were similar across all pens after litter had been shoveled out of all experimental pens (P = 0.5096; Figure 5). This result was expected because at this point in the experiment the washed and unwashed pens were treated the same (litter removed only). Pens that were then pressure washed with cold water did not show further reduction in bacterial counts compared with the pens that had only the litter removed (P = 0.6531). Application of the disinfectant to washed pens provided a significant reduction in the population of Enterobacteriaceae bacteria compared with those pens that only had the litter removed (P = 0.0294).
View larger version (15K):
[in this window]
[in a new window]
|
Figure 5. Effect of cleaning and disinfecting steps on the mean log total Enterobacteriaceae bacterial counts. Litter removed = all samples were collected after the birds had been shipped and all litter had been shoveled out of the pens. +/– Pressure wash = half of the experimental pens were washed with cold water the other half only had litter removed (no washing). Disinfect = application of Virkon [7] to washed and unwashed pens. a,bBars that do not share common letters within each cleaning step are statistically significantly at P 0.05; n = 8. Each mean is the average of four samples taken in 2 pens.
|
|
Total Aerobic Bacteria.
As with the Enterobacteriaceae counts, total aerobic bacterial counts recovered from samples after litter had been removed did not vary among (P = 0.8553; Figure 6). In addition pens that were washed with cold water did not show further reduction in the bacterial population compared with those pens that only had the litter removed (P = 0.0584). However, with the application of the disinfectant there was a significant reduction in total aerobic bacterial counts in those pens that had received the cold-water wash as compared with those that only had litter removed (P < 0.0001).
View larger version (17K):
[in this window]
[in a new window]
|
Figure 6. Effect of cleaning and disinfecting steps on the mean log total aerobic bacteria counts. Litter Removed = all samples were collected after the birds had been shipped and all litter had been shoveled out of the pens. +/– Pressure wash = half of the experimental pens were washed with cold water the other half only had litter removed (no washing). Disinfect = application of Virkon [7] to washed and unwashed pens. a,bBars that do not share common letters within each cleaning step are statistically significantly at P 0.05; n = 8. Each mean is the average of four samples taken in 2 pens.
|
|
The results from this experiment showed that the use of cold-water wash did not significantly reduce the overall bacterial load within the experimental pens. This was observed for total aerobic bacteria and Enterobacteriaceae counts. The data did provide evidence that in the presence of organic matter (unwashed pens) the efficacy of Virkon was reduced. This result was observed for the Enterobacteriaceae and total aerobic bacterial counts after the disinfectant had been applied to the pens. The research in the present study supports previous research showing that the efficacy of Virkon is reduced when applied in the presence of organic matter [3]. The on-farm food safety program currently being implemented by Canadian broiler producers requires that litter be removed after every cycle. The data presented in the 2 experiments supports this policy and further emphasizes the need to be diligent in the removal of as much organic material as possible prior to disinfectant application. This policy has become even more important since the avian influenza outbreak that occurred in British Columbia, Canada in 2004. Removal of organic material could be performed by blowing out the barn with compressed air or washing the barn with water (cold or hot). By removal of as much organic matter as possible the efficacy of the disinfectant will be maximized when the manufacturer’s recommended concentration is applied.
|
CONCLUSIONS AND APPLICATIONS
|
|---|
- Sampling location in wood- and metal-walled barns affected the number of total aerobic bacteria recovered. In both barn types, there were more total aerobic bacteria recovered from corners than from a wood loading door (wood barn) or a metal service door (metal barn). Sampling location did not affect the counts of Enterobacteriaceae isolated.
- The 2 main cleaning steps (litter removal and litter removal plus washing) examined on one commercial broiler farm did result in decreases in mean log total counts of Enterobacteriaceae. Mean log total counts of aerobic bacteria were not affected by the cleaning steps conducted.
- Data obtained from a controlled research setting support previous research showing that the mean log total counts of Enterobacteriaceae and aerobic bacteria had the greatest reduction in pens that were pressure washed with cold water prior to the application of the disinfectant.
- Further research is needed to examine cleaning methods and different disinfectants on other farms in an effort to provide scientific based information to develop effective standardized cleaning and disinfecting methods in commercial broiler barns.
|
ACKNOWLEDGMENTS
|
|---|
This project was supported by the University of Alberta Department of Agricultural, Food, and Nutritional Science, the University of Alberta Poultry Research Center, the Alberta Chicken Producers, the Alberta Livestock Industry Development Fund, and the Saskatchewan Chicken Industry Development Fund. The authors acknowledge M. Therrien, E. O’Dea, R. McEgan, and Jacob Middlekamp for their assistance.
|
FOOTNOTES
|
|---|
1 The use of trade names in this publication does not imply endorsement of the products mentioned or criticism of similar products not mentioned. 
|
REFERENCES AND NOTES
|
|---|
- FAO. 2002. Risk assessments of Salmonella in eggs and broiler chickens: Interpretative summary. WHO, Rome. http://www.who.int/foodsafety/publications/micro/en/salm_summary.pdf Accessed Jan. 2004.
- Mortimore, S., and C. Wallace. 1994. HACCP: A Practical Approach. Chapman and Hall, New York, NY.
- Amass, S. F., D. Ragland, and P. Spicer. 2001. Evaluation of the efficacy of a peroxygen compound, Virkon(r), as boot bath disinfectant. J. Swine Health Prod. 9:121–123.
- The 2 wood barns measured 60 m2 and had capacities of 9,810 broilers. The metal barn measured 110 m2 and housed 21,420 broilers. All barns had cement floors. The wood barns had negative ventilation, and the metal barn had tunnel ventilation. Eight sample locations were predetermined based on the lengths of the barns and locations of ventilation equipment, main service door, and maintenance doors.
