Prevalence and Antimicrobial Resistance of Salmonella Isolated from Cattle Feces in United States Feedlots in 2011

Abstract

The objective of this study was to determine the prevalence and characteristics of Salmonella spp. isolated from feces of cattle in feedlots in the United States. Fecal samples were collected from up to three pens of cattle in each of 68 feedlots in 12 states. Samples included up to 25 individual fecal pats from the pen floors and up to five composite samples from the floors of the same pens. The prevalence of Salmonella-positive samples was 9.1% (460/5050) and 11.3% (114/1009) for individual and composite samples, respectively. The prevalences of Salmonella at the pen level were 35.6% (72/202) and 22.8% (46/202) for individual and composite samples, respectively. Dietary factors, including inclusion of cottonseed hulls, coccidiostats, and antimicrobial drugs, were associated with differences in prevalence of Salmonella isolation. Overall, 32 serotypes of Salmonella were identified, but six serotypes accounted for 69.1% (495/716) of the isolates. Nearly two-thirds (64.7%, 44/68) of feedlots had at least one positive sample. All isolates were evaluated for susceptibility to a panel of 15 antimicrobial drugs. Most isolates (74.4%, 533/716) were susceptible to all antimicrobial drugs in the panel. When resistance was detected, it was most commonly to tetracycline (21.7%, 155/716 of isolates) or sulfisoxazole (12.4%, 89/716 of isolates). Less than 10% of the isolates were resistant to any other antimicrobials in the panel. The results of this study indicate that the prevalence of Salmonella in individual fecal samples was less than 10%, but that Salmonella is widely distributed among feedlot cattle. Furthermore, when Salmonella is present in feedlot cattle, there is a low occurrence of antimicrobial resistance with the exception of tetracycline and sulfisoxazole. More research is indicated to understand the ecology of Salmonella and antimicrobial resistance, when present, in cattle-feeding operations.

Introduction

Salmonellae are a cause of animal and human disease. In the United States salmonellae cause an estimated 1.03 million cases of domestically acquired foodborne illness in humans per year resulting in 19,336 hospitalizations and 378 deaths (Scallan et al., 2011). Substantial efforts have been devoted to application of strategies in slaughter facilities to lower the burden of foodborne illness caused by Salmonella (Stopforth and Sofos, 2006). Increasingly, there is interest in applying strategies even earlier in the production cycle (i.e., preharvest) to augment postharvest procedures (Sofos, 2002; Oliver et al., 2008). Effective mitigation of risks earlier in the production system requires a better understanding of the ecology of Salmonella at all phases of the animal production environment.

Previous studies of cattle feedlots have reported on prevalence and characteristics of Salmonella in cattle feedlots in 1994 and 1999 (Dargatz et al., 2002; Dargatz et al., 2003; Van Donkersgoed, 2009; Rao, 2010). There is potential for much to have changed over the preceding 12 years with regard to prevalent serotypes and their characteristics. Some of these changes may be related to the inclusion of new nutritional management strategies and other contemporary management changes for cattle in feedlots. In addition, as in other ecologic settings (e.g., other livestock production streams or public health), there appears to be temporal variation in serotype distribution and characteristics that remains unexplained. For these and other reasons, there is a need for updated information for feedlot cattle in the United States. Furthermore, there is a need to identify optimal surveillance strategies for pathogen prevalence. Collection and testing of large numbers of samples can be cumbersome and expensive. The objectives of this study were (1) to provide more current estimates of the prevalence of Salmonella in cattle feedlots in the United States, (2) to describe the serotype and resistance characteristics of Salmonella isolates, (3) to compare these data when collected by two sampling strategies (individual and composite samples), and (4) to investigate regional and dietary associations with Salmonella prevalence.

