Effectiveness of a Direct-Fed Microbial Product Containing Lactobacillus acidophilus and Lactobacillus casei in Reducing Fecal Shedding of Escherichia coli O157:H7 in Commercial Feedlot Cattle

Abstract

The objective of this study was to evaluate the effectiveness of a direct-fed microbial (DFM) product in reducing fecal shedding of Escherichia coli O157:H7 in finishing commercial feedlot cattle in Kansas (KS) and Nebraska (NE). Utilizing a randomized complete block design within the feedlot (KS, n = 1; NE, n = 1), cattle were randomly allocated to 20 pens, grouped in blocks of two based on allocation date, and then, within the block, randomly assigned to a treatment group (DFM or negative control). The DFM product was included in the diet at a targeted daily dose of 1 × 10⁹ colony-forming units (CFU) of the Lactobacillus acidophilus and Lactobacillus casei combination per animal for at least 60 d before sampling. Feedlots were sampled for four consecutive weeks; weekly sampling consisted of collecting 20 pen floor fecal samples per pen. Fecal samples were subjected to culture-based methods for detection and isolation of E. coli O157, and positive samples were quantified using real-time polymerase chain reaction. Primary outcomes of interest were fecal prevalence of E. coli O157:H7 and E. coli O157 supershedding (≥10⁴ CFU/g of feces) prevalence. Data for each feedlot were analyzed at the pen level using mixed models accounting for the study design features. Model-adjusted mean E. coli O157:H7 fecal prevalence estimates (standard error of the mean [SEM]) for DFM and control groups were 8.2% (SEM = 2.2%) and 9.9% (SEM = 2.5%) in KS and 14.6% (SEM = 2.8%) versus 14.3% (SEM = 2.6%) in NE; prevalence did not differ significantly between treatment groups at either site (KS, p = 0.51; NE, p = 0.92). Mean E. coli O157 supershedding prevalence estimates for DFM and control groups were 2.2% (SEM = 0.7%) versus 1.8% (SEM = 0.7%) in KS (p = 0.66) and 6.7% (SEM = 1.5%) versus 3.2% (SEM = 1.0%) in NE (p = 0.04). In conclusion, administering the DFM product in the finishing diet of feedlot cattle did not significantly reduce E. coli O157:H7 fecal prevalence or supershedding prevalence in study pens at either commercial feedlot.

Introduction

Seven Shiga toxin-producing Escherichia coli serogroups, including serotype E. coli O157:H7, are considered adulterants in raw, nonintact beef products by the Food Safety and Inspection Service (USDA-FSIS, 2012). Based on U.S. data from 2000 to 2006, it is estimated that ∼33% of human foodborne illnesses due to E. coli O157:H7 are attributed to ground beef (Withee et al., 2009). Cattle shed these bacteria in their feces, and fecal material may contaminate the hides of cattle in the production environment, during transport, and/or in lairage, which then can serve as the primary source of carcass and subsequent beef product contamination (Loneragan and Brashears, 2005; Fox et al., 2008; Jacob et al., 2010b). Therefore, reducing the fecal shedding prevalence and concentration of E. coli O157:H7 in cattle conceivably reduces the risk of contamination of beef products.

A subset of cattle, termed supershedders, shed E. coli O157:H7 at high concentrations (≥10⁴ colony-forming units [CFU]/g of feces) and these animals have been shown to be associated with most of the within-pen transmission of E. coli O157:H7 in the cattle production environment (Omisakin et al., 2003). However, supershedding has been described as transient or intermittent, not continuous, in individuals over time (Munns et al., 2014; Williams et al., 2014); therefore, deeming individual animals as supershedders may be a mischaracterization. While the role of these supershedding events in the feedlot environment is not completely understood, it is clear that they pose a risk to beef safety. Targeting E. coli O157:H7 in the bovine reservoir before harvest offers an opportunity to decrease the bacterial load in the host and in the environment while reducing the potential for contamination of hides and subsequent food products.

Preharvest interventions to reduce E. coli O157:H7 fecal shedding in cattle have been at the forefront of beef safety research for over two decades (LeJeune and Wetzel, 2007; Marder et al., 2018). Evaluated preharvest interventions include diet interventions and management, direct-fed microbials (DFMs), antimicrobials, and vaccines (Loneragan and Brashears, 2005; Callaway et al., 2013). Despite conflicting research findings on various DFM products, a meta-analysis demonstrated that DFMs, including many strains of bacteria, yeast, molds, and combinations, may be effective as a preharvest intervention in reducing fecal prevalence of E. coli O157:H7 in beef cattle (Wisener et al., 2015). Currently, there are limited data evaluating the impact of DFM products on the prevalence and concentration of E. coli O157:H7 in the production environment (Stephens et al., 2007a; Arthur et al., 2010; Cernicchiaro et al., 2010; Cull et al., 2012; Brown et al., 2020). Thus, the objective of this study was to evaluate the effectiveness of a commercially available DFM product, containing a proprietary blend of Lactobacillus acidophilus and Lactobacillus casei, in reducing fecal shedding of E. coli O157:H7 and supershedding events in finishing pens of commercial feedlot cattle in Kansas (KS) and Nebraska (NE) during the summer months.

