*2.10. Statistical Analysis*

Bacterial recovery data were subjected to a one-way ANOVA using the general linear model procedure (SAS Institute Inc., Cary, NC, USA), with means deemed significantly different at *p* < 0.05 and separated using Duncan's multiple range test.

#### **3. Results and Discussion**

The objective for lab trial 1 was to spike layer feces with *Salmonella* Typhimurium and evaluate whether a high-pressure water rinse (HPWR) step prior to or following the PAA + FA treatment would be an added benefit in reducing aerobic bacteria and Salmonella (Table 1). Transportation coops treated with PAA + FA alone or with a HPWR step prior to or following the treatment in both replications were statistically similar (*p* < 0.05) in reducing aerobic bacteria (4.10 to 5.17 logs) and Salmonella (3.99 to 4.58). The LPWR consistently had the lowest reductions in both replications when reducing aerobic bacteria (2.09 and 2.14 logs) and Salmonella (2.10 and 2.16 logs).

The objective for lab trial 2 was to spike the feces with *Salmonella* Typhimurium and evaluate whether a HPWR step prior to or following a FC would be an added benefit in reducing aerobic bacteria and Salmonella (Table 2). Treatments using a FC varied statistically in both replications. In replication 1, HPWR prior to the FC and the FC used alone had the greatest reductions and were statistically similar (*p* < 0.05) in reducing aerobic bacteria (4.05 and 4.23 logs, respectively). The FC followed by the HPWR was statistically different (*p* < 0.05) from all other treatments at 3.5 log10 reductions of aerobic bacteria and was greater than the LPWR. The LPWR had the lowest reduction of aerobic bacteria at 1.12 logs.


**Table 1.** Lab Trial 1: Reduction of aerobic bacteria and Salmonella on broiler transportation coops following a compressed air foam application of disinfectant and a



pre-treatment samples. 4 Data are presented as mean ± SE, log10 reduction; *n* = 10 pooled samples per treatment; log reductions are subjected to a one-way ANOVA using the GLM procedure, with means deemed significantly different at *p* < 0.05 and separated using Duncan's multiple range test.5 Salmonella incidence data is described as "X" out of 10 samples or 10/10 for 100% positive samples. \*, a-b Column values with different superscripts differ significantly (*p* < 0.05).

In the same lab trial *Salmonella* Typhimurium recovery was also evaluated and all three treatments using the FC were statistically similar (*p* < 0.05) to one another (3.17 to 3.65 logs). The LPWR had the lowest reduction at 1.82 logs of Salmonella and was statistically different than all other treatments. These data demonstrate that the FC is effective in reducing not only aerobes but Salmonella as well.

In replication 2 of lab trial 2 (Table 2) aerobic bacteria reductions for all coops were statistically different from one another. The greatest reduction was achieved from the HPWR followed by the FC, which was a 4.84 log10 reduction of aerobic bacteria. Another significant reduction came from the FC used alone with a reduction at 3.59 logs of aerobes. The FC followed by the HPWR with a reduction of 2.78 logs of aerobic bacteria also had a significant reduction. The lowest reduction was observed with the LPWR treatment at 0.98 log reduction of aerobic bacteria.

Replication 2 also evaluated the reductions of Salmonella. The authors found that the HPWR followed by the FC had the greatest statistically significant reduction of 3.90 logs of aerobic bacteria. The HPWR used prior to the use of the FC consistently proved to be the most effective way to reduce aerobic bacteria and Salmonella in both replications, which could be because the organic matter was removed prior to the product being applied. The organic matter that was used for lab trials had water added and *Salmonella* Typhimurium was blended to allow the slurry to be thicker in consistency and truer to organic matter that is naturally present on broiler transportation coops. According to Dvorak and Peterson 2009 [18] the removal of organic matter first is essential because it acts as a barrier to the microorganisms present and affects the efficacy of the disinfectant. They concluded that the efficacy of hypochlorites is rapidly reduced when a large amount of organic matter is present. Perhaps this is why we saw better results from the coops treated by the HPWR first followed by the disinfectant or cleaner in this lab trial. The FC used alone and followed by the HPWR statistically had similar reductions (2.82 and 3.18 logs). Finally, the LPWR statistically showed that it had the lowest reductions of Salmonella at 0.65 logs of aerobic bacteria.

