Comparison of the prevalence of Salmonella infection in layer hens from commercial layer farms with high and low rodent densities

A comparison on the prevalence of Salmonella infection in layer hens from commercial layer farms with high and low rodent densities was investigated. Out of 280 laying hens sampled from three commercial layer farms with high rodent densities, Salmonella enterica subsp. enterica serovar Enteritidis (Salmonella Enteritidis) was isolated from 20 (7.14%) hens and Salmonella enterica subsp. enterica serovar Infantis (Salmonella Infantis) from three (1.07%) hens. In contrast, layer hens sampled from four commercial layer farms with low rodent densities were negative for any salmonellae. Significant differences (P < 0.05) in the isolation rates of Salmonella from various organs of infected layer hens were also noted. For Salmonella Enteritidis, liver (55.0%) and the oviduct (55.0%) had the highest isolation rates while all Salmonella Infantis isolates were from the oviduct. Pulsed field gel electrophoresis (PFGE) analysis of BlnI-digested chromosomal DNA of Salmonella Enteritidis isolated from layer hens and rodents showed similar patterns. PFGE analysis of Salmonella Infantis isolated from layer hens, rodents, eggs, and the environment yielded identical patterns. In this study, the significantly higher prevalence rate (P < 0.05) of Salmonella Enteritidis and Salmonella Infantis in layer hens from high rodent density farms could be attributed to the high rodent population density. The persistent Salmonella Enteritidis and Salmonella Infantis infection inside layer houses may have been amplified by the increasing numbers in the rodent population over the years, which increased the opportunity for environment-rodent-chicken interaction and the transmission of salmonellae to chickens. Monitoring of salmonellae from rodents inside poultry premises is recommended to be an effective additional tool in the assessment of the Salmonella status of layer flocks.     

The relationship between the duration of fecal shedding and the production of contaminated eggs by laying hens infected with strains of Salmonella enteritidis and Salmonella Heidelberg

Egg contamination by Salmonella Enteritidis has remained a significant public health problem for nearly two decades, and Salmonella Heidelberg has also been recently implicated in egg-transmitted human illness. Colonization of the intestinal tract is a necessary precursor to the invasion of reproductive organs and subsequent deposition inside eggs laid by infected hens, but the relationship between the persistence of Salmonella in the intestinal tract and the likelihood of egg contamination has been uncertain. In this study, groups of laying hens were inoculated with large oral doses of strains of Salmonella Enteritidis and Salmonella Heidelberg, including variants of the original parent strains that had been reisolated from eggs laid by infected hens in a prior study. The shedding of Salmonella in voided feces was monitored for 6 wk postinoculation, and all eggs laid by infected hens between 5 and 22 days postinoculation were cultured for Salmonella in their contents. The mean duration of fecal shedding was significantly longer for the previously passaged Salmonella strains (26.7 days) than for the original parent strains (17.5 days), and the passaged strains caused a significantly higher frequency of egg contamination (6.4%) than did the parent strains (3.3%). However, the duration of fecal shedding and the frequency of egg contamination were not correlated for any of the Salmonella Enteritidis or Salmonella Heidelberg strains.     

Early immune dynamics following infection with Salmonella enterica serovars Enteritidis, Infantis, Pullorum and Gallinarum: cytokine and chemokine gene expression profile and cellular changes of chicken cecal tonsils

Salmonella enterica subspecies enterica infection remains a serious problem in a wide range of animals and in man. Poultry-derived food is the main source of human infection with the non-host-adapted serovars while fowl typhoid and pullorum disease are important diseases of poultry. We have assessed cecal colonization and immune responses of newly hatched and older chickens to Salmonella serotypes Enteritidis, Infantis, Gallinarum and Pullorum. S. Enteritidis and S. Infantis colonized the ceca more efficiently than S. Gallinarum and S. Pullorum. Salmonella infection was also associated with increased staining for B-lymphocytes and macrophages in the cecal tonsils of infected birds. S. Enteritidis infection in newly hatched birds stimulated the expression of CXCLi1 and CXCLi2 chemokines in the cecal tonsils, while S. Gallinarum up-regulated the expression of LITAF. In older chickens, S. Enteritidis infection resulted in a significantly higher expression of CXCLi2, iNOS, LITAF and IL-10 while S. Pullorum appeared to down-regulate CXCLi1 expression in the cecal tonsils. Data from spleens showed either no expression or down-regulation of the tested genes.     