Materials and Methods

Study population, design, and sample collection

Commercial feedlots were selected based on their willingness to conduct research, ability to feed a DFM product and a control diet concurrently, and capacity to fill 20 pens with cattle on a finishing diet during the summer. The study population comprised crossbred beef cattle in 40 study pens, 20 pens per feedlot (10 per treatment group), with projected harvest dates between August and September 2018 at two commercial feedlots. One feedlot was located in KS (feedlot capacity = 30,000 cattle) and the other was located in NE (feedlot capacity = 16,000 cattle). Study pens were enrolled between April and May 2018. Cattle were procured, processed, and managed according to the standard operating procedures of each feedlot. Standard operating procedures of the commercial operations were followed to prevent mixing of the two study diets. However, cattle were managed under typical commercial conditions and thus DFM and control pens could share fence lines and waterers. Study pens in NE were confined to a single alley and treatment groups commonly shared fence lines and waterers. In contrast, at the KS site, blocks of cattle were housed in close proximity to each other, but study blocks were distributed throughout the feedlot, and the majority of study pens did not share fence lines with other study pens.

This field trial consisted of a randomized complete block design with the pen as the experimental unit and repeated sampling. In each feedlot, cattle were randomly allocated, within arrival dates, to pens grouped in blocks of two (DFM or control); within a block, pens were randomly allocated to DFM (n = 10) or control groups (n = 10). The total number of pens (20 per treatment group) were estimated based on design parameters and previous data described by Cull et al. (2012), assuming a mean E. coli O157:H7 prevalence of 40% and 25% for control and DFM groups, respectively (α = 0.05, β = 0.20). Cattle in DFM pens were administered the DFM product in their feed at a targeted daily dose of 1 × 10⁹ CFU of the L. acidophilus and L. casei combination (50 mg per animal per day of BactaShield™; Legacy Animal Nutrition, LLC; Wamego, KS), whereas cattle in control pens received no DFM product. Feed testing was not conducted to determine as-fed DFM concentrations. Study pens were fed the allocated diet for at least 60 d before the first sampling.

Each of the enrolled pens was sampled weekly for four consecutive weeks with 20 fecal samples collected weekly from each pen. The KS feedlot was sampled during the first four consecutive weeks and the NE feedlot was sampled the following four consecutive weeks. A sample comprised ∼30 g of freshly eliminated feces obtained off the pen floor and collected in individual plastic bags (WHIRL-PAK^®; Nasco, Fort Atkinson, WI) using plastic spoons. All fecal samples were placed in a cooler with ice packs and transported to the Preharvest Food Safety Laboratory at Kansas State University for processing within 24h. One sampling crew collected all fecal samples for this study. At the time of sampling, data on pen conditions and weather also were documented using a standardized data capture form. Observed pen conditions were classified as dry/dusty, normal, wet, and very wet. Weather data from the National Weather Service mobile application included temperature (°F), precipitation (yes or no), and humidity at the time of sampling for each pen.

Detection and quantification of E. coli O157:H7

Culture methods, including the immunomagnetic separation (IMS) technique utilized, have been described in detail previously (Dewsbury et al., 2015). Following enrichment, IMS, plating on Sorbitol MacConkey agar supplemented with cefixime and potassium tellurite (CT-SMAC), and overnight incubation at 37°C, putative colonies were tested for the O157 antigen by latex agglutination. If all subcultured colonies, maximum of six, tested negative for the O157 antigen by latex agglutination, the sample was considered negative for E. coli O157:H7 and no further testing was done. If a subcultured colony tested positive for the O157 antigen by latex agglutination, it was tested by conventional multiplex polymerase chain reaction (PCR) assay targeting the rfbE, fliC _H7, eae, stx1, stx2, and ehxA genes (Bai et al., 2010). A sample was considered positive for E. coli O157:H7 if the isolate tested positive for rfbE, fliC _H7, eae, stx1, and/or stx2. Pre-enriched samples were subjected to quantitative PCR (qPCR) (Noll et al., 2015) to identify supershedding events only if the enriched sample was positive for E. coli O157:H7. A sample was considered a supershedding event if the pre-enriched sample yielded a qPCR (Noll et al., 2015) average end-point threshold cycle for the rfbE gene that was less than or equal to 37.8 (E. coli O157 concentration ≥10⁴ CFU/g of feces).

Statistical analyses

Unadjusted E. coli O157:H7 sample-level prevalence was calculated as the number of fecal samples that tested positive for E. coli O157:H7 divided by the total number of samples subjected to culture methods. Similarly, unadjusted E. coli O157 supershedding prevalence was calculated as the number of fecal samples with estimated concentrations ≥10⁴ CFU E. coli O157/g of feces divided by the number of samples subjected to culture methods.

