Retrospective Analysis of Salmonella,Campylobacter,Escherichia coli,and Enterococcus in Animal Feed Ingredients

Abstract

The presence and antimicrobial susceptibility of foodborne pathogens and indicator organisms in animal feed are not well understood. In this study, a total of 201 feed ingredient samples (animal byproducts, n=122; plant byproducts, n=79) were collected in 2002 and 2003 from representative rendering plants and the oilseed (or cereal grain) industry across the United States. The occurrence and antimicrobial susceptibility of four bacterial genera (Salmonella, Campylobacter, Escherichia coli, and Enterococcus) were determined. Salmonella isolates were further characterized by serotyping and pulsed-field gel electrophoresis (PFGE). None of the samples yielded Campylobacter or E. coli O157:H7, whereas Salmonella, generic E. coli, and Enterococcus were present in 22.9%, 39.3%, and 86.6% of samples, respectively. A large percentage (47.8%) of Salmonella-positive samples harbored two serovars, and the vast majority (88.4%) of Enterococcus isolates were E. faecium. Animal byproducts had a significantly higher Salmonella contamination rate (34.4%) than plant byproducts (5.1%) (p<0.05). Among 74 Salmonella isolates recovered, 27 serovars and 55 PFGE patterns were identified; all were pan-susceptible to 17 antimicrobials tested. E. coli isolates (n=131) demonstrated similar susceptibility to these antimicrobials except for tetracycline (15.3% resistance), sulfamethoxazole (7.6%), streptomycin (4.6%), ampicillin (3.8%), and nalidixic acid (1.5%). Enterococcus isolates (n=362) were also resistant to five of 17 antimicrobials tested, ranging from 1.1% to penicillin to 14.6% to tetracycline. Resistance rates were generally higher among isolates recovered from animal byproducts. Taken together, our findings suggest that diverse populations of Salmonella, E. coli, and Enterococcus are commonly present in animal feed ingredients, but antimicrobial resistance is not common. Future large-scale studies to monitor these pathogenic and indicator organisms in feed commodities is warranted.

Introduction

T he United States is one of the largest animal feed producers in the world (Best, 2012). Over the last 10 years, the U.S. feed production estimate has ranged from 116.0 million to 124.7 million metric tons annually (Muirhead, 2012). Ingredients used to formulate animal feed rations are derived primarily from the processing of plants (e.g., grains, oilseed meals) and animals (e.g., rendered products, fish meal) (Sapkota et al., 2007). By using these byproducts as feed ingredients, the industry efficiently recycles what would otherwise be waste materials into valuable products for agriculture (Muirhead, 2003).

However, growing concerns about feed safety have been raised recently, including, but not limited to, the relationship between bacterial contamination of animal feed and human foodborne illness (Crump et al., 2002), and the contribution of feed to the development and dissemination of antimicrobial resistance (The Panel on Biological Hazards, 2008). Given that feed ingredients cover a wide range of raw materials, it is not surprising for them to harbor enteric pathogens such as Salmonella (Sapkota et al., 2007). In a 1993 survey, the U.S. Food and Drug Administration's (FDA) Center for Veterinary Medicine (FDA/CVM) detected Salmonella in close to 50% of 151 feed ingredients sampled (McChesney et al., 1995). Numerous other studies worldwide also documented the presence of Salmonella in various feed products (Crump et al., 2002; The Panel on Biological Hazards, 2008; Williams, 1981). In contrast, the occurrence of other foodborne pathogens such as Campylobacter and pathogenic E. coli in feed is much less well documented (Davis et al., 2003; The Panel on Biological Hazards, 2008). Additionally, there are scarce data on the antimicrobial susceptibility of foodborne pathogenic and indicator organisms in animal feed. In that regard, Escherichia coli and Enterococcus are often tested as sentinel organisms due to their ubiquitous nature in animals and humans and their potential to serve as resistance gene reservoirs for human enteric pathogens (Sapkota et al., 2011; Zhao et al., 2012).

The objectives of this study were to determine the occurrence of Salmonella, Campylobacter, E. coli, and Enterococcus in animal- and plant-derived feed ingredients, to further elucidate Salmonella serovars present among these commodities, and also to characterize the antimicrobial susceptibility profiles of the organisms recovered.

Materials and Methods

Sample collection and preparation

In 2002 and 2003, a total of 201 feed ingredient samples of animal (n=122) and plant (n=79) origins (Table 1) were randomly collected by FDA field investigators from representative rendering plants and the oilseed (or cereal grain) industry across the United States. Animal byproducts were of six types, while plant byproducts were of three main types and 10 subtypes (Table 1).

Table 1.

