PROCESSING, PRODUCTS, AND FOOD SAFETY Salmonella Populations and Prevalence in Layer Feces from Commercial High-Rise Houses and Characterization of the Salmonella Isolates by Serotyping, Antibiotic Resistance Analysis, and Pulsed Field Gel Electrophoresis X. Li,* J. B. Payne,† F. B. Santos,† J. F. Levine,* K. E. Anderson,† and B. W. Sheldon†1 *Department of Population Health and Pathobiology, and †Department of Poultry Science, North Carolina State University, Raleigh 27695 wk (second peak of production cycle)]. Bird ages and molting practice did not significantly affect (P > 0.05) Salmonella populations with an average of 1.25, 1.27, 1.20, and 1.14 log most probable number/g for the 18-, 25- to 28-, 66- to 74-, and 75- to 7-wk birds, respectively. However, the 18-wk birds had the highest prevalence of Salmonella (55.6%), followed by the 25- to 28-wk birds (41.7%), 75- to 78-wk birds (16.7%), and 66- to 74-wk birds (5.5%). Of the 45 Salmonella isolates characterized, the most predominant serovar was Salmonella Kentucky (62%). Thirtyfive percent of the Salmonella isolates were resistant to at least 1 antibiotic. As expected, considerable genetic diversity was observed within and across the different serovars.
ABSTRACT Salmonella species are recognized as a major cause of foodborne illnesses that are closely associated with the consumption of contaminated poultry and egg products. The objectives of this study were to evaluate the Salmonella populations and prevalence in layer feces during the laying cycle and molting of the hen and to characterize the layer fecal Salmonella isolates by serotyping, antibiotic resistance analysis, and pulsed field gel electrophoresis. Fecal samples were collected from a commercial layer complex consisting of 12 houses. Composite fecal samples across each row were collected as a function of bird age [18 wk (at placement), 25 to 28 wk (first peak of production cycle), 66 to 74 wk (molting), and 75 to 78
Key words: Salmonella, layer feces, population, prevalence, characterization 2007 Poultry Science 86:591–597
confinement in large populations. High-rise poultry housing consists of an arrangement of cages for laying hens in which manure generated in each cage is dropped through an open space beneath the cages and stored below on the lower poultry house floor. Typically, a 10,000bird commercial layer flock can generate more than 100 tons of manure per year on a wet-weight basis (Patterson and Lorenz, 1996; Lorimor and Xin, 1999). In general, high-rise houses only require manure removal once a year during the molting period or when flocks are changed. After removal, manure is generally used as fertilizer and applied to fields for growing agricultural commodities (US Poultry and Egg Association, 1998). The large volume of manure produced in concentrated poultry production areas can contribute to the contamination of ground and surface waters (Mallin and Cahoon, 2003). Most environmental concerns over land application of animal manure have focused on either the effect of applied nutrients, especially N and P; on surrounding water quality; or have emphasized odor problems and air quality issues. However, the presence of human pathogens such as Salmonella in agricultural soils amended with manure may also pose a public health risk. Salmonella spp. have frequently been found in broiler (Payne et al., 2005)
INTRODUCTION The Centers for Disease Control and Prevention (CDC) estimate that approximately 40,000 cases of salmonellosis are diagnosed annually in the United States, resulting in approximately 600 deaths (Mead et al., 1999). Although Salmonella are generally inactivated during standard cooking practices and consumer education efforts advocate careful handling and preparation of meat products, contaminated eggs and animal meat products continue to contribute to the number of human salmonellosis cases. Raw meat and poultry are recognized as the primary sources for transmitting Salmonella species to humans, with 40% of the clinical cases attributed to the consumption of egg and poultry products (Sanchez et al., 2002). Modern livestock industries, including poultry production, are frequently unprofitable unless a significant economy of scale can be achieved (Bossman, 2005). To achieve this economy of scale, poultry are generally reared under
©2007 Poultry Science Association Inc. Received August 19, 2006. Accepted December 2, 2006. 1 Corresponding author:
[email protected]
591
592
LI ET AL.
and turkey (Santos et al., 2005) feces. For laying hens, factors such as the egg production cycle may affect Salmonella shedding; however, their specific role in the dynamics of Salmonella ecology in commercial egg production facilities is not clearly defined. These factors may influence the prevalence and populations of Salmonella in layer feces. Accordingly, we measured the prevalence and populations of Salmonella in feces collected from a commercial laying hen complex during different phases of the production cycle: at pullet replacement, peak of first egg production cycle, molting, and peak of the second egg production cycle. Salmonella isolates obtained were also characterized by serotyping, antibiotic resistance analysis, and pulsed field gel electrophoresis (PFGE). Although the focus of this study was not specifically directed at quantifying the actual safety risks associated with land application of layer wastes, it does provide an initial assessment of populations, serotypes, and antibiotic resistance and PFGE profiles of Salmonella species found in layer wastes, which would benefit future risk assessment studies directed at these waste management practices.