Detection of Salmonella in spent hens and eggs associated with foodborne infections

About 16,000 spent hens from 23 farms in the northern area of Japan were purchased in 1996, 1997, 1998, and 1999 to isolate Salmonella in two poultry processing plants. Salmonella was detected in 12 of 23 farms (52.2%). In particular, the serotypes Enteritidis and Infantis were detected in four and three farms, respectively. The prevalence rates in the hens' ceca, immature eggs, and the yolk of mature eggs in oviducts were 14%, 7.2%, and 6.8%, respectively. A total of 23 serotypes were detected. The major serotypes of the strains were Enteritidis, Corvallis, Typhimurium, and Infantis, but most of the strains were untypable. In the same area during 1992 to 1996, Salmonella was detected in eggs associated with four outbreaks of Salmonella Enteritidis infection and one outbreak of Salmonella Infantis infection. The ratio of contamination was approximately 1%, and the level was estimated to be 93 MPN(most probable number)/100 g in one outbreak. In farms that produced the eggs associated with all of the five outbreaks of Salmonella, the serotype Enteritidis or Infantis was isolated from hens. Farms where Salmonella was not detected were not related to any of the outbreaks.     

Protection conferred by a live Salmonella Enteritidis vaccine against fowl typhoid in laying hens

Fowl typhoid is under control in poultry farms of developed countries, but it still endemically subsists in commercial laying hen farms of some countries. It has been demonstrated that Salmonella live vaccines can elicit cross-immunity against members of the same Kauffmann-White scheme serogroup. In this work, we explored the protection conferred by TAD Salmonella vac E, a live Salmonella enterica serovar Enteritidis vaccine, against fowl typhoid. Three groups of laying hens were vaccinated with different vaccination schedules starting on the first day of life, and afterwards were infected with 2 x 10(5) CFU of a virulent Salmonella Gallinarum strain, either at wk 28 or wk 52. Mortality, fecal shedding, and organ invasion of Salmonella Gallinarum were assessed. In this work we demonstrated that this Salmonella Enteritidis vaccine is able to cross-immunize against Salmonella Gallinarum. At wk 28, hens vaccinated with three oral doses or with two oral doses combined with one subcutaneous dose were protected by the vaccine. At wk 52, when hens were infected 36 wk after the final immunization, the vaccine was not able to confer protection. Thus, revaccination every 3 mo would be highly recommended. In countries where Salmonella Gallinarum subsists together with Salmonella Enteritidis, control programs should include vaccination of laying hens using safe attenuated Salmonella strains.     

An epidemiological analysis of Salmonella enteritidis contamination in a rat-infested chicken layer farm, an egg processing facility, and liquid egg samples by pulsed-field gel electrophoresis

In order to determine the epidemiological link between the Salmonella Enteritidis contamination in a rat-infested chicken layer farm, an attached egg processing facility and liquid egg samples, several S. Enteritidis isolates were analyzed by pulsed-field gel electrophoresis (PFGE) and bacteriophage typing. A total of 33 S. Enteritidis strains were isolated from a total of 4,081 samples. Similar pulsed-field patterns were generated by S. Enteritidis isolates from liquid eggs, rats and effluent water. Additionally, only two phage types were detected among the S. Enteritidis isolates, PT 1b and PT 6. These results suggest that S. Enteritidis isolates from rats, egg processing facility, and liquid eggs are genetically related. Furthermore, S. Enteritidis infection in rats in layer farms poses a serious public health concern and should be included in future epidemiological studies.     