Prevalence of Salmonella, Escherichia coli, and Enterococcus in 201 Feed Ingredient Samples

		No. (%) of positive samples for: ^b
Feed type ^a	No. of samples	Salmonella	Escherichia coli	Enterococcus
Animal derived	122	42 (34.4)	45 (36.9)	103 (84.4)
Meat and bone meal	72	28 (38.9)^A	26 (36.1)^A	62 (86.1)^A
Poultry meal	17	3 (17.6)^A	5 (29.4)^A	15 (88.2)^A
Blood meal	16	5 (31.3)^A	6 (37.5)^A	13 (81.3)^A
Feather meal	10	1 (10)	3 (30)	7 (70)
Fish meal	5	4 (80)	4 (80)	5 (100)
Bone meal	2	1 (50)	1 (50)	1 (50)
Plant derived	79	4 (5.1)	34 (43)	71 (89.9)
Alfalfa meal	13	—	6 (46.2)^AB	13 (100)^A
Oilseed byproducts	49	4 (8.2)	25 (51)^A	47 (95.9)^A
Soybean meal	31	3 (9.7)^a	14 (45.2)^a	31 (100)^a
Cottonseed meal	8	1 (12.5)^a	6 (75)^a	7 (87.5)^a
Sunflower meal	5	—	3 (60)^a	5 (100)^a
Linseed meal	3	—	1 (33.3)	3 (100)
Canola meal	2	—	1 (50)	1 (50)
Corn products	17	—	3 (17.6)^B	11 (64.7)^B
Corn gluten	10	—	—	5 (50)^a
Corn meal	5	—	2 (40)	5 (100)^a
Corn germ	1	—	—	1 (100)
Hominy	1	—	1 (100)	—
Total	201	46 (22.9)	79 (39.3)	174 (86.6)

Underlined are subtypes of plant-based feed ingredients within two main types: oilseed byproducts and corn products.

No sample tested positive for E. coli O157:H7.

In each column within animal- or plant-derived feed ingredients, percentages followed by different uppercase (A or B) letters are statistically significant (p<0.05). Among oilseed byproducts or corn products, in each column, percentages followed by the same lowercase italic letter (a) are not statistically significant (p>0.05). Statistical analysis was not performed for feed types having less than 10% of the total samples in each category or subcategory.

One sample consisted of 10 subsamples each weighing approximately 100 g. The preparation, handling, and shipping of samples were in accordance with FDA's standard procedures (FDA, 2012b). Upon arrival at the laboratory, the 10 subsamples representing one sample were aseptically combined into two composites for analysis. Of each composite, 25-g portions were used for bacterial isolation.

Bacterial isolation and identification

Media and reagents were obtained from BD Diagnostic Systems (Sparks, MD) unless specified otherwise. Salmonella, Campylobacter, and Escherichia coli were isolated following methods specified in FDA's Bacteriological Analytical Manual (FDA, 2001) with modifications, while the method of the National Antimicrobial Resistance Monitoring System (NARMS) (FDA, 2012c) was used for Enterococcus with modifications.

For Salmonella isolation, all incubations were conducted at 35°C for 24 h unless specified otherwise. Briefly, 25 g of feed ingredient was mixed thoroughly with 225 mL of lactose broth. After incubation, an aliquot (0.1 mL) was inoculated into 10 mL of Rappaport-Vassiliadis R10 broth and incubated at 42°C for 24 h. The enrichment broth was then streaked onto xylose lysine Tergitol 4 agar and incubated. Suspect Salmonella colonies were subcultured on Trypticase soy agar with 5% sheep blood. After biochemical confirmation using the Vitek system (bioMérieux, Hazelwood, MO), Salmonella isolates were stored at −80°C in brucella broth containing 20% glycerol. Serovars were determined by the White-Kauffmann-Le Minor scheme (Grimont and Weill, 2007) using commercial antisera.

For Campylobacter isolation, all incubations were conducted at 42°C for 24–48 h under microaerophilic conditions (5% O_2, 10% CO₂, and 85% N₂). Briefly, 25 g of feed ingredient was mixed well with 225 mL of Bolton broth (Oxoid Inc., Ogdensburg, NY) and incubated for 24 h. The enrichment broth was inoculated onto Campy-Cefex agar using a cotton-tipped swab followed by streaking. After 48-h incubation, presumptive Campylobacter colonies were transferred to blood agar and incubated for 24 h. Morphological examinations, Gram stain, and the AccuProbe Campylobacter Culture Identification Test (Gen-Probe, San Diego, CA) were performed following the manufacturers' instructions.

For Escherichia coli, all incubations were conducted at 35°C for 24 h unless specified otherwise. Briefly, 25 g of feed ingredient was mixed thoroughly with 225 mL of lauryl tryptose broth and incubated. The broth culture was inoculated onto tellurite cefixime–sorbitol MacConkey agar and incubated for E. coli O157:H7 isolation. Additionally, an aliquot (100 μL) was inoculated into 10 mL of EC broth and incubated at 45°C for 24–48 h. The EC broth showing gas formation was inoculated onto eosin methylene blue agar and incubated. Presumptive E. coli O157:H7 colonies on tellurite cefixime–sorbitol MacConkey agar and E. coli colonies on eosin methylene blue agar were subcultured on blood agar. E. coli isolates were confirmed by biochemical tests using the Vitek system. Suspect E. coli O157:H7 isolates were tested for Stx1 and Stx2 production using a Shiga toxin enzyme immunoassay (Premier EHEC; Meridian Bioscience, Inc., Cincinnati, OH).