MATERIALS AND METHODS Study Site and Sample Collection This study was conducted in a commercial layer facility containing 12 high-rise houses. Each house measured 57 × 480 ft (17.4 × 146.3 m) and housed approximately 77,000 egg-laying White Leghorn hens in cages arranged across 6 rows. Temperatures in the layer houses were maintained from 20 to 28°C. Pullets were placed at 18 wk of age. Each flock replacement per house took 2 to 3 d. All flocks were molted from 66 to 74 wk, resulting in 2 egg production cycles. Birds were fed a low-protein energy basal diet during the first 2 wk of the molting period, a lowenergy Ca diet during the third through sixth weeks, and a postmolt diet during the seventh and eighth weeks. Approximately 300 g of fresh feces was collected under the cages across the entire length of each row, mixed, and stored in a sterile Whirl-Pak bag (Fisher Scientific Int., Bohemia, NY). The samples were kept in coolers containing “blue ice” and analyzed for Salmonella on the same day of sample collection.
Salmonella Enumeration and Detection A most probable number (MPN) procedure was used to estimate Salmonella populations in layer feces. Upon arrival at the laboratory, the fresh fecal samples were thoroughly mixed by hand while contained in the WhirlPak bags. Twenty-five grams of each composite sample was placed into a sterile stomacher filter bag containing 50 mL of buffered peptone water (BPW; Oxoid, Ogdensburg, NY). Individual bags were stomached for 1 min. For preenrichment, the mixtures were serially diluted in BPW, incubated at 37°C for 18 to 24 h (Morinigo et al., 1986; Tate and Miller, 1990; Tate et al., 1992), and then 0.1 mL of the appropriate dilutions from each tube were
transferred to 3 tubes containing 10 mL of RappaportVassiliadis (RV) broth (Oxoid). All RV tubes were incubated at 42°C for 18 to 24 h for selective enrichment of Salmonella spp. Following incubation, 1 loopful (approximately 10 L) from each RV tube was streaked onto modified Lys-Fe agar (Oxoid), selective medium for Salmonella, and incubated at 37°C for 18 to 24 h. Suspect black colonies on modified Lys-Fe agar plates were picked and confirmed for Salmonella by inoculation onto triple-sugar Fe slants (Oxoid) and agglutination using Salmonella polyO antiserum (Difco Laboratories, Detroit, MI). Populations of Salmonella spp. were calculated using Thomas’ approximation of MPN/g = P/(NT)1/2, where P = the number of positive tubes; N = the total quantity of sample (g) in all negative tubes; and T = the total quantity of sample (g) in all tubes (Swanson et al., 2001). The minimum detection limit of this method was 10 organisms/ g of excreta (1 log). The prevalence of Salmonella in the excreta was also tested as follows. Each 25-g sample was placed into a sterile stomacher filter bag containing 100 mL of BPW. The bags were stomached for 1 min, and then an additional 125 mL of BPW was added, mixed thoroughly, and incubated at 37°C for 18 to 24 h. One milliliter from each bag was added to a bottle containing 100 mL of RV broth and incubated at 42°C for 18 to 24 h. The remaining procedures for isolating and identifying Salmonella were as described above.
Serotyping Forty-five Salmonella isolates (approximately 50% of total isolates) collected from laying hen feces were selected for serotyping (18, 17, 7, and 3 isolates taken from the 18-, 25- to 28-, 66- to 74-, and 75- to 76-wk-old layer samples, respectively). For serotyping, the Salmonella isolates were transferred onto tryptic soy agar (Difco Laboratories) slants, grown overnight at 37°C, and shipped by overnight courier to the USDA National Veterinary Service Laboratories in Ames, Iowa.