The objective of the field trial was to evaluate whether PAA + FA alone or after a HPWR step would be effective in reducing aerobic bacteria on freshly contaminated broiler transportation coops from a poultry processing facility (Table 3). Similar results were seen in both replications. Significant reductions (1.72 and 2.32 logs, respectively) of aerobic bacteria were observed from coops treated with HPWR followed by PAA + FA in both replications. The HPWR proved to be effective in a field setting, which may be due to the removal of organic matter present that had not been washed previously.


**Table 3.** Field Trial—Reduction of aerobic on broiler transportation coops following a compressed air foam application of disinfectant and a high-pressure water rinse.1.

1 All treatments were given a 10 min contact time and were followed by a LPWR of the transportation coops to remove any residual chemical. 2 LPWR = Low-pressure water rinse; PAA + FA = Peroxyacetic acid with a foam additive, and HPWR = High-pressure water rinse. 3 Values for reductions in aerobic bacteria were calculated by subtracting post-treatment from pre-treatment samples. 4 Data are presented as mean ± SE, log10 reduction; *n* = 10 pooled samples per treatment; log reductions are subjected to a one-way ANOVA using the GLM procedure, with means deemed significantly different at *p* < 0.05 and separated using Duncan's multiple range test. \*, a-b Column valueswithdifferentsuperscriptsdiffersignificantly(*p* <0.05).

Researchers have suggested that high-pressure rinsing may be more effective to significantly reduce bacterial load than a LPWR and was what was seen in the results for our lab trials [3]. Their hypothesis to apply a HPWR proved to be effective in a field setting along with removal of organic matter which is what previous research and literature suggests. Stringfellow and co-workers [19] concluded that when using disinfectants, correct contact time, temperature, and amount of organic matter affects product efficacy. Furthermore, a multi-step protocol is required to effectively reduce the bacterial load found in cages [20]. The higher amount of organic matter seen in the present study led to

the conclusion that the addition of a HPWR will further reduce bacterial load present on transportation coops. The PAA + FA used alone had a significant reduction of aerobic bacteria (0.88 and 0.80 logs). The LPWR had the lowest reduction concentrations (0.0 and 0.42 logs) for the field trials conducted.

Transport coop floors were power washed and treated with relatively high concentrations of disinfectant or cleaner during the lab and field trials. The researchers were surprised to continue to find bacteria present on surfaces that appeared to be smooth and clean. To further investigate this observation, a separate experiment was conducted within a microbiology lab. A coupon of broiler transport coop flooring was inoculated with *Enterococcus faecalis* and evaluated by a light and electron microscope (Figures 1–4). We found that the apparent smooth surfaces were actually scratched and covered in pits where bacteria could accumulate. It is possible that microbes may never encounter a cleaner or disinfectant due to the protection provided due to these imperfections.

**Figure 1.** Light micrograph of uninoculated fiberglass flooring material depicting a hole (white arrow) in the material as well as subsurface air bubbles (red arrow). Sub-surface bubbles can be exposed to surface contamination as the surface wears with age.

**Figure 2.** White arrow indicates hole in the surface of the fiberglass floor like that shown in Figure 1 (white arrow). Black arrow indicates bacteria colonizing the surface of the floor.

**Figure 3.** Higher magnification of a hole as seen in Figure 2 at 72 hours post inoculation with bacterial aggregates (white arrows).

**Figure 4.** Micrograph of the bottom of the hole seen in Figure 3. Large aggregates of bacteria are evident adhering to the area of the hole (white arrows).Figures 2–4: Scanning electron micrographs of bacteria inoculated flooring at various timepoints after inoculation and magnifications.

The current study did not demonstrate whether the bacteria present were killed on the coops or were physically washed away but shows whether the bacteria were reduced or removed. As such, this observation may be irrelevant since the bacteria are no longer present on the transportation coops that can be a vehicle for cross-contamination. Continued research in a commercial setting may be needed. Furthermore, evaluations of bacterial counts on carcasses can be compared when taken from washed transport coops versus unwashed coops to further determine bacterial load. These products have already been approved by the Environmental Protection Agency, which means that they may be implemented in being used at a poultry processing facility. These data sugges<sup>t</sup> that a CAFS application of cleaners and disinfectants may be used to significantly reduce Salmonella and aerobic bacteria on broiler transport coops. While a direct comparison was not made, coops from a commercial setting were found to be more difficult to clean and disinfect than coops which were contaminated in the laboratory.