For Enterococcus isolation, 25 g of feed ingredient was mixed thoroughly with 225 mL of enterococcosel broth for selective enrichment at 45°C for up to 48 h. If blackening was observed, the broth was streaked onto enterococcosel agar and incubated at 35°C for 24 h. Presumptive Enterococcus colonies were transferred to blood agar and incubated. Gram stain, catalase test, PYR test, and the AccuProbe Enterococcus Culture Identification Test (Gen-Probe) were performed following the manufacturers' instructions. Enterococcus species were determined using the Vitek system.

Pulsed-field gel electrophoresis (PFGE)

PFGE was performed in 2009 to assess Salmonella genetic relatedness using the the Centers for Disease Control and Prevention's PulseNet protocol (Ribot et al., 2006). Agarose-embedded DNA was digested with XbaI or BlnI (Roche, Indianapolis, IN) and restriction fragments were separated by pulsed electrophoresis (CHEF Mapper; Bio-Rad, Hercules, CA). PFGE patterns were analyzed using the BioNumerics software (version 5.1; Applied-Maths, Kortrijk, Belgium) with a 1.5% band position tolerance. PFGE patterns of leading Salmonella serovars identified in this study were searched against the Centers for Disease Control and Prevention's PulseNet database.

Antimicrobial susceptibility testing

The minimal inhibitory concentrations for Salmonella, E. coli, and Enterococcus isolates were determined by broth microdilution and interpreted according to guidelines of the Clinical and Laboratory Standards Institute (CLSI, formerly National Committee for Clinical Laboratory Standards, NCCLS) (NCCLS, 2002a; 2002b). Custom 96-well microtiter plates (Sensititre; TREK Diagnostic Systems, Inc., Cleveland, OH) developed by NARMS (FDA, 2012c) were used. Quality control organisms were E. coli ATCC 25922, Enterococcus faecalis ATCC 29212, Staphylococcus aureus ATCC 29213, and Pseudomonas aeruginosa ATCC 27853.

Statistical analysis

Prevalence data sorted by target organism and feed type were analyzed by using analysis of variance (SAS for Windows, version 9; Cary, NC). Resistance data were compared by the chi-square test. Differences between the means were significant when p<0.05.

Results

Prevalence of Salmonella, Campylobacter, E. coli, and Enterococcus

None of the 201 feed ingredients yielded Campylobacter or E. coli O157:H7, whereas Salmonella, generic E. coli, and Enterococcus were recovered from 22.9%, 39.3%, and 86.6% of samples, respectively (Table 1). Animal byproducts had a significantly higher Salmonella contamination rate (34.4%) than plant byproducts (5.1%) (p<0.05). Conversely, plant byproducts were more frequently contaminated with E. coli and Enterococcus (p>0.05).

Among animal byproducts, the prevalence of Salmonella ranged from 10% in feather meal to 80% in fish meal (Table 1). Meat and bone meal, which comprised the majority (59%) of animal byproducts sampled, harbored Salmonella at a rate of 38.9%. The prevalence of Salmonella, E. coli, or Enterococcus did not differ significantly among the three major animal byproducts sampled (n ≥10) (i.e., meat and bone meal, poultry meal, or blood meal) (p>0.05). It is noteworthy that the five fish-meal samples also had the highest prevalence rates for E. coli (80%) and Enterococcus (100%).

Among plant byproducts, only soybean meal (9.7%) and cottonseed meal (12.5%) tested positive for Salmonella (Table 1). The prevalence of E. coli and Enterococcus among alfalfa meal and oilseed byproducts was significantly higher than that in corn products (p<0.05).

Salmonella serovar distribution and PFGE patterns

Among 74 Salmonella isolates recovered, 70 were from animal byproducts and four were from plant byproducts. Excluding one that was untypable, a total of 27 Salmonella serovars were identified (Table 2). Tennessee was most prevalent, but it comprised less than 10% of the isolates (Table 2). Montevideo was associated with the highest number of feed types (n=4). Among animal byproducts, meat and bone meal contained the highest number of Salmonella serovars (n=21), followed by blood meal (n=6), poultry and fish meal (n=4 each), and feather and bone meal (n=2 each). In plant byproducts, soybean meal had three Salmonella serovars, Anatum, Godesberg, and Schwarzengrund, while Tennessee was the only one identified in cottonseed meal. Notably, a large percentage (22 of 46; 47.8%) of Salmonella-positive samples harbored two serovars (data not shown).

Table 2.