Antibiotic Resistance Analysis The antimicrobial susceptibility of the 45 Salmonella isolates submitted for serotyping was determined for 15 antimicrobials listed in Table 1 using the disk diffusion method described by the National Committee for Clinical Laboratory Standards (2000). Antimicrobials selected for these assays reflect the recommendations of the National Committee for Clinical Laboratory Standards. The susceptibility tests were conducted using the Sensititre susceptibility system (Sensititre, Trek Diagnostic Systems Inc., Cleveland, OH). The system encompasses a microtiter plate that is dosed with 15 individual antimicrobial agents at specified concentrations. A standard protocol and resistance break point developed under the National Antimicrobial Resistance Monitoring System was followed to study the susceptibility of the Salmonella isolates. Each Salmonella isolate was cultured on brain heart infu-
593
SALMONELLA POPULATION AND PREVALENCE IN LAYER FECES Table 1. Antibiotics, their concentrations in the microtiter plate, and minimum inhibitory concentration (MIC) break point
Antibiotic (abbreviation) Amikacin (Ami) Ampicillin (Amp) Amoxicillin-clavulanic acid (Amo) Ceftriaxone (Ceftri) Chloramphenicol (Chlor) Ciprofloxacin (Cip) Trimethoprim-Sulfamethoxazole (Sulfa) Cefoxitin (Cefox) Gentamicin (Gen) Kanamycin (Kan) Nalidixic acid (Nal) Sulfisoxazole (Sulfi) Tetracycline (Tet) Streptomycin (Strep) Ceftiofur (Ceftio)
MIC break point (g/mL)
Concentration (g/mL)
Resistant
Sensitive
0.5 to 64 1 to 32 1/0.5 to 32/16 0.25 to 64 2 to 32 0.015 to 4 0.12/2.38 to 4/76 0.5 to 32 0.25 to 16 8 to 64 0.5 to 32 16 to 256 4 to 32 32 to 64 0.12 to 8
32 32 32/16 64 32 4 4/76 32 8 25 32 350 16 64 8
4 8 8/4 8 8 1 1/19 8 4 16 8 100 4 32 4
sion agar plates (Oxoid) at 37°C for 18 to 24 h. A nephelometer (Promega, Madison, WI) was calibrated using a 0.5 McFarland BaSO4 turbidity standard (Sensititre, Trek Diagnostic Systems Inc.). One to 2 colonies from the brain heart infusion agar plate were transferred to 5 mL of sterile saline (0.9% NaCl) and adjusted to a 0.5 McFarland reading using the nephelometer. Seventy-five microliters of the saline cell suspension was then transferred to 10 mL of Mueller Hinton broth (Oxoid). After transfer of 50 L of each Mueller Hinton broth cell suspension into separate wells of a microtiter plate containing 15 antimicrobials as listed in Table 1, the plate was sealed with an adhesive seal and incubated at 37°C for 18 to 24 h. The contents of the wells were manually read for bacterial growth under a fluorescent lamp. Salmonella Typhimurium DT104 (ATCC 700408) was used as the quality control strain for this assay.
PFGE The CDC PulseNet 1-d standardized laboratory protocol for molecular subtyping of Escherichia coli O157H7, nontyphoidal Salmonella serotypes, and Shigella sonnei by PFGE was used for extraction of Salmonella genomic DNA and for establishing the analysis conditions (CDC, 2004). After electrophoresis, the gels were stained for 20 min in 0.01% ethidium bromide solution and then destained 3 times for 30 min each in distilled water. Gel images were digitally captured using an Alphaimager (model 3300, Alpha Innotech Corp., San Leandro, CA). Gels were visually analyzed with PFGE profiles differing in 1 or more bands considered different.
Statistical Analysis The composite fresh fecal sample collected under each row of cages served as an independent trial replicate. There were 18, 24, 18, and 18 replicates evaluated for the 18-wk (pullet placement), 25- to 28-wk (peak of first production cycle), 66- to 74-wk (molting), and 75- to 76-
wk-old (peak of second production cycle) birds, respectively. The independent variable was bird ages. The Salmonella population data were analyzed by the GLM procedure for ANOVA. The residual replicate × age MS was used for testing the main effects (age, replicates; SAS Institute, 2000). When a significant effect was observed, means were compared using the PDIFF option of SAS. Moreover, the χ2 test was used to determine if Salmonella prevalence differences were significant among the 4 ages of birds. Model and parameter adequacy were considered significant at P ≤ 0.05 unless otherwise noted.
RESULTS AND DISCUSSION Salmonella Populations and Prevalence Salmonella populations and prevalence for the 18, 25to 28-, 66- to 74-, and 75- to 76-wk-old birds are shown in Table 2. No significant difference (P > 0.05) in Salmonella populations was detected among the 4 age groups (overall mean of log 1.21 MPN/g). This finding may be due in part to the large SD and relatively low population of Salmonella recovered from the feces (mean of
