Effect of Feed Acidification and Conditioning Temperature on Feed Hygiene and Salmonella Recovery from Mash and Pelleted Broiler Feed

Poultry feed is routinely pelleted for many reasons such as increased feed intake, feed efficiency, and feed hygiene (Abdollahi et al., 2013). However, pelleting is not without its drawbacks, particularly when using higher pelleting temperatures (Abdollahi et al., 2010). The industry has been trending towards ever higher conditioning temperatures due, at least in part, to efforts to improve feed hygiene and thus decrease reliance on in-feed antibiotics for prevention and treatment of disease. To our knowledge, the present study is the first designed to understand the interaction between feed acidification and conditioning temperature on survival of Salmonella on feed specifically.

Four 100 kg lots of a wheat/SBM-based broiler diet (Table 1) were treated with Amasil NA at 0, 4, 7, and 10 g/kg. The next day, feed samples were inoculated with previously prepared stock solutions of Salmonella at 4 log₁₀ cfu/g of feed.

Preparation of inoculum: Previously prepared stock culture of Salmonella enterica subsp. enterica serovar Enteritidis (ATCC 13076) maintained at -80°C was transferred to fresh trypticase soy broth (TSB), incubated at 37°C, with subsequent transfers to fresh TSB for a total culture volume of 1L for each feed sample to be inoculated. The culture was then directly applied to the feed using a spray applicator with thorough mixing.

Sampling pre-pelleting: Following inoculation and mixing, 1 sample of each contaminated batch of feed (0, 4, 7, and 10 g/kg of Amasil NA) was collected. These samples were taken back to the lab, stored at 20°C, and used to enumerate Salmonella in the mash feed samples at d 0, 1, 2, 7, and 14 post-inoculation.

Sampling post-pelleting: The remainder of the inoculated feed in each lot was then placed in the hopper of the pellet mill (California Pellet Mill CL5 Laboratory Pellet Mill) and conditioning temperature increased to a target of 60, 75, or 90°C for 45 seconds. Once the target temperature was achieved, hot pellets were collected aseptically and immediately placed on ice. This first pelleted sample was later reinoculated with Salmonella (ATCC 13076) to simulate post-pelleting recontamination of feed within the mill. These samples were used to enumerate Salmonella on d 0, 1, 2, 7 and 14 post-re-inoculation. A second sample of hot pellets was collected and cooled for 10 minutes using a pilot-scale room-air pellet-cooler. Following cooling, this second pelleted sample was also placed on ice and used to determine the impact of conditioning temperature on Salmonella enumeration without being reinoculated. These later samples were used to enumerate Salmonella on d 0, 1, 2, 7 and 14 post-pelleting.

Salmonella enumeration: Enumeration was conducted according to methods previously described in Huss et al. (2017). In brief, a subsample of the previously collected feed samples was collected aseptically, serially diluted with buffered peptone water, and the appropriate dilutions spread plated in duplicate on xylose lysine deoxycholate agar (XLD; Becton, Dickinson and Company, Franklin Lakes, NJ). All XLD plates were incubated at 37°C for 24 hr. After incubation, all plates were enumerated by counting black colonies, typical for Salmonella and serotyped by the National Veterinary Services Laboratory (Des Moines, IA). The number of observed colonies was then multiplied by the dilution factor to determine the total count in cfu/g for the sample. For all plates with no growth, overnightenriched samples were serial diluted and spread plated to XLD and incubated. Growth on enriched plates were noted as positive or negative for Salmonella but not counted.

Statistical analysis: Calculated cfu/g data was converted to log₁₀ and analyzed using the GLIMMIX procedure of SAS with fixed effects of treatment (0, 4, 7, and 10 g/kg of Amasil NA), day (0, 1, 2, 7, and 14), and form (mash vs pellet). Contrasts for mash vs. pellet were also evaluated. Data from the two sets of pelleted samples (with or without reinoculation) were analyzed separately.

Inoculation of the acidified mash feed resulted in between 3.39 and 3.88 log₁₀ cfu/g on d 0 post-inoculation, with numerically lower recoveries in diets containing higher concentrations of Amasil NA (Table 2). The main effect of time decreased Salmonella recovery from mash feed (3.74, 2.54, 2.17, 1.48, 0.80 log10 cfu/g; P < 0.01). The main effect of 7 and 10 g/kg of Amasil NA decreased Salmonella relative to the control (2.92 vs 1.74 and 1.14 log10 cfu/g, respectively; P < 0.05). The 7 and 10 g/kg inclusion levels resulted in no detectable Salmonella growth with 14 and 7 days of inoculation, respectively. These results clearly show that acidification of mash broiler feed can greatly improve feed hygiene through direct elimination of Salmonella.

In the present study, pelleting at 60, 75, or 90°C resulted in feed from which no Salmonella could be enumerated at any time point (data not shown). Enrichment of pelleted samples revealed that, for feed containing 0, 4, and 7 g/kg of Amasil NA, the 60°C pellets were still Salmonella positive for the first 24-hr or so, despite not being culturable by direct plating on XLD. Starting on day 2, no Salmonella could be cultured from any of the pelleted samples. It has previously been shown that thermal or pH stress can push Salmonella in to a non-reproductive phase temporarily (Carrique-Mas et al., 2007). The additional enrichment step enables detection of these populations. However, pelleting at 75 or 90°C for 45 seconds, regardless of level of acidification, resulted in no detectable Salmonella even after enrichment. Interestingly, the combination of 10 g/kg of Amasil NA plus pelleting at 60°C also resulted in no detectable Salmonella after enrichment. This would indicate that acidification and pelleting temperature are potentially additive.

Re-inoculation of acidified broiler pellets with Salmonella resulted in approximately 6 to 7 log₁₀ cfu/g on day 0 (Figure 1). Salmonella recovery from the re-inoculated samples naturally decreased over time, even in the unacidified mash/pellets, but reduced more rapidly the higher the inclusion of Amasil NA. On day 0, the feed samples containing 4, 7 and 10 g/kg Amasil NA were an average of 0.14, 0.26 and 0.72 log₁₀ cfu/g lower (SE = 0.23), respectively, relative to the unacidified samples. This indicates that, at higher concentrations, feed acidification offers some level of immediate protection against recontamination of pellets and mash feed with Salmonella. As highlighted by Davies and Wray (1997), the pellet cooler can itself serve as a reservoir of Salmonella that can potentially undo the work of thermal treatment on feed hygiene. In feed originally inoculated with roughly 7 log₁₀ cfu/g of Salmonella, acidification with 10 g/kg of Amasil NA reduced the level of quantifiable Salmonella to 0 cfu/g in less than 7 days, although it could still be detected by overnight enrichment, and to undetectable even with enrichment within 2 weeks.

Both pelleting and acidification with Amasil NA were shown to be effective at increasing the hygiene of broiler feed. Pelleting at 75 and 90°C for 45 seconds were equally effective regarding the pre-pelleting inoculation, but offered only minimal protection against post pelleting recontamination. Acidification was effective regardless of feed form or conditioning temperature of pellets. Lower pelleting temperatures plus feed acidification with Amasil NA are shown to be complementary tools for promoting feed hygiene.