Salmonella Serovar Distribution Across Various Types (Subtypes) of Feed Ingredient Samples

	No. (%) of positive feed ingredient samples harboring specific Salmonella serovar
		Animal-derived proteins						Plant-derived proteins
Salmonella serovar ^a	All types (n=201)	Meat and bone meal (n=72)	Poultry meal (n=17)	Blood meal (n=16)	Feather meal (n=10)	Fish meal (n=5)	Bone meal (n=2)	Soybean meal (n=31)	Cottonseed meal (n=8)
Tennessee	7 (3.5)	6 (8.3)							1 (12.5)
Cerro	5 (2.5)	4 (5.6)	1 (5.9)
Montevideo	5 (2.5)	2 (2.8)		1 (6.3)	1 (10)	1 (20)
Oranienburg	5 (2.5)	4 (5.6)	1 (5.9)
Anatum var. 15+	4 (2)	1 (1.4)				2 (40)	1 (50)
Bareilly	4 (2)	4 (5.6)
Anatum	3 (1.5)	1 (1.4)		1 (6.3)				1 (3.2)
Lille var. 14+	3 (1.5)	3 (4.2)
Ohio	3 (1.5)	3 (4.2)
Senftenberg	3 (1.5)	2 (2.8)			1 (10)
Amager	2 (1)	2 (2.8)
Infantis	2 (1)	2 (2.8)
Johannesburg	2 (1)	1 (1.4)	1 (5.9)
Mbandaka	2 (1)	1 (1.4)					1 (50)
Meleagridis	2 (1)	2 (2.8)
Muenster	2 (1)	1 (1.4)		1 (6.3)
Schwarzengrund	2 (1)	1 (1.4)						1 (3.2)
Waycross	2 (1)		1 (5.9)	1 (6.3)
All serovars^b	46 (22.9)	28 (38.9)	3 (17.6)	5 (31.3)	1 (10)	4 (80)	1 (50)	3 (9.7)	1 (12.5)

Only Salmonella serovars identified in two or more feed ingredient samples are listed. Additional serovars isolated from single feed ingredient samples include Agona, Bietri, Kiambu, and Muenster var. 15⁺ from meat and bone meal samples, Bredeney and Orion from blood meal, Molade and Putten from fish meal, and Godesberg from soybean meal.

In this row, the total number of feed ingredient samples harboring any Salmonella serovar are tabulated. A large percentage (22 of 46; 47.8%) of positive samples had two Salmonella serovars.

Fifty-five distinct Salmonella PFGE patterns were identified. Regardless of the source of isolation, isolates of the same serovar generally clustered together, except for those belonging to Bareilly, Montevideo, Ohio, and Oranienburg (data not shown). In addition, isolates from all of the four leading Salmonella serovars (Tennessee, Cerro, Montevideo, and Oranienburg) had indistinguishable PFGE-XbaI patterns with those from human clinical cases deposited in PulseNet (data not shown).

Composition of E. coli and Enterococcus isolates

A total of 131 E. coli isolates were recovered, including 77 (58.8%) from animal byproducts and 54 (41.2%) from plant byproducts.

The vast majority (320 of 362, 88.4%) of Enterococcus isolates were Enterococcus faecium. Among isolates from animal byproducts (n=200), E. faecium comprised the highest percentage (86.5%) followed by E. faecalis (7.5%), E. gallinarum (2%), other Enterococcus spp. (2%), E. hirae (1.5%), and E. avium (0.5%). Among isolates from plant byproducts (n=162), E. faecium again was most prevalent (90.7%), followed by other Enterococcus spp. (6.8%) and E. faecalis (2.5%).

Antimicrobial susceptibility profiles

Salmonella isolates (n=74) were pansusceptible to 17 antimicrobial agents tested using the NARMS Gram-negative panel. E. coli isolates (n=131) demonstrated similar susceptibility to these antimicrobials except for tetracycline (15.3% resistance), sulfamethoxazole (7.6%), streptomycin (4.6%), ampicillin (3.8%), and nalidixic acid (1.5%) (Table 3). Additionally, E. coli from animal byproducts had higher resistance rates than those from plant byproducts except for nalidixic acid, although the difference was significant only for tetracycline (p<0.05).

Table 3.

Antimicrobial Resistance Profiles Among 131 Escherichia coli Isolates Recovered by Commodity Group

			No. (%) of resistant isolates ^a
Antimicrobial class	Antimicrobial agent	Resistant breakpoint (μg/mL)	All (n=131)	Animal-derived (n=77)	Plant-derived (n=54)
Aminoglycosides	Amikacin	≥64	—	—	—
	Gentamicin	≥16	—	—	—
	Kanamycin	≥64	—	—	—
	Streptomycin	≥64	6 (4.6)	5 (6.5)^A	1 (1.9)^A
β-Lactam/β-lactamase inhibitor combinations	Amoxicillin-clavulanic acid	≥32/16	—	—	—
Cephems	Cefoxitin	≥32	—	—	—
	Ceftiofur	≥8	—	—	—
	Ceftriaxone	≥64	—	—	—
	Cephalothin	≥32	—	—	—
Folate pathway inhibitors	Sulfamethoxazole	≥512	10 (7.6)	8 (10.4)^A	2 (3.7)^A
	Trimethoprim-sulfamethoxazole	≥4/76	—	—	—
Penicillins	Ampicillin	≥32	5 (3.8)	4 (5.2)^A	1 (1.9)^A
Phenicols	Chloramphenicol	≥32	—	—	—
Quinolones	Ciprofloxacin	≥4	—	—	—
	Nalidixic acid	≥32	2 (1.5)	1 (1.3)^A	1 (1.9)^A
Tetracyclines	Tetracycline	≥16	20 (15.3)	16 (20.8)^A	4 (7.4)^B

In each row, percentages followed by different uppercase letters (A or B) are significantly different (p<0.05) between isolates recovered from animal- or plant-derived ingredients.