Materials and Methods

Sample collection

Feedlots participating in the USDA National Animal Health Monitoring System (NAHMS) Feedlot 2011 study were eligible to participate in the pathogen sampling component of the study (USDA, 2013). Sixty-eight feedlots distributed across 12 states were recruited for sample collection. In each feedlot three pens of cattle were selected for sampling, including the pen of cattle that had been in the feedlot the shortest time, the pen of cattle that had been in the feedlot the longest time, and a randomly selected pen (pen type for subsequent analysis). Samples were collected between April 2 and August 15, 2011. In each pen up to 25 fecal samples were collected from individual fecal pats from the pen floor during a serpentine walk of the pen (individual samples). Efforts were made to minimize the chance of collecting multiple samples from the same animal. An additional five composite samples were collected from fecal pats on the pen floor by conducting another serpentine walk of the pen and aggregating feces from up to five fecal pats into a single collection (composite samples). Samples (individual or composite) were collected in separate plastic bags, chilled, and transported by overnight delivery to the USDA–ARS Bacterial Epidemiology and Antimicrobial Resistance research laboratory for bacterial culture and initial characterization beginning on the day of arrival.

Salmonella isolation

Fecal samples were cultured as previously described (Wells et al., 2001). Fecal samples were enriched in selective media (1 g of sample in 10 mL of broth: GN Hajna incubated for 24 h at 37°C and Tetrathionate Broth incubated 48 h at 37°C), and 0.1 mL of each selective enrichment culture was transferred to Rappaport R-10 broth (10 mL) for a secondary selective enrichment and incubated for 24 h at 37°C. Broth was streaked for isolation on selective brilliant green sulfa agar (BGS) and Xylose Lysine Tergitol 4 (XLT4) agar plates and incubated for 24 h at 37°C. On BGS plates, presumptive Salmonella colonies were lactose negative, forming a pink colony, and on XLT4 plates, presumptive Salmonella colonies were typically yellow to red with a black center (indicating hydrogen sulfide production). Up to four presumptive Salmonella colonies were inoculated on triple sugar iron (TSI) and lysine iron agar slants and incubated for 24 h at 37°C. On TSI, the presence of Salmonella produced a K/A+ alkaline (red) reaction on the slant surface and an acid (yellow) reaction with hydrogen sulfide production in the stabbed agar butt. Up to four presumptive Salmonella colonies from each sample were serogrouped by slide agglutination with serogroup-specific antisera (Difco; BD; personal communication). When all isolates were of the same serogroup, a single isolate was chosen for serotyping.

Serotyping of the isolates

All the isolates were submitted to the USDA–APHIS National Veterinary Services Laboratories in Ames, IA, for serotyping using previously described methods (Ewing, 1986). Antigenic formulae for somatic (O) and flagellar (H) antigens were used to determine serotype (Grimont and Weill, 2007).

Antimicrobial susceptibility testing

The susceptibility of each of the isolates was tested using a semiautomated system (Sensititre; TREK Diagnostic Systems, Inc.) using a panel of 15 antimicrobial drugs. To allow comparison with other antimicrobial resistance surveillance systems, the Sensititre Gram Negative Plate (CMV3AGPF) used in the National Antimicrobial Resistance Monitoring System (NARMS) was used and included the following antimicrobial agents: amoxicillin–clavulanic acid, ampicillin, azithromycin, cefoxitin, ceftiofur, ceftriaxone, chloramphenicol, ciprofloxacin, gentamicin, kanamycin, nalidixic acid, streptomycin, sulfisoxazole, tetracycline, and trimethoprim–sulfamethoxazole. Breakpoints used in the NARMS program were used to classify the isolates as susceptible, intermediate, or resistant (FDA, 2013). The quality control organism used for the study was Escherichia coli ATCC 25922. Isolates were classified as multidrug resistant if they were resistant to three or more antimicrobial drugs in the panel.

Statistical analysis

At the time of sample collection, a questionnaire was administered to collect data on animal gender, breed type, average placement weight, average current weight, use of E. coli or Salmonella vaccines, metaphylaxis,¹ morbidity levels, mortality levels, single source or multisource, chlorination of drinking water, feed ingredients used in the last 7 d, and any use of specific feed or water antimicrobials since arrival (questionnaire available at www.aphis.usda.gov/aphis/ourfocus/animalhealth/monitoring-and-surveillance/nahms/feedlot_questionnaires). Questionnaire data were merged with the laboratory results from the sample testing. For the purposes of this analysis, isolates that were classified as intermediate based on their minimum inhibitory concentration value were collapsed into the susceptible category. Feedlot locations were attributed to regions in two ways. First, feedlots in Colorado, Kansas, Nebraska, Oklahoma, and Texas were classified into the Central region, whereas feedlots in Arizona, California, Iowa, Idaho, New Mexico, South Dakota, and Washington were classified in the Other region. Second, to explore the potential influence of latitude, feedlots in California, Texas, Oklahoma, New Mexico, and Arizona were classified into the South region, whereas feedlots in Colorado, Iowa, Idaho, Kansas, Nebraska, South Dakota, and Washington were classified in the North region.