All analyses were performed at the pen level (experimental unit) with data analyzed for each study site separately. Outcomes of interest consisted of pen-level fecal prevalence and pen-level supershedding prevalence, which were modeled as the number of positive samples in each pen divided by the total number of samples collected in each pen at each sampling visit (events/trials). Models were fitted using generalized linear mixed models in Proc Glimmix (SAS 9.3; SAS Institute, Inc., Cary, NC) with a binomial distribution, restricted pseudolikelihood estimation, logit link, Kenward–Roger degrees of freedom, and Newton–Raphson and Ridging optimization procedures. In the model, the treatment group (DFM or control), sampling visit (1, 2, 3, and 4), and a treatment-by-sampling visit interaction were included as fixed effects. A random effect for the block and a first-order autoregressive covariance structure at the pen level to account for repeated measures also were included in the model. Tukey–Kramer methods were used to adjust for multiple comparisons. Treatment effects were considered significant when p-values were ≤0.05. Additionally, for all blocks, a simple t-test was performed to evaluate if mean days on treatment differed between study sites; similar analyses were performed at the pen level to evaluate if mean number of animals per pen and mean initial body weights differed between KS and NE. Descriptive statistics were tabulated for observed pen conditions and weather data.

Results

Study population and sample collection

Study population and sampling data are provided for each study site in Table 1. At enrollment, the average body weights ± standard deviations (SDs) of KS (379.2 ± 46.4 kg; range = 257.2 to 437.3 kg) and NE (365.5 ± 27.2 kg; range = 326.6 to 403.2 kg) study cattle were not significantly different (p = 0.27). The average pen size at the KS study site was significantly greater than the NE study site (p < 0.01); the average number of animals per study pen in KS and NE was 129.8 (SD = 35.3 animals/pen; range = 59 to 192 animals/pen) and 78.7 (SD = 6.6 animals/pen; range = 70 to 90 animals/pen), respectively. Study sites differed in rations fed; notably the amounts of wet distillers' grains (WDG) included in KS and NE were 22% and 44%, respectively.

Table 1.

Feedlot, Study Population, and Sampling Characteristics by Study Site

Characteristic, unit	Kansas	Nebraska
Feedlot capacity, no. of animals	30,000	16,000
Average enrollment pen size (range), no. of animals	130 (59–192)	79 (70–90)
Sex	Heifer	Steer
Average enrollment weight (range), kg	379 (257–437)	366 (327–403)
Average days on treatment diet at first sampling (range)	90 (68–102)	60 (60–60)
Sampling visit, date
1	7/23/2018	8/20/2018
2	7/30/2018	8/27/2018
3	8/5/2018	9/3/2018
4	8/12/2018	9/10/2018
Sampling visit,
No. of pens sampled (no. of samples collected)
1	20 (400)	20 (400)
2	20 (400)	20 (400)
3	20 (400)	20 (400)
4	14 (280)	20 (400)
Total pens sampled	20	20
Total samples collected	1480	1600

At the time of the first sampling at the KS and NE feedlots, study pens were on the allocated treatment diet for an average of 90 d (SD = 13.8 d; range = 68 to 102 d) and 60.0 d (SD = 0.0 d; range = 60 to 60 d), respectively; mean days on treatment significantly differed between study sites (p < 0.01). At the KS feedlot, 14 of 20 pens were sampled for the entire study period; however, 6 pens were only sampled for three consecutive weeks as they were sent to harvest before the fourth sampling visit. All 20 pens were sampled the entire 4-week period at the NE feedlot. Overall, 3080 fecal samples were collected (KS, n = 1480; NE, n = 1600). In KS, pen conditions were normal on sampling of most pens (64.9%; 48/74) or very wet (33.8%; 25/74), whereas in NE, the majority of the pens were wet (43.8%; 35/80) or very wet (32.5%; 26/80) at the time of sample collection throughout the 4-week sampling period. The observed pen conditions and weather data are summarized for each sampling week by study site in Table 2.

Table 2.

Pen Conditions and Weather Data by Sampling Visit for Each Study Site

State	Sampling visit	Pens sampled^a	Pen conditions, no. of pens (% pens sampled)				Average temperature (range), °F	Average humidity (range), %	Raining
State	Sampling visit	Pens sampled^a	Dry/dusty	Normal	Wet	Very wet	Average temperature (range), °F	Average humidity (range), %	Raining
Kansas
	1	20	1 (5.0)	17 (85.0)	0 (0.0)	2 (10.0)	74 (69–80)	81 (64–93)	No
	2	20	0 (0.0)	0 (0.0)	0 (0.0)	20 (100.0)	66 (61–72)	84 (66–93)	No
	3	20	0 (0.0)	17 (85.0)	0 (0.0)	3 (15.0)	73 (66–80)	63 (52–78)	No
	4	14	0 (0.0)	14 (100.0)	0 (0.0)	0 (0.0)	65 (58–71)	77 (59–90)	No
	Overall	74	1 (1.4)	48 (64.9)	0 (0.0)	25 (33.8)	70 (58–80)	76 (52–93)
Nebraska
	1	20	0 (0.0)	0 (0.0)	20 (100.0)	0 (0.0)	59 (58–60)	93 (89–96)	No
	2	20	0 (0.0)	19 (95.0)	0 (0.0)	1 (5.0)	67 (63–73)	88 (75–96)	No
	3	20	0 (0.0)	0 (0.0)	5 (25.0)	15 (75.0)	66 (66–67)	100 (99–100)	Yes
	4	20	0 (0.0)	0 (0.0)	10 (50.0)	10 (50.0)	60 (58–65)	93 (83–97)	No
	Overall	80	0 (0.0)	19 (23.8)	35^b (43.8)	26^b (32.5)	63 (58–73)	94 (75–100)

The same pens were sampled at each sampling visit; however, six pens were sent to harvest before sampling visit 4 in Kansas and were unable to be sampled.