Using the NARMS Gram-positive panel, Enterococcus isolates (n=362) were resistant to 5 out of 17 antimicrobials tested, ranging from 1.1% to penicillin to 14.6% to tetracycline (Table 4). As expected, all of the 19 E. faecalis isolates were intrinsically resistant to quinupristin/dalfopristin (Singh et al., 2002). Four E. gallinarum isolates from animal byproducts exhibited intermediate susceptibility to vancomycin (minimal inhibitory concentration=16 μg/mL). Again, isolates from animal byproducts carried higher resistance rates than those from plant byproducts, except for erythromycin. All of the Enterococcus isolates demonstrating resistance to ciprofloxacin, penicillin, and streptomycin were E. faecium.

Table 4.

Antimicrobial Resistance Profiles Among 362 Enterococcus Isolates Recovered by Commodity Group

			No. (%) of resistant isolates ^a
Antimicrobial class	Antimicrobial agent	Resistant breakpoint (μg/mL)	All (n=362)	Animal-derived (n=200)	Plant-derived (n=162)
Aminoglycosides	Gentamicin	≥500	—	—	—
	Kanamycin	≥512	—	—	—
	Streptomycin^b	≥1000	7 (1.9)	5 (2.5)^A	2 (1.2)^A
Glycopeptides	Vancomycin^c	≥32	—	—	—
Lincosamides	Lincomycin	≥32	—	—	—
Macrolides	Erythromycin	≥8	24 (6.6)	9 (4.5)^A	15 (9.3)^A
	Tylosin	≥32	—	—	—
Nitrofurans	Nitrofurantoin	≥128	—	—	—
Oxazolidinones	Linezolid	≥8	—	—	—
Penicillins	Penicillin^b	≥16	4 (1.1)	4 (2)	—
Phenicols	Chloramphenicol	≥32	—	—	—
Phosphoglycolipids	Flavomycin	≥32	—	—	—
Polyether ionophore	Salinomycin	≥32	—	—	—
Polypeptides	Bacitracin	≥128	—	—	—
Quinolones	Ciprofloxacin^b	≥4	26 (7.2)	16 (8)^A	10 (6.2)^A
Streptogramins	Quinupristin/dalfopristin^d	≥4	19 (5.2)	15 (7.5)^A	4 (2.5)^B
Tetracyclines	Tetracycline	≥16	53 (14.6)	37 (18.5)^A	16 (9.9)^B

In each row, percentages followed by different upper case letters (A or B) are significantly different (p<0.05) between isolates recovered from animal- or plant-derived ingredients.

All isolates resistant to ciprofloxacin, penicillin, and streptomycin were Enterococcus faecium.

Four Enterococcus gallinarum isolates from animal-derived products exhibited intermediate susceptibility to vancomycin (minimum inhibitory concentration=16 μg/mL).

All isolates resistant to quinupristin/dalfopristin were Enterococcus faecalis, which is intrinsic resistant to this antimicrobial.

Discussion

This survey was conducted in 2002 and 2003, 10 years after the FDA/CVM's 1993 survey that established the baseline for Salmonella in animal feed ingredients (McChesney et al., 1995). Compared to that survey, the overall prevalence of Salmonella decreased from 49.7% to 22.9%. The decrease was observed among both animal (from 56.4% to 34.4%) and plant byproducts (from 36% to 5.1%). Specifically, the rate among meat and bone meal decreased from 64.3% to 38.9%, while the reduction was from 24.2% to 9.7% in soybean meal. Since 2002, the FDA/CVM has established a Salmonella surveillance program to monitor the trend of Salmonella contamination in animal feed (FDA, 2012a). A recent report from that program showed a significant overall Salmonella reduction in feed ingredients (from 30.9% in 2002–2006 to 19.4% in 2007–2009; p<0.05) (Li et al., 2012). This reduction was primarily due to the decrease in animal-derived ingredients (from 66.1% in 2002–2006 to 41.3% in 2007–2009; p<0.05). Notable differences between samples collected in the three surveys were that samples in the surveillance program have entered interstate commerce while those in the two ingredient surveys were directly from processing plants, and some samples in the surveillance program were imported.