Analyses were done at the sample level and at the pen level—separately for individual and composite samples—using SAS version 9.3 (SAS Institute version 9.3). For sample-level inferential analyses, logistic regression models were constructed using SAS PROC GENMOD, with samples nested within pens and feedlots. Variables significant at p < 0.10 in univariable models were included in a multivariable model, and variables significant at p < 0.05 were retained in the final multivariable model after backward elimination. Pen-level inferential analyses were done similarly, with pens nested within feedlots.

Results

Prevalence of Salmonella

Overall 5050 individual and 1009 composite samples were collected from 202 pens in 68 feedlots (Table 1). Among the 68 feedlots, 44 had at least one Salmonella-positive sample. Individual samples were positive on 41 feedlots and composite samples were positive on 30 feedlots.

Table 1.

Pen- and Sample-Level Prevalence for Individual and Composite Samples by Type of Pen

	Pen				Sample
Sample type	Individual		Composite		Individual		Composite
Pen type	No. of pens examined	Percent positive	No. of pens examined	Percent positive	No. of samples examined	Percent positive	No. of samples examined	Percent positive
Short-fed	68	39.7	68	25.0	1700	8.5	339	10.0
Long-fed	68	33.8	68	22.1	1700	9.8	340	14.4
Random	66	33.3	66	21.2	1650	9.0	330	9.4
Total	202	35.6	202	22.8	5050	9.1	1009	11.3

The overall pen-level prevalences for individual samples and composite samples were 35.6% (72/202) and 22.8% (46/202), respectively, with no differences by type of pen (p = 0.59 and p = 0.88, respectively). Pens in feedlots from the Central region were more likely to have a positive individual sample (42.4%, p = 0.043) compared with pens from the Other region (24.7%). The prevalence of positive pens based on individual samples was not different among the North and South regions. However, pens from the South region were more likely to have positive composite samples (39.2%, p = 0.028) compared with pens from the North region (17.2%).

The sample-level prevalences for individual and composite samples were 9.1% and 11.3%, respectively, again with no differences by type of pen (p = 0.93 and p = 0.55, respectively). The prevalence of positive individual samples was higher from the Central region (11.3%, p = 0.045) compared with the Other region (5.6%) (Table 2). Similarly, the prevalence of positive composite samples was also higher in the Central region (14.1%, p = 0.034) compared with the Other region (6.8%). Considering feedlot latitude, the prevalence of positive individual samples was higher in the South region (20.0%, p = 0.003) compared with the North region (5.4%). The prevalence of positive composite samples was also higher in the South region (25.9%, p = 0.001) compared with the North region (6.4%).

Table 2.

Pen- and Sample-Level Prevalence for Individual and Composite Samples by Regions

	Pen				Sample
Sample type	Individual		Composite		Individual		Composite
Region	No. of pens examined	Percent positive	No. of pens examined	Percent positive	No. of samples examined	Percent positive	No. of samples examined	Percent positive
Central	125	42.4	125	26.4	3125	11.3	624	14.1
Other	77	24.7	77	16.9	1925	5.6	385	6.8
South	51	45.1	51	39.2	1275	20.0	255	25.9
North	151	32.5	151	17.2	3775	5.4	754	6.4

Serotyping of isolates

Among the 716 Salmonella isolates from the individual and composite samples (n = 6059), there were 33 serotypes represented. The top six serotypes comprised 69.1% of the isolates (Table 3). The relative proportions of isolates by serotype were similar regardless of sample type.

Table 3.