During fecal sample collection in Nebraska, some pens were wet and very muddy (n = 35) and some pens (n = 26) contained large pools of standing water (very wet).

E. coli O157:H7 fecal prevalence

Overall, 436 of 3080 (14.2%) fecal samples tested positive for E. coli O157:H7. Unadjusted cumulative fecal prevalence in the KS and NE feedlots was 10.8% (160/1480) and 17.3% (276/1600), respectively. Model-adjusted mean E. coli O157:H7 pen-level prevalence and standard error of the mean (SEM) are reported by treatment, sampling visit, and treatment-by-sampling visit interaction for KS and NE study sites in Table 3. Effects of the DFM product on prevalence of E. coli O157:H7 did not significantly differ by sampling visit in KS (p = 0.77) or NE (p = 0.32); that is, the treatment-by-sampling visit interaction terms were not significant. In KS, mean E. coli O157:H7 fecal prevalence estimates for DFM and control groups were 8.2% (SEM = 2.2%) and 9.9% (SEM = 2.5%), respectively (p = 0.51). At the NE study site, mean E. coli O157:H7 fecal prevalence estimates for DFM and control groups were 14.6% (SEM = 2.8%) and 14.3% (SEM = 2.6%), respectively (p = 0.92). Model-adjusted mean E. coli O157:H7 prevalence estimates significantly differed among sampling visits for data from both KS (p < 0.01) and NE (p < 0.01).

Table 3.

Model-Adjusted^* Mean Escherichia coli O157:H7 Fecal Prevalence (Standard Error of the Mean) by Treatment, Sampling Visit, and Treatment-by-Sampling Visit Interaction for Kansas and Nebraska Study Sites

Variable	Kansas			Nebraska
Variable	Mean prevalence ^† , %	SEM, %	p-Value	Mean prevalence ^† , %	SEM, %	p-Value
Treatment			0.51			0.92
Control	9.9	2.5		14.3	2.6
DFM	8.2	2.2		14.6	2.8
Sampling visit			<0.01			<0.01
1	3.8^a	1.5		28.7^a	4.3
2	14.4^b	3.4		19.7^a,b	3.6
3	12.4^b	3.1		13.2^b	2.9
4	9.1^a,b	2.9		5.2^c	1.7
Treatment × sampling visit interaction			0.77			0.32
Control—sampling visit 1	3.6	1.9		25.0	5.1
Control—sampling visit 2	14.6	4.2		16.0	4.2
Control—sampling visit 3	15.6	4.4		15.6	4.1
Control—sampling visit 4	10.6	4.0		6.2	2.6
DFM—sampling visit 1	4.1	2.0		32.6	5.7
DFM—sampling visit 2	14.2	4.1		24.0	5.1
DFM—sampling visit 3	9.8	3.3		11.1	3.5
DFM—sampling visit 4	7.7	3.4		4.3	2.1

Presented results are from a generalized linear mixed model(s), modeling E. coli O157:H7 fecal prevalence with a binomial distribution and logit link, including treatment, sampling visit, and treatment-by-sampling visit interaction as fixed effects, a random effect accounting for block, and an autoregressive (1) covariance structure for repeated measures. p-Values in bold are statistically significant.

†

Estimates with differing superscripts are significantly different at p < 0.05, estimates with shared letter superscripts do not differ significantly at p < 0.05.

DFM, direct-fed microbial; SEM, standard error of the mean.

E. coli O157 supershedding prevalence

There were 130 (4.2%) of 3073 fecal samples that tested positive for E. coli O157 at a concentration ≥10⁴ CFU per gram of feces. Of samples positive for E. coli O157:H7, 29.8% (130/436) contained supershedding concentrations of E. coli O157. Unadjusted cumulative E. coli O157 supershedding prevalence was 2.3% (34/1473) and 6.0% (96/1600) for the KS and NE study sites, respectively. Seven samples (DFM, n = 3; control, n = 4) from sampling visit 4 at the KS feedlot were culture positive for E. coli O157:H7, but pre-enrichment broth samples were unavailable for quantification. Model-adjusted mean E. coli O157 supershedding pen-level prevalence and SEM are reported by treatment, sampling visit, and treatment-by-sampling visit interaction for KS and NE study sites in Table 4. Effects of treatment on the mean prevalence of E. coli O157 supershedding did not significantly differ by sampling visit in KS (p = 0.97) or NE (p = 0.14). Thus, mean E. coli O157 supershedding prevalence was not significantly reduced by feeding the DFM product at either feedlot. Mean E. coli O157 supershedding prevalence estimates were significantly higher (p = 0.04) for the DFM group in NE (6.7% ± 1.5%) compared with the control group (3.2% ± 1.0%). Supershedding prevalence did not significantly differ between treatment groups in KS (p = 0.66). Model-adjusted mean E. coli O157 supershedding prevalence estimates did not significantly differ across sampling visits (p = 0.18) in KS, but did in NE (p < 0.01).