Despite the decrease, the present survey supports that Salmonella is a major hazard in feed ingredients (The Panel on Biological Hazards, 2008). The overall prevalence (22.9%) fell within the broad range (0.8%–69.5%) reported since the 1950s (Rutqvist and Thal, 1958; Bensink, 1979; Veldman et al., 1995; Davies and Wray, 1997; Krytenburg et al., 1998; Kidd et al., 2002; Davis et al., 2003; Whyte et al., 2003; Franco, 2005; Papadopoulou et al., 2009; Kinley et al., 2010; Wierup and Haggblom, 2010; Torres et al., 2011; Shilangale et al., 2012). The significantly higher occurrence of Salmonella among animal over plant byproducts also corroborated two FDA/CVM surveys (McChesney et al., 1995; Li et al., 2012). Rendered products, such as meat and bone meal, meat meal, and poultry meal frequently harbor Salmonella (Bensink, 1979; Williams, 1981; Franco, 2005; Kinley et al., 2010; Shilangale et al., 2012). Other studies (Veldman et al., 1995; Papadopoulou et al., 2009) reported fish meal to be especially problematic. Among plant byproducts, cotton-seed meal had the highest odds of contamination in Spanish feed mills (Torres et al., 2011), while 14.6% of soybean meal imported to Sweden contained Salmonella (Wierup and Haggblom, 2010). In a large Great Britain survey between 1987 and 2006, soybean meal, rapeseed meal, and fish meal were the top three feed types with Salmonella contaminations (Papadopoulou et al., 2009). Not surprisingly, we found relatively high Salmonella incidence among certain animal byproducts (fish meal, meat and bone meal, and blood meal) and in soybean and cottonseed meal.

Diverse and exotic Salmonella serovars (n=27) with extensive PFGE patterns were recovered. To our knowledge, this is the first report showing a high percentage (47.8%) of positive samples with two Salmonella serovars. Previously, animals seropositive to one or more Salmonella serogroups were observed in most cattle operations, likely due to the consumption of feed contaminated with multiple Salmonella serovars (Smith et al., 1994). Among Salmonella serovars identified, Agona, Anatum, Cerro, Mbandaka, Montevideo, Ohio, Schwarzengrund, Senftenberg, and Tennessee were frequently reported in previous feed surveys (McChesney et al., 1995; Davis et al., 2003; Franco, 2005; Papadopoulou et al., 2009; Kinley et al., 2010; Wierup and Haggblom, 2010; Torres et al., 2011; Li et al., 2012). Agona, Anatum, Montevideo, and Senftenberg were also common serovars submitted to animal diagnostic laboratories in 2002 and 2003 (Ferris et al., 2002; Ferris et al., 2003). However, Salmonella serovars concerning target animal health were not isolated, such as Newport or Dublin in cattle, biovars Gallinarum or Pullorum in poultry, and Choleraesuis in swine (FDA, 2010). From a public health perspective, Enteritidis, Newport, and Typhimurium consistently ranked as the top three Salmonella serovars causing human foodborne illnesses in the United States (CDC, 2012). Although not found in this survey, all of these three serovars have been recovered previously from feed ingredients (McChesney et al., 1995; Krytenburg et al., 1998; Davies and Wray, 1997; Kidd et al., 2002; Davis et al., 2003; Dargatz et al., 2005; Franco, 2005; Papadopoulou et al., 2009; Wierup and Haggblom, 2010; Torres et al., 2011). Importantly, other leading Salmonella serovars causing human foodborne illnesses (CDC, 2012) (i.e., Infantis and Montevideo), were identified in this survey. Furthermore, isolates from all of the leading Salmonella serovars possessed similar PFGE patterns as human isolates in PulseNet, suggesting their potentials to cause human illness.

In direct contrast to Salmonella, the failure to recover Campylobacter or E. coli O157:H7 confirms that feed ingredient is a far less important source for these foodborne pathogens (The Panel on Biological Hazards, 2008). Previously, E. coli O157 has been detected in cattle feed (Davis et al., 2003), but not Campylobacter (Whyte et al., 2003), although the number of studies examining these pathogens in feed is rather scarce.

E. coli and Enterococcus were present in 39.3% and 86.6% of samples, respectively. As an indicator organism, E. coli has been found in feed ingredients with rates ranging from 0 to 48.2% (Lynn et al., 1998; Dargatz et al., 2005; da Costa et al., 2007; Kinley et al., 2010). In comparison, previous surveys of retail meats demonstrated much higher E. coli contamination rates (Zhao et al., 2012). Conversely, the high prevalence of Enterococcus is well aligned with previous surveys on feed ingredients and retail meats (Hayes et al., 2003; da Costa et al., 2007; Sapkota et al., 2011; Aslam et al., 2012). One notable difference is that among retail meats, E. faecalis was more dominant than in animal feed ingredients, where E. faecium was the vast majority (Hayes et al., 2003; Aslam et al., 2012).

The potential development and dissemination of antimicrobial resistance via animal feed is of important public health concern (The Panel on Biological Hazards, 2008). Previously, Salmonella isolates resistant to several antimicrobials have been recovered from feed ingredients, although at a low rate (Hofacre et al., 2001; Dargatz et al., 2005). In contrast, feed ingredients were extensively contaminated with resistant enterococci (da Costa et al., 2007), and to a lesser extent E. coli (Dargatz et al., 2005; da Costa et al., 2007). The numbers of antimicrobials resistant to and resistance rates found in this survey were low for all three bacterial genera, in comparison with those reported in isolates from humans, animals, and retail meats (CDC/FDA/USDA, 2013). Nonetheless, the current scattered and sparse data on antimicrobial resistance among bacteria in feed are inadequate and warrant further large-scale studies (The Panel on Biological Hazards, 2008).