Salmonella Serotypes Isolated

	All samples		Individual samples		Composite samples
Serotype	No. isolates	%	No. isolates	%	No. isolates	%
Montevideo	129	18.0	98	17.2	31	21.4
Anatum	127	17.7	103	18.1	24	16.6
Kentucky	108	15.1	87	15.2	21	14.5
Idikan	48	6.7	38	6.7	10	6.9
Cerro	43	6.0	35	6.1	8	5.5
Mbandaka	40	5.6	31	5.4	9	6.2
Meleagridis	39	5.4	31	5.4	8	5.5
Typhimurium	30	4.2	23	4.0	7	4.8
Muenster	24	3.4	19	3.3	5	3.4
Newport	19	2.7	15	2.6	4	2.8
Others^a	109	15.2	91	15.9	18	12.4
Total	716	100.0	571	100.0	145	100.0

6;7:g;m;s:e;n;z15, Worthington, Agona, Infantis, Schwarzengrund, Reading, Dublin, Give, Muenchen, Sundsvall, Bareilly, Derby, Minnesota, Thompson, Uganda, Cubana, Javiana, Kiambu, Lexington var 15+, Orion, Rough 0:l;v:1;7, Senftenberg, and Soerenga.

Antimicrobial resistance phenotypes

Most of the Salmonella isolates (74.4%; 533/716) were susceptible to all antimicrobial drugs in the panel tested (Table 4). An additional 16.3% (117/716) of isolates were resistant to only one antimicrobial drug tested. When resistance was present it was most commonly to tetracycline (21.7%; 155/216) or sulfisoxazole (12.4%; 89/216) (Table 5). The resistance characteristics of Salmonella isolates were similar regardless of whether they were from individual or composite samples. Isolates among some serotypes were more likely to be multidrug resistant (i.e., resistant to three or more antimicrobials) (Table 6).

Table 4.

Number and Percentage of Isolates by Number of Antimicrobial Drugs to Which They Were Resistant

Resistance number	No. isolates	%
0	533	74.4
1	117	16.3
2	4	0.6
3	2	0.3
4	7	1.0
6	1	0.1
8	3	0.4
9	40	5.6
10	8	1.1
11	1	0.1

Table 5.

Percentage of Resistant Isolates by Antimicrobial

Antimicrobial	No. isolates resistant (total n = 716)	Percent resistant
Amoxicillin–clavulanic acid	52	7.3
Ampicillin	53	7.4
Azithromycin	0	0
Cefoxitin	52	7.3
Ceftiofur	52	7.3
Ceftriaxone	52	7.3
Chloramphenicol	60	8.4
Ciprofloxacin	3	0.4
Gentamicin	0	0.0
Kanamycin	7	1.0
Nalidixic acid	6	0.8
Streptomycin	58	8.1
Sulfisoxazole	89	12.4
Tetracycline	155	21.7
Trimethoprim–sulfamethoxazole	1	0.1

Table 6.

Salmonella Serotypes Resistant to Three or More Antimicrobials

Serotype	Number isolates	Isolates resistant to 3 or more antimicrobials	% Resistant to 3 or more antimicrobials
Montevideo	129	5	3.9
Anatum	127	0	0
Kentucky	108	0	0
Idikan	48	0	0
Cerro	43	0	0
Mbandaka	40	0	0
Meleagridis	39	5	12.8
Typhimurium	30	17	56.7
Muenster	24	1	4.2
Newport	19	17	89.5
Others^a	109	17	15.6
Total	716	62	8.7

Includes serotypes 6;7:g;m;s:e;n;z15, Worthington, Agona, Infantis, Schwarzengrund, Reading, Dublin, Give, Muenchen, Sundsvall, Bareilly, Derby, Minnesota, Thompson, Uganda, Cubana, Javiana, Kiambu, Lexington var 15+, Orion, Rough 0:l;v:1;7, Senftenberg, and Soerenga.

For serotypes represented by at least 10 isolates, the highest proportion of multidrug resistance was seen for serotypes Newport (89.5%, 17/19), Typhimurium (56.7%, 17/30), Infantis (33.3%, 4/12), Meleagridis (12.8%, 5/39), Muenster (4.2%, 1/24), and Montevideo (3.9%, 5/129). Among the 49 isolates that were resistant to 9 or more antimicrobial drugs, the serotypes were Montevideo (n = 3), Typhimurium (n = 14), Newport (n = 15), Infantis (n = 4), Reading (n = 7), Dublin (n = 4), Uganda (n = 1), and S. Rough O:I;v:1;7 (n = 1).