- Sample locations in wood barns: location 1 = main overhead wood loading door, location 2 = corner, location 3 = one-third of the length of the barn, location 4 = two-thirds of the length of the barn, location 5 = wood service door, location 6 = one-third of the length of opposing wall, near an exhaust fan; location 7 = two-thirds of the length of the opposing wall, near an exhaust fan; and location 8 = corner. At each sample site, 2 locations were chosen, one at 25 cm above the barn floor and the other directly above the first sample site at 125 cm above the barn floor. In the wood and metal barns all samples were taken on the walls and doors, and no samples were taken from the floors of either barn.
- Sample locations in metal barn: location 1 = main overhead wood loading door, location 2 = metal service door, location 3 = one-third of the length of the barn, location 4 = two-thirds of the length of the barn, location 5 = rear wall, near fan, location 6 = corner, location 7 = one-third of the length of opposing wall, and location 8 = two-thirds of the length of the opposing wall. At each sample site, 2 locations were chosen, one at 25 cm above the barn floor and the other directly above the first sample site at 125 cm above the barn floor.
- Virkon, Antec International Ltd., Suffolk, UK.
- Each sample site was 10 x 50 cm. Sterile gauze pads (Johnson and Johnson sterile gauze pads, Fisher Scientific, Whitby, Ontario, Canada) moistened with sterile 0.1% peptone water (0.1% sterile peptone, Difco, Becton-Dickson and Co., Sparks, MD) were held in the sampler’s hand, which was covered with a sterile glove. The moistened gauze pad was then pressed firmly against the surface of the site and swiped 10 times horizontally and 10 times vertically. After wiping, the gauze pads were put into 24-oz Nasco Whirl-Pak bags (Fisher Scientific, Edmonton, Alberta, Canada) and placed in a cooler at 4°C for transportation to the laboratory.
- After the samples were collected, they were kept in a cooler at 4°C for 24 h prior to microbial analysis. Tenfold serial dilutions were prepared for each sample using sterile 0.1% peptone water. Total aerobic bacteria were enumerated on prepoured plates of plate count agar (PCA) (Difco, Becton-Dickson Co., Sparks, MD) after aerobic incubation at 23°C for 48 h. Enterobacteriaceae were enumerated by pour-plate and spread-plate techniques on violet red bile agar (Difco) with 1% added glucose [Fisher Scientific, Edmonton, Alberta, Canada (VRBGA)] after incubation at 37°C for 24 h.
- The experimental design for both studies was a randomized complete block design consisting of pens or barns, sampling day or sample sites, or both as independent variables. Microbiological data were converted to log colony-forming units per centimeter squared and were subjected to analysis of variance according to the PROC GLM procedure of SAS [11]. Least squares means were used to determine level of significance at P 0.05 and where appropriate, mean separation was performed using the Student-Newman Keuls’ multiple range test.
- SAS Institute. 2001. Version 8 for Windows. SAS Institute, Inc., Cary, NC.
- Pens measured 5.49 x 2.29 m. Each pen was made up of a cement floor, one solid metal wall extending up 0.35 m from the floor (with wire walls extending up from the 0.35 m point on the wall), 3 painted cinder block walls (with wire walls extending up from the 0.35 m point on 2 of the walls), and a door made of metal and wire (floor to 0.35 m was solid metal; 0.35m up was wire). The barn had a negative ventilation system.
- Location 1 = cement floor in middle of pen under the nipple drinker, location 2 = cement floor 0.5 m diagonally out from left corner, location 3 = cinder block wall halfway down the length of the pen, and location 4 = metal wall halfway down the length of the pen. All wall samples were taken 13 cm above the floor. The first wall samples were taken 225 cm from the corner, the second samples were taken 275 cm from the corner, and third samples were taken 325 cm from the corner. Location 1 cement floor samples were taken 115 cm away from the side walls of the pen; first floor sample was taken 225 cm down from the end wall of the pen, second sample was taken 275 cm down from the end wall of the pen, and the third sample was taken 325 cm down from the end wall of the pen. Location 2 cement floor sample: the first sample was taken 50 cm (diagonally) from left corner of the pen, the second sample was taken 60 cm (diagonally) from the left corner of the pen, and the third sample was taken 70 cm (diagonal) from the left corner of the pen. The sample area for each site was 500 cm2 [8].
- Pens on the north and south sides of the barn were separated by 3.9 m wide hallway.
- Initially 200 broiler chicks were placed in all pens. At shipping the pens contained the following numbers of birds: pen 1 = 182 birds, pen 2 = 193 birds, pen 3 = 187 birds, pen 4 = 190 birds, pen 5 = 187 birds, pen 6 = 186 birds, pen 7 = 188 birds, and pen 8 = 186 birds. Birds were shipped at 39 d of age.
- Pressure washing was done with a 1,000 psi Karcher pressure washer (Alfred Karcher GmbH and Co. K.G., Winnenden, Germany), and approximately 5 L of cold tap water was applied to each of the 4 pens designated to this cleaning treatment.
- Virkon, Antec International Ltd., Suffolk, UK. Virkon was diluted to 1% solution in cold tap water, and 7 L of this solution was then applied to all experimental pens.
This article has been cited by other articles:
|
|
|
|
 
B. M. Rathgeber, K. L. Thompson, C. M. Ronalds, and K. L. Budgell
Microbiological evaluation of poultry house wall materials and industrial cleaning agents
J. Appl. Poult. Res.,
January 1, 2009;
18(3):
579 - 582.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
SUMMARY
DESCRIPTION OF PROBLEM