Conclusions

This survey demonstrated that diverse populations of Salmonella, E. coli, and Enterococcus are commonly present in animal feed ingredients but antimicrobial resistance is uncommon. Because feed ingredients are at the very beginning of the farm-to-table continuum, these findings highlight potential food safety hazards and warrant further large-scale studies to continuously monitor these pathogenic and indicator organisms in feed.

Footnotes

Acknowledgments

We thank FDA field investigators at the Office of Regulatory Affairs for their assistance in obtaining feed ingredient samples used in this study.

Disclosure Statement

No competing financial interests exist.

References

Aslam

, Diarra

, Checkley

, Bohaychuk

, Masson

. Characterization of antimicrobial resistance and virulence genes in Enterococcus spp. isolated from retail meats in Alberta, Canada. Int J Food Microbiol, 2012; 156:222–230.

Bensink

. Salmonella contamination of meat and bone meal. Aust Vet J, 1979; 55:13–5.

Best

. World feed panorama: Once again, industry increases its volume. Feed Int, 2012; 33:8–12.

[CDC] Centers for Disease Control and Prevention. FoodNet 2011 preliminary data. 2012. http://www.cdc.gov/foodnet/data/trends/tables-2011.html. 2012 August 1.

Crump

, Griffin

, Angulo

. Bacterial contamination of animal feed and its relationship to human foodborne illness. Clin Infect Dis, 2002; 35:859–865.

da Costa

, Oliveira

, Bica

, Vaz-Pires

, Bernardo

. Antimicrobial resistance in Enterococcus spp. and Escherichia coli isolated from poultry feed and feed ingredients. Vet Microbiol, 2007; 120:122–131.

Dargatz

, Strohmeyer

, Morley

, Hyatt

, Salman

. Characterization of Escherichia coli and Salmonella enterica from cattle feed ingredients. Foodborne Pathog Dis, 2005; 2:341–347.

Davies

, Wray

. Distribution of Salmonella contamination in ten animal feedmills. Vet Microbiol, 1997; 57:159–169.

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.

10.

[FDA] Food and Drug Administration. Bacteriological Analytical Manual online. 2001. http://www.fda.gov/Food/ScienceResearch/LaboratoryMethods/BacteriologicalAnalyticalManualBAM/. 2012 May 25.

11.

[FDA]. Draft Compliance Policy Guide Sec. 690.800. Salmonella in Animal Feed, 2010. http://www.regulations.gov/#!documentDetail;D=FDA-2010-D-0378-0002. 2012 February 14.

12.

[FDA]. Investigations Operations Manual. CVM Feed Sampling Survey Assignments, 2012a. http://www.fda.gov/AnimalVeterinary/Products/AnimalFoodFeeds/Contaminants/default.htm. 2012 October 11.

13.

[FDA]. Investigations Operations Manual. 2012b. http://www.fda.gov/ICECI/Inspections/IOM/default.htm. 2012 May 25.

14.

[FDA]. NARMS Methods and Data. 2012c. http://www.fda.gov/AnimalVeterinary/SafetyHealth/AntimicrobialResistance/NationalAntimicrobialResistanceMonitoringSystem/ucm059103.htm. 2012 October 15.

15.

[FDA]. NARMS Retail Meat Annual Report. 2011; 2013. http://www.fda.gov/AnimalVeterinary/SafetyHealth/AntimicrobialResistance/NationalAntimicrobialResistanceMonitoringSystem/ucm334828.htm. 2013 February 5.

16.

Ferris

, Aalsburg

, Iseminger

. Salmonella serotypes from animals and related sources reported during July 2001–June 2002. Proc US Anim Health Assoc, 2002; 106:467–497.

17.

Ferris

, Aalsburg

, Palmer

, Hostetler

. Salmonella serotypes from animals and related sources reported during July 2002–June 2003. Proc US Anim Health Assoc, 2003; 107:463–469.

18.

Franco

. A survey of Salmonella serovars and most probable numbers in rendered-animal-protein meals: Inferences for animal and human health. J Environ Health, 2005; 67:18–22.

19.

Grimont

PAD

, Weill

. Antigenic Formulae of the Salmonella Serovars. 9 ^th World Health Organization Collaborating Center for Reference and Research. Salmonella . Paris: Institut Pasteur, 2007.

20.

Hayes

, English

, Carter

, Proescholdt

, Lee

, Wagner

, White

. Prevalence and antimicrobial resistance of Enterococcus species isolated from retail meats. Appl Environ Microbiol, 2003; 69:7153–7160.

21.

Hofacre

, White

, Maurer

, Morales

, Lobsinger

, Hudson

. Characterization of antibiotic-resistant bacteria in rendered animal products. Avian Dis, 2001; 45:953–961.

22.

Kidd

, Rossignol

, Gamroth

. Salmonella and other Enterobacteriaceae in dairy-cow feed ingredients: Antimicrobial resistance in western Oregon. J Environ Health, 2002; 64:9–1632quiz 35–36.

23.

Kinley

, Rieck

, Dawson

, Jiang

. Analysis of Salmonella and enterococci isolated from rendered animal products. Can J Microbiol, 2010; 56:65–73.

24.