Factors associated with Salmonella presence

Modeling the prevalence of Salmonella in individual samples, three factors remained in the multivariable model (Table 7). Ration inclusion of cottonseed hulls in the last 7 d (increasing prevalence), inclusion of decoquinate (decreasing prevalence), and inclusion of oxytetracycline (decreasing prevalence) any time since arrival were associated with the outcome.

Table 7.

Factors Associated with Prevalence of Positive Samples or Pens

Sample type	Level	Factor	Number of samples	Factor level	Predicted percent positive	p
Individual samples	Sample	Cotton seed hulls	250	Yes	13.0	0.026
			4800	No	1.3
		Decoquinate	250	Yes	1.4	0.006
			4800	No	12.6
		Oxytetracycline	275	Yes	2.3	0.018
			4775	No	8.0
Composite samples	Sample	Beta-agonist	121	Yes	32.7	0.016
			888	No	8.3
		Cotton seed hulls	50	Yes	36.4	0.019
			959	No	7.1
		Decoquinate	50	Yes	6.9	0.027
			959	No	37.1
		Region	255	South	27.8	0.014
			754	North	10.4
Individual samples	Pen	Lasalocid	16	Yes	64.0	0.047
			186	No	33.3
Composite samples	Pen	Beta-agonist	24	Yes	57.0	0.019
			178	No	23.3
		Days on feed (coeff. = −0.0051)				0.009
				80 d	44.0
				160 d	34.3
		Region	51	South	54.9	0.013
			151	North	24.8

In modeling the prevalence of Salmonella in composite samples, four factors remained in the model, including the inclusion of a beta-agonist in the last 7 d (increasing prevalence), inclusion of cottonseed hulls in the last 7 d (increasing prevalence), inclusion of decoquinate (decreasing prevalence) any time since arrival, and being located in the South region (increasing prevalence).

For the model of pen-level Salmonella prevalence based on results from individual samples, only one factor remained in the model. Pens that had lasalocid in the ration any time since arrival were more likely to be positive than pens not receiving lasalocid.

Finally, for the model of pen-level prevalence based on results of composite samples, three factors remain in the model. Ration inclusion of a beta-agonist in the last 7 d (increasing prevalence), longer time on feed (decreasing prevalence), and being located in the South region (increasing prevalence) were associated with the outcome.

Factors associated with Salmonella resistance

Modeling the prevalence of Salmonella resistant to tetracycline or sulfisoxazole in individual samples, two factors remained in the multivariable models (Table 8)—ration inclusion of cottonseed hulls in the last 7 d (decreasing prevalence) and region location of the feedlot. Samples positive for Salmonella from animals collected in feedlots in the Central region were more likely to have tetracycline- or sulfisoxazole-resistant Salmonella than samples with Salmonella from feedlots in the Other region.

Table 8.

Factors Associated with Prevalence of Samples with Resistant Salmonellae

Antimicrobial resistance	Factor	Number of samples	Level	Predicted percent positive	p
Sulfisoxazole	Cotton seed hulls	232	Yes	0.7	0.007
		484	No	12.4
	Region	563	Central	6.7	0.009
		153	Other	1.4
Tetracycline	Cotton seed hulls	232	Yes	0.7	0.006
		484	No	19.2
	Region	563	Central	7.9	0.020
		153	Other	1.9

Discussion

The sample and pen-level prevalence of Salmonella from this study were similar to those described for feedlots from 1999 (Dargatz et al., 2003). Prevalence estimates were similar regardless of whether individual or composite samples were used. This result suggests that when the outcome of interest is an overall prevalence, and distribution of serotypes across a broad population of animals in multiple facilities that a more cost-effective design for sample collection could involve the collection of composite samples from the pen floors rather than the collection of a larger number of samples representing individual animals. While the precision of the estimates may be lower with composite sampling and some rare serotypes may not be identified, the overall description of prevalence and distribution are not severely compromised.