Table 4.

Model-Adjusted^* Mean Escherichia coli O157 Supershedding Prevalence (Standard Error of the Mean) by Treatment, Sampling Visit, and Treatment-by-Sampling Visit Interaction for Kansas and Nebraska Study Sites

Variable	Kansas			Nebraska
Variable	Mean prevalence ^† , %	SEM, %	p-Value	Mean prevalence ^† , %	SEM, %	p-Value
Treatment			0.66			0.04
Control	1.8	0.7		3.2^b	1.0
DFM	2.2	0.7		6.7^a	1.5
Sampling visit			0.18			<0.01
1	1.5	0.7		10.5^a	2.2
2	3.7	1.1		2.9^b	1.4
3	2.5	0.9		4.4^a,b	1.4
4	1.1	0.7		3.4^b	1.2
Treatment × sampling visit interaction			0.97			0.14
Control—sampling visit 1	1.5	1.0		7.4	2.4
Control—sampling visit 2	3.5	1.5		1.0	0.9
Control—sampling visit 3	2.5	1.3		4.9	2.0
Control—sampling visit 4	0.7	0.8		2.9	1.5
DFM—sampling visit 1	1.5	1.0		14.8	3.4
DFM—sampling visit 2	4.0	1.6		8.3	2.6
DFM—sampling visit 3	2.5	1.3		3.9	1.7
DFM—sampling visit 4	1.5	1.2		3.9	1.7

Presented results are from a GLMM, modeling E. coli O157 supershedding with a binomial distribution and logit link, including treatment, sampling visit, and treatment-by-sampling visit interaction as fixed effects, a random effect accounting for block, and an AR (1) covariance structure for repeated measures. p-Values in bold are statistically significant.

†

Estimates with differing superscripts are significantly different at p < 0.05, estimates with shared letter superscripts do not differ significantly at p < 0.05.

Discussion

The findings of this field study indicated that the prevalence of E. coli O157:H7 fecal shedding and E. coli O157 supershedding was not significantly reduced by including this DFM product in the finishing diet of these commercial feedlot cattle in KS and NE. Mean fecal shedding prevalence of E. coli O157:H7 significantly differed by sampling visit in KS and NE, demonstrating the variability in E. coli O157:H7 fecal shedding patterns over time. The well-documented seasonal and intermittent shedding pattern of E. coli O157:H7 creates inherent challenges in understanding the ecology of the pathogen and potential effectiveness of interventions in cattle production environments (Besser et al., 1997; Hancock et al., 1997; Sargeant et al., 2000). Due to the lack of understanding of the competitive exclusion mechanism of action in the gastrointestinal tract, and with the complexity of E. coli O157:H7 in cattle and feedlot production environments, DFM efficacy has been largely characterized as inconsistent (Callaway et al., 2008). However, given the public health importance and shift away from antibiotic usage in production agriculture, DFM products offer a potential alternative to antibiotics to decrease pathogenic bacterial populations in the host.

For over a decade, DFM studies have been inconsistent in demonstrating effectiveness and repeatability as a preharvest intervention under commercial field conditions. While not always beneficial in reducing fecal prevalence of foodborne pathogens when included in the diet, DFMs may benefit cattle weight gain and feed efficiency (Vasconcelos et al., 2008; Hanford et al., 2011; Cull et al., 2015). The most widely researched DFM products in the field are Lactobacillus based, particularly those including L. acidophilus, formulated at various dosages and administered for a range of durations. While many studies have demonstrated L. acidophilus DFM products to be effective (Brashears et al., 2003; Younts-Dahl et al., 2004, 2005; Stephens et al., 2007a, 2007b; Tabe et al., 2008) in the commercial production environment, others have not (Stephens et al., 2007a, 2010; Cull et al., 2012; Luedtke et al., 2016). The lack of effectiveness in reducing E. coli O157:H7 shedding in this study could be due to many reasons, including study limitations, environmental conditions, and/or a true lack of product effectiveness in commercial feedlot settings.

A priori sample size calculations were estimated based on parameters from the study by Cull et al. (2012), assuming a mean E. coli O157:H7 prevalence of 40% and 25% for control and DFM treatment groups, respectively. In KS and NE, the mean observed control group E. coli O157:H7 prevalence estimates were 9.9% and 14.3%, respectively; therefore, the overall pathogen level observed in our study was much lower than expected, which may have limited our ability to demonstrate intervention effectiveness given the study design. In this current study, supershedding events were compared by dichotomizing qPCR results for estimated E. coli O157 concentrations (≥10⁴ or <10⁴ CFU/g of feces; Noll et al., 2015) within only culture-positive samples, rather than estimating specific concentrations within all samples. Although a more quantitative evaluation of specific concentrations of E. coli O157:H7 in all fecal samples may have provided additional information relative to estimates of pathogen loads, the semiquantitative method employed was useful for addressing this study's primary objective of comparing pen-level prevalence estimates between treatment groups administered different interventions.