Krytenburg

, Hancock

, Rice

, Besser

, Gay

. A pilot survey of Salmonella eneterica contamination of cattle feeds in the Pacific northwestern USA. Anim Feed Sci Technol, 1998; 75:75–79.

25.

, Bethune

, Jia

, Lovell

, Proescholdt

, Benz

, Schell

, Kaplan

, McChesney

. Surveillance of Salmonella prevalence in animal feeds and characterization of the Salmonella isolates by serotyping and antimicrobial susceptibility. Foodborne Pathog Dis, 2012; 9:692–698.

26.

Lynn

, Hancock

, Besser

, Harrison

, Rice

, Stewart

, Rowan

. The occurrence and replication of Escherichia coli in cattle feeds. J Dairy Sci, 1998; 81:1102–1108.

27.

McChesney

, Kaplan

, Gardner

. FDA survey determines Salmonella contamination. Feedstuffs, 1995; 67:20–23.

28.

Muirhead

. Structure of U.S. feed industry has evolved to meet changing needs of producers, consumers. Feedstuffs, 2003; 75:14–17.

29.

Muirhead

. Feed marketing & distribution. Feedstuffs, 2012; 84:3–14.

30.

[NCCLS] National Committee for Clinical Laboratory Standards. Performance Standards for Antimicrobial Disk and Dilution Susceptibility Tests for Bacteria Isolated from Animals; Approved Standard-Second Edition (M31-A2) Wayne, PA: NCCLS, 2002a.

31.

[NCCLS] Performance Standards for Antimicrobial Susceptibility Testing; Twelveth Informational Supplement (M100-S12). Wayne, PA: NCCLS, 2002b.

32.

Papadopoulou

, Carrique-Mas

, Davies

, Sayers

. Retrospective analysis of Salmonella isolates recovered from animal feed in Great Britain. Vet Rec, 2009; 165:681–688.

33.

Ribot

, Fair

, Gautom

, Cameron

, Hunter

, Swaminathan

, Barrett

. Standardization of pulsed-field gel electrophoresis protocols for the subtyping of Escherichia coli O157:H7, Salmonella, and Shigella for PulseNet. Foodborne Pathog Dis, 2006; 3:59–67.

34.

Rutqvist

, Thal

. Salmonella isolated from animals and animal products in Sweden during 1956–1957. Nord Vet Med, 1958; 10:234–244.

35.

Sapkota

, Hulet

, Zhang

, McDermott

, Kinney

, Schwab

, Joseph

. Lower prevalence of antibiotic-resistant enterococci on U.S. conventional poultry farms that transitioned to organic practices. Environ Health Perspect, 2011; 119:1622–1628.

36.

Sapkota

, Lefferts

, McKenzie

, Walker

. What do we feed to food-production animals? A review of animal feed ingredients and their potential impacts on human health. Environ Health Perspect, 2007; 115:663–670.

37.

Shilangale

, Di Giannatale

, Chimwamurombe

, Kaaya

. Prevalence and antimicrobial resistance pattern of Salmonella in animal feed produced in Namibia. Vet Ital, 2012; 48:125–132.

38.

Singh

, Weinstock

, Murray

. An Enterococcus faecalis ABC homologue (Lsa) is required for the resistance of this species to clindamycin and quinupristin-dalfopristin. Antimicrob Agents Chemother, 2002; 46:1845–1850.

39.

Smith

, Da Roden

, Thurmond

, Dilling

, Konrad

, Pelton

, Picanso

. Prevalence of salmonellae in cattle and in the environment on California dairies. J Am Vet Med Assoc, 1994; 205:467–471.

40.

The Panel on Biological Hazards. European Commission on Microbiological Risk Assessment in feedingstuffs for food-producing animals, Scientific Opinion of the Panel on Biological Hazards on a request from the Health and Consumer Protection Directorate General. EFSA J, 2008; 720:1–84.

41.

Torres

, Piquer

, Algarra

, de Frutos

, Sobrino

. The prevalence of Salmonella enterica in Spanish feed mills and potential feed-related risk factors for contamination. Prev Vet Med, 2011; 98:81–87.

42.

Veldman

, Vahl

, Borggreve

, Fuller

. A survey of the incidence of Salmonella species and Enterobacteriaceae in poultry feeds and feed components. Vet Rec, 1995; 136:169–172.

43.

Whyte

, McGill

, Collins

. A survey of the prevalence of Salmonella and other enteric pathogens in a commercial poultry feed mill. J Food Safety, 2003; 23:13–24.

44.

Wierup

, Haggblom

. An assessment of soybeans and other vegetable proteins as source of Salmonella contamination in pig production. Acta Vet Scand, 2010; 52:15.

45.

Williams

. Salmonellas in poultry feeds—A worldwide review. World Poultry Sci J, 1981; 37:6–25.

46.

Zhao

, Blickenstaff

, Bodeis-Jones

, Gaines

, Tong

, McDermott

. Comparison of the prevalences and antimicrobial resistances of Escherichia coli isolates from different retail meats in the United States, 2002 to 2008. Appl Environ Microbiol, 2012; 78:1701–1707.