Relatively few variables were associated with finding Salmonella either in individual samples or in composite samples collected in cattle feedlots. Most of the variables associated with positive samples were related to diet components. Some of these dietary factors were the same or similar to those associated with Salmonella-positive samples from two previous similar studies (Losinger et al., 1997; Green et al., 2010). In the 1994 and 1999 studies, the odds of a positive sample were 3.5 and 5.2 times higher if cottonseed hulls were included in the ration. Cottonseeds or cottonseed meal have been associated with Salmonella contamination in a number of studies evaluating the prevalence of contamination of animal feed components (Van et al., 1966; Rutqvist et al., 1983; Jardy and Michard, 1992; Davis et al., 2003; Jones and Richardson, 2004; Myint et al., 2007).

In the current study, as with the 1999 study, the use of a coccidiostat—decoquinate—in the ration decreased the likelihood of a sample positive for Salmonella. However, in some cases the inclusion of lasalocid, another compound with anticoccidial activity, was associated with a higher expected percentage of positive samples. Inclusion of coccidiostats in the diet of poultry has also been associated with decreased prevalence of Salmonella shedding (Bolder et al., 1999). Inclusion of tetracycline in the ration was associated with a decreased likelihood of positive samples in both studies as well. The finding of a higher predicted percentage of positive samples with animals receiving a beta agonist is consistent with findings from others (Edrington et al., 2006).

Samples from more southern latitudes were more likely to be positive for Salmonella in the current study and in the 1994 and 1999 studies. This observation is consistent with anecdotal observations and other research to document the presence of Salmonella in samples from cattle in feedlots across a broad geographic continuum from the southern edge of the United States, where the prevalence is expected to be higher to Canada where the prevalence is very low. The reasons for this regional variation are unclear at this time.

Relatively few of the Salmonella isolates were resistant to one or more of the antimicrobial drugs tested in the panel. Less than 10% of the isolates were resistant to antimicrobial drugs considered critical for treatment of human salmonellosis, such as cephalosporins and fluoroquinolones. However, the low frequency of resistance preclude an extensive look at the factors associated with resistance among the isolates. While the prevalence of Salmonella was positively associated with the inclusion of cottonseed or cottonseed meal in the ration, the salmonellae were less likely to be resistant, with 89.3% (293/328) of those isolates being pansusceptible. The most common serotypes of Salmonella among the cattle pens exposed to cottonseed or cottonseed meal were S. Anatum (29.9% = 98/328), followed by S. Kentucky (19.8% = 65/328), S. Montevideo (19.5% = 64/328), and S. Cerro (12.8% = 42/328). The remainder of the isolates (18.0% = 59/328) were from 10 additional serotypes. More work is indicated to define those modifiable factors that might serve to mitigate the risk of both resistant and nonresistant Salmonella shedding by feedlot cattle.

Footnotes

Acknowledgments

The skilled technical assistance of Sandra House and Takiyah Ball of the USDA–ARS Bacterial Epidemiology and Antimicrobial Resistance unit is gratefully acknowledged.

Disclosure Statement

No competing financial interests exist.

References

Bolder

, Wagenaar

, Putirulan

, Veldman

, Sommer

. The effect of flavophospholipol (Flavomycin®) and salinomycin sodium (Sacox®) on the excretion of Clostridium perfringens, Salmonella enteritidis, and Campylobacter jejuni in broilers after experimental infection. Poultry Sci, 1999; 78:1681–1689.

Dargatz

, Fedorka-Cray

, Ladely

, Ferris

, Green

, Headrick

. Antimicrobial susceptibility patterns of Salmonella isolates from cattle in feedlots. J Am Vet Med Assoc, 2002; 221:268–272.

Dargatz

, Fedorka-Cray

, Ladely

, Kopral

, Ferris

, Headrick

. Prevalence and antimicrobial susceptibility of Salmonella spp. isolates from U.S. cattle in feedlots in 1999 and 2000. J Appl Microbiol, 2003; 95:753–761.

Davis

, Hancock

, Rice

, Call

, DiGiacomo

, Samadpour

, Besser

. Feedstuffs as a vehicle of cattle exposure to Escherichia coli O157:H7 and Salmonella enterica . Vet Microbiol, 2003; 95:199–210.

Edrington

, Callaway

, Ives

, Engler

, Welsh

, Hallford

, Genovese

, Anderson

, Nisbet

. Effect of ractopamine HCl supplementation on fecal shedding of Escherichia coli O157:H7 and Salmonella in feedlot cattle. Current Microbiol, 2006; 53:340–345.