A notable factor observed in this study was the amount and difference in rainfall during the sampling periods between the two study feedlot locations (U.S. Climate Data, 2019). The KS feedlot site received ∼15.2 inches of rain between April and August 2018 and nearly 5 inches of rainfall during the sampling period (July to August 2018) with observed pen conditions ranging from dry and dusty to wet (Table 2). However, the feedlot in NE had an even more uncharacteristically wet summer, and resulting pen conditions during the sampling period were very muddy and contained large amounts of standing water, in some areas spanning between pens. At the NE site, nearly 23 inches of rainfall were received during the period of March to September 2018, with 9 inches of rainfall during the 2 months sampled (August and September 2018). It has been documented that ambient temperature and moisture are associated with longer survival of E. coli O157:H7 in the environment, promoting reexposure, and subsequently leading to higher fecal prevalence within the pen (reviewed by Smith, 2014). The large amount of rainfall during this study, particularly for the NE study site, likely affected the prevalence and distribution of E. coli O157:H7 within and among the study pens, perhaps due to the longer survival of E. coli O157:H7 in the environment, lack of complete independence and separation between pens due to standing water/slurry, and reexposure to organisms from the environment and/or grooming contaminated hides. Environmental conditions likely impacted observed prevalence and may have negatively impacted the ability to demonstrate treatment effects among the study pens that received different treatments, but were all located within the same production environment and thus similarly exposed to drivers of prevalence.

Additional factors influencing fecal shedding of E. coli O157:H7 include breed, sex, diet, and stocking density (Smith et al., 2001; Callaway et al., 2009; Jeon et al., 2013). Due to utilization of a randomized complete block design, the distribution of known and unknown risk factors and management factors should be similar across treatment groups within study blocks and feedlot sites. However, several differences existed between feedlots, including differences in sex, pen size, days the DFM product was fed, and WDG included in the diet and the aforementioned differences in environmental conditions. In KS and NE, the amounts of WDG fed in total mixed rations were 22% and 44%, respectively. It has been demonstrated that cattle fed 40% WDG in their diets were associated with significantly higher fecal prevalence of E. coli O157:H7 and supershedding prevalence compared with cattle fed 20% WDG (Jacob et al., 2010a). In addition, sex of study cattle differed between KS and NE study sites as did average pen size and days the DFM product was fed. Thus, the observed variability between KS and NE feedlot results for E. coli O157:H7 prevalence (Table 3) and supershedding prevalence (Table 4) should not be surprising. The variability among these factors, particularly diet and environmental conditions, and the variability of prevalence and shedding over time are all rather typical observations for commercial feedlot production settings and are inherent challenges for evaluating pen-level interventions for reducing E. coli O157:H7 shedding. Given the potential for pathogen survival and dissemination in the environment and between pens, and herd immunity issues that result in control cattle having pathogen levels biased toward the levels in the treated cattle (Peterson et al., 2007; Dodd et al., 2011), future studies of preharvest interventions in commercial feedlots may warrant approaches other than the typical, side-by-side, pen-level study design.

Conclusions

This pen-level field trial did not demonstrate effectiveness for the DFM product reducing E. coli O157:H7 shedding in commercial feedlot cattle. There were no significant reductions in fecal prevalence of E. coli O157:H7 or supershedding prevalence for cattle fed the DFM product versus cattle fed a negative control diet. However, there were significant differences in shedding over time and variability between study sites with regard to cattle and environmental data. While the exact reason(s) for the lack of effectiveness remain unknown, this study illustrates the challenges in demonstrating DFM products as effective preharvest interventions for reducing the prevalence of E. coli O157:H7 in commercial feedlot environments.

Footnotes

Acknowledgments

The authors would like to acknowledge the participating feedlot personnel, field sampling crews (Hannah Seger, Christy Hanthorn, Joaquin Baruch, Pius Ekong, and Tariku Jibat Beyene), laboratory personnel (Leigh Ann Feuerbacher, Xiaorong Shi, and Neil Wallace), and undergraduate students (Monica Anderson and Katie Hoch) for their efforts on this project.

Disclosure Statement

None of the authors have conflicts of interest. However, Kansas State University employees were responsible for the study design, sampling, laboratory work, statistical analysis, result interpretation, and manuscript preparation.

Funding Information

This study was funded by Legacy Animal Nutrition, LLC.

References

Arthur

, Bosilevac

, Kalchayanand

, Wells

, Shackelford

, Wheeler

, Koohmaraie

. Evaluation of a direct-fed microbial product effect on the prevalence and load of Escherichia coli O157:H7 in feedlot cattle. J Food Prot, 2010; 73:366–371.

Bai

, Shi

, Nagaraja

. A multiplex PCR procedure for the detection of six major virulence genes in Escherichia coli O157:H7. J Microbiol Methods, 2010; 82:85–89.

Besser

, Hancock

, Pritchett

, McRae

, Rice

, Tarr

. Duration of detection of fecal excretion of Escherichia coli O157:H7 in cattle. J Infect Dis, 1997; 175:726–729.

Brashears

, Galyean

, Loneragan

, Mann

, Killinger-Mann

. Prevalence of Escherichia coli O157:H7 and performance by beef feedlot cattle given Lactobacillus direct-fed microbials. J Food Prot, 2003; 66:748–754.