Ewing

. Edwards and Ewing's identification of Enterobacteriaceae. 4th ed. New York: Elsevier Science, 1986.

FDA. National Antimicrobial Resistance Monitoring System–Enteric Bacteria (NARMS): 2011 Executive Report. Rockville, MD: U.S. Department of Health and Human Services, Food and Drug Administration, 2013.

Green

, Dargatz

, Wagner

, Hyatt

, Fedorka-Cray

, Ladely

, Kopral

. Analysis of risk factors for Salmonella spp. isolates from U.S. feedlot cattle. Foodborne Path and Dis, 2010; 7:825–833.

Grimont

PAD

, Weill

. Antigenic Formulae of the Salmonella Serovars . 9th ed. Paris: World Health Organization Collaborating Centre for Reference and Research on Salmonella, 2007.

10.

Jardy

, Michard

. Salmonella contamination in raw feed components. Microbiol Aliments Nutr, 1992; 10:233–240.

11.

Jones

, Richardson

. Salmonella in commercially manufactured feeds. Poultry Sci, 2004; 83:384–391.

12.

Losinger

, Garber

, Smith

, Hurd

, Biehl

, Fedorka-Cray

, Thomas

, Ferris

. Management and nutritional factors associated with the detection of Salmonella sp. from cattle fecal specimens from feedlot operations in the United States. Prev Vet Med, 1997; 31:231–244.

13.

Myint

, Johnson

, Paige

, Bautista

. A cross-sectional study of bacterial contamination in plant-protein feed from feed stores in Northern Virginia and Maryland. Anim Feed Sci Technol, 2007; 133:137–148.

14.

Oliver

, Patel

, Callaway

, Torrence

. ASAS centennial paper: Developments and future outlook for preharvest food safety. J Anim Sci, 2009; 87:419–437.

15.

Rao

, Van Donkersgoed

, Bohaychuk

, Besser

, Song

X-M

, Wagner

, Hancock

, Renter

, Dargatz

, Morley

. Antimicrobial drug use and antimicrobial resistance in enteric bacteria among cattle from alberta feedlots. Foodborne Pathog Dis, 2010; 7:449–457.

16.

Rutqvist

, Holmberg

, Fallquist

U-

. Aflatoxin and Salmonella in animal feeds commercially available in Sweden. Svensk Veterinärtidning, 1983; 35:363–366.

17.

Scallan

, Hoekstra

, Angulo

, Tauxe

, Widdowson

M-A

, Roy

, et al. Foodborne illness acquired in the United States—major pathogens. Emerg Infect Dis, 2011; 17:1338–1339.

18.

Sofos

. Approaches to pre-harvest food safety assurance. In Food Safety Assurance and Veterinary Public Health, v. 1, Food Safety Assurance in the Pre-Harvest Phase, Smulders

FJM

& Collins

(Eds.). Wageningen, The Netherlands: Wageningen Academic Publishers, 2002, pp 23–48.

19.

Stopforth

, Sofos

. Recent advances in pre- and post-slaughter intervention strategies for control of meat contamination. In: Advances in Microbial Food Safety, ACS Symposium 931. Recent Advances in Intervention Strategies to Improve Food Safety, Juneja

, Cherry

, Tunick

(Eds.). Washington, DC: American Chemical Society, Oxford University Press, 2006. pp. 66–86.

20.

USDA. Feedlot 2011 “Part IV: Health and Health Management on U.S. Feedlots with a Capacity of 1,000 or More Head” USDA–APHIS–VS–CEAH–NAHMS. Fort Collins, CO #638.0913, 2013.

21.

Van

, Frick

, Van

. Steam pelleting of pig feed against Salmonella infection. Proc World Congress Anim Feed, 1966; 2:359–3646.

22.

Van Donkersgoed

, Bohaychuk

, Besser

, Song

, Wagner

, Hancock

, Renter

, Dargatz

. Occurrence of foodborne bacteria in Alberta feedlots. Can Vet J, 2009; 50:166–172.

23.

Wells

S.J.

, Fedorka-Cray

P.J.

, Dargatz

D.A.

, Ferris

, and Green

Fecal shedding of Salmonella spp. by dairy cows on farm and at cull cow markets. J Food Prot, 2001; 64:3–11.