Brown

, Edrington

, Genovese

, He

, Anderson

, Nisbet

. Evaluation of the efficacy of three direct fed microbial cocktails to reduce fecal shedding of Escherichia coli O157:H7 in naturally colonized cattle and fecal shedding and peripheral lymph node carriage of Salmonella in experimentally infected cattle. J Food Prot, 2020; 83:28–36.

Callaway

, Carr

, Edrington

, Anderson

, Nisbet

. Diet, Escherichia coli O157:H7, and cattle: A review after 10 years. Curr Issues Mol Biol, 2009; 11:67–79.

Callaway

, Edrington

, Anderson

, Harvey

, Genovese

, Kennedy

, Venn

, Nisbet

. Probiotics, prebiotics and competitive exclusion for prophylaxis against bacterial disease. Anim Health Res Rev, 2008; 9:217–225.

Callaway

, Edrington

, Loneragan

, Carr

, Nisbet

. Review: Shiga toxin-producing Escherichia coli (STEC) ecology in cattle and management based options for reducing fecal shedding. Agric Food Anal Bacteriol, 2013; 3:39–69.

Cernicchiaro

, Pearl

, McEwen

, Zerby

, Fluharty

, Loerch

, Kauffman

, Bard

, LeJeune

. A randomized controlled trial to assess the impact of dietary energy sources, feed supplements, and the presence of super-shedders on the detection of Escherichia coli O157:H7 in feedlot cattle using different diagnostic procedures. Foodborne Pathog Dis, 2010; 7:1071–1081.

10.

Cull

, Paddock

, Nagaraja

, Bello

, Babcock

, Renter

. Efficacy of a vaccine and a direct-fed microbial against fecal shedding of Escherichia coli O157:H7 in a randomized pen-level field trial of commercial feedlot cattle. Vaccine, 2012; 30:6210–6215.

11.

Cull

, Renter

, Bello

, Ives

, Babcock

. Performance and carcass characteristics of commercial feedlot cattle from a study of vaccine and direct-fed microbial effects on Escherichia coli O157:H7 fecal shedding. J Anim Sci, 2015; 93:3144–3151.

12.

Dewsbury

DMA

, Shridhar

, Noll

, Shi

, Nagaraja

, Cernicchiaro

. Summer and winter prevalence of Shiga Toxin-producing Escherichia coli (STEC) O26, O45, O103, O111, O121, O145, and O157 in feces of feedlot cattle. Foodborne Pathog Dis, 2015; 12:726–732.

13.

Dodd

, Sanderson

, Jacob

, Renter

. Modeling preharvest and harvest interventions for Escherichia coli O157 contamination of beef cattle carcasses. J Food Prot, 2011; 74:1422–1433.

14.

Fox

, Renter

, Sanderson

, Nutsch

, Shi

, Nagaraja

. Associations between the presence and magnitude of Escherichia coli O157 in feces at harvest and contamination of preintervention beef carcasses. J Food Prot, 2008; 71:1761–1767.

15.

Hancock

, Besser

, Rice

, Herriott

, Tarr

. A longitudinal study of Escherichia coli O157 in fourteen cattle herds. Epidemiol Infect, 1997; 118:193–195.

16.

Hanford

, Kreikemeier

, Ware

. The effect of Bovamine on feedlot performance of finishing cattle: A meta-analysis. J Anim Sci, 2011; 89(Suppl. 1):258–261.

17.

Jacob

, Paddock

, Renter

, Lechtenberg

, Nagaraja

. Inclusion of dried or wet distillers' grains at different levels in diets of feedlot cattle affects fecal shedding of Escherichia coli O157:H7. Appl Environ Microbiol, 2010a;76:7238–7242.

18.

Jacob

, Renter

, Nagaraja

. Animal- and truckload-level associations between Escherichia coli O157:H7 in feces and on hides at harvest and contamination of preevisceration beef carcasses. J Food Prot, 2010b;73:1030–1037.

19.

Jeon

, Elzo

, DiLorenzo

, Lamb

, Jeong

. Evaluation of animal genetic and physiological factors that affect the prevalence of Escherichia coli O157 in cattle. PLoS One, 2013; 8:e55728.

20.

LeJeune

, Wetzel

. Preharvest control of Escherichia coli O157 in cattle. J Anim Sci, 2007; 85:E73–E80.

21.

Loneragan

, Brashears

. Pre-harvest interventions to reduce carriage of E. coli O157 by harvest-ready feedlot cattle. Meat Sci, 2005; 71:72–78.

22.

Luedtke

, Bosilevac

, Harhay

, Arthur

. Effect of direct-fed microbial dosage on fecal concentrations of enterohemorrhagic Escherichia coli in feedlot cattle. Foodborne Pathog Dis, 2016; 13:190–195.

23.

Marder

, Griffin

, Cieslak

, Dunn

, Hurd

, Jervis

, Lathrop

, Muse

, Ryan

, Smith

, Tobin-D'Angelo

, Vugia

, Holt

, Wolpert

, Tauxe

, Geissler

. Preliminary incidence and trends of infections with pathogens transmitted commonly through food—Foodborne Diseases Active Surveillance Network, 10 U.S. Sites, 2006–2017. Morb Mortal Wkly Rep, 2018; 67:324–328.

24.

Munns

, Selinger

, Stanford

, Selinger

, McAllister

. Are super-shedder feedlot cattle really super?. Foodborne Pathog Dis, 2014; 11:329–331.

25.

Noll

, Shridhar

, Shi

, Baoyan

, Cernicchiaro

, Renter

, Nagaraja

, Bai

. A four-plex real-time PCR assay, based on rfbE, stx1, stx2, and eae genes for the detection and quantification of Shiga toxin-producing Escherichia coli O157 in cattle feces. Foodborne Pathog Dis, 2015; 12:787–794.

26.

Omisakin

, MacRae

, Ogden

, Strachan

NJC

. Concentration and prevalence of Escherichia coli O157 in cattle feces at slaughter. Appl Environ Microbiol, 2003; 69:2444–2447.

27.

Peterson

, Klopfenstein

, Moxley

, Erickson

, Hinkley

, Rogan

, Smith

. Efficacy of dose regimen and observation of herd immunity from a vaccine against Escherichia coli O157:H7 for feedlot cattle. J Food Prot, 2007; 70:2561–2567.

28.

Sargeant

, Gillespie

, Oberst

, Phebus

, Hyatt

, Bohra

, Galland

. Results of a longitudinal study of the prevalence of Escherichia coli O157:H7 on cow-calf farms. Am J Vet Res, 2000; 61:1375–1379.

29.

Smith

, Blackford

, Younts

, Moxley

, Gray

, Hungerford

, Milton

, Klopfenstein

. Ecological relationships between the prevalence of cattle shedding Escherichia coli O157:H7 and characteristics of the cattle or conditions of the feedlot pen. J Food Prot, 2001; 64:1899–1903.

30.

Smith

DR.

Cattle Production Systems: Ecology of existing and emerging Escherichia coli types related to foodborne illness. Annu Rev Anim Biosci, 2014; 2:445–468.

31.

Stephens

, Loneragan

, Chichester

, Brashears

Prevalence and enumeration of Escherichia coli O157 in steers receiving various strains of Lactobacillus-based direct-fed microbials. J Food Prot, 2007a;70:1252–1255.

32.

Stephens

, Loneragan

, Karunasena

, Brashears

. Reduction of Escherichia coli O157 and Salmonella in feces and on hides of feedlot cattle using various doses of a direct-fed microbial. J Food Prot, 2007b;70:2386–2391.

33.

Stephens

, Stanford

, Rode

, Booker

, Vogstad

, Schunicht

, Jim

, Wildman

, Perrett

, McAllister

. Effect of a direct-fed microbial on animal performance, carcass characteristics and the shedding of Escherichia coli O157 by feedlot cattle. Anim Feed Sci Technol, 2010; 158:65–72.

34.

Tabe

, Oloya

, Doetkott

, Bauer

, Gibbs

, Khaitsa

. Comparative effect of direct-fed microbials on fecal shedding of Escherichia coli O157:H7 and Salmonella in naturally infected feedlot cattle. J Food Prot, 2008; 71:539–544.

35.

United States Department of Agriculture, Food Safety and Inspection Service (USDA-FSIS). 9 CFR Parts 416, 417, and 430. Fed Regist, 2012; 77:31975–31981.

36.

US Climate Data. 2019. Available at: www.usclimatedata.com, accessed January 8, 2019 .

37.

Vasconcelos

, Elam

, Brashears

, Galyean

. Effects of increasing dose of live cultures of Lactobacillus acidophilus (stain NP 51) combined with a single dose of Propionibacterium freudenreichii (strain NP 24) on performance and carcass characteristics of finishing beef steers. J Anim Sci, 2008; 86:756–762.

38.

Williams

, Ward

, Dhungyel

. Longitudinal study of Escherichia coli O157 shedding and super shedding in dairy heifers. J Food Prot, 2014; 78:636–642.

39.

Wisener

, Sargeant

, O'Connor

, Faires

, Glass-Kaastra

. The use of direct-fed microbials to reduce shedding of Escherichia coli O157 in beef cattle: A systematic review and meta-analysis. Zoonoses Public Health, 2015; 62:75–89.

40.

Withee

, Williams

, Disney

, Schlosser

, Bauer

, Ebel

. Streamlined analysis for evaluating the use of preharvest interventions intended to prevent Escherichia coli O157:H7 illness in humans. Foodborn Pathog Dis, 2009; 6:817–825.

41.

Younts-Dahl

, Galyean

, Loneragan

, Elam

, Brashears

. Dietary supplementation with Lactobacillus-and Propionibacterium-based direct-fed microbials and prevalence of Escherichia coli O157 in beef feedlot cattle and on hides at harvest. J Food Prot, 2004; 67:889–893.

42.

Younts-Dahl

, Osborn

, Galyean

, Rivera

, Loneragan

, Brashears

. Reduction of Escherichia coli O157 in finishing beef cattle by various doses of Lactobacillus acidophilus in direct-fed microbials. J Food Prot, 2005; 68:6–10.