Salmonella enterica serovar Enteritidis colonization of the chicken caecum requires the HilA regulatory protein

Invasion of Salmonella into intestinal epithelial cells is believed to be essential for the pathogenesis of Salmonella infections. Invasion is mediated by genes located on the Salmonella pathogenicity Island I (SPI-1), which are needed for assembling a type three secretion system, that mediates injection of bacterial proteins into the cytosol of epithelial cells, resulting in cytoskeletal rearrangements and as a consequence invasion. HilA is the key regulator of the Salmonella Pathogenicity Island I. To assess the role of hilA in colonization of gut and internal organs in poultry, animals were infected with 10(8) CFU of a delta hilA mutant of S. Enteritidis and its parent strain at day of hatch. Very low numbers of delta hilA mutant strain were able to colonize the internal organs shortly after infection, but they were not eliminated from internal organs at 4 weeks post-infection. At that time, the colonization level of the wild type bacteria in internal organs was decreased to the same low level compared with delta hilA mutant strain bacteria. Shedding of the delta hilA mutant strain and colonization of the caeca was seriously decreased relative to the parent strain starting from Day 5 post-infection. At 4 weeks post-infection, the delta hilA mutant strain was more or less eliminated from the chicken gut, while the parent strain was still shed to a high level and colonized the caeca to a high extent (more than 10(7) CFU/g). It is concluded that hilA is involved in long-term shedding and colonization of S. Enteritidis in the chicken caeca.     

Comparison of Salmonella enterica serovar enteritidis levels in crops of fed or fasted infected hens

Long-term feed withdrawal has been shown to increase ileocecal intestinal colonization and fecal shedding of Salmonella enterica serovar Enteritidis in challenged hens. Less information is available regarding effects of fasting on crop colonization. Two trials were conducted to compare effects of 14-day feed withdrawal vs. full feed on crop colonization in hens challenged with Salmonella Enteritidis. The levels of Salmonella Enteritidis in the crops of fasted hens were significantly higher than in nonfasted hens on days 3 and 10 and days 3, 9, and 16 postinfection (PI) in trials 1 and 2, respectively. Fecal shedding of Salmonella Enteritidis was significantly increased in the fasted hens on day 10 PI in trial 1. Analysis of crop IgA anti-Salmonella Enteritidis lipopolysaccharide levels in crop lavage samples of hens in trial 1 revealed a humoral response PI in both treatment groups with no significant differences, although peak response for fasted hens occurred 1 wk later. Histologic evaluation of hematoxylin and eosin-stained crop sections from trial 1 birds revealed mild to moderate heterophilic infiltration within the crop lamina propria (LP) or LP and epithelium of nonfasted infected hens at 24 and 96 hr PI. In comparison, heterophils in crops of fasted hens infected at this time point were sparse, indicating a possible diminished heterophil response in the fasted birds. Multifocal areas of tissue inflammation, as indicated by marked heterophil infiltration, with necrosis and sloughing of epithelium, were observed in crops from fasted hens at day 11 PI (14th day of feed withdrawal) but not in the fed groups. This severe heterophilic inflammation was observed in both challenged and nonchallenged fasted hens, suggesting that some factor other than Salmonella Enteritidis was responsible. These results indicate that feed withdrawal can have a dramatic effect on the integrity of the crop and its ultimate response to infection.     

Extended spectrum beta-lactamase SHV-12-producing Salmonella from poultry

Salmonella strains isolated from poultry and poultry products over the period 2005-2006 have been investigated in order to ascertain the presence of extended spectrum cephalosporins (ESC) resistance. Twelve (ESC)-resistant isolates (n=1 S. Enteritidis, n=1 S. Braenderup and n=10 S. Livingstone) were characterized as SHV-12-positive. The multi-drug resistant S. Livingstone SHV-12-producing isolates, untypeable by pulsed-field gel electrophoresis (PFGE), showed a clonal relationship by random amplified polymorphic DNA (RAPD) analysis. The SHV-12 beta-lactamase is reported for the first time in Salmonella enterica strains isolated from poultry in Italy. The results suggest poultry as a source of Salmonella carrying extended spectrum beta-lactamases (ESBLs) genes and highlights the need of monitoring animal productions to prevent spreading of (ESC)-resistant strains.     

Immune dynamics following infection of avian macrophages and epithelial cells with typhoidal and non-typhoidal Salmonella enterica serovars; bacterial invasion and persistence, nitric oxide and oxygen production, differential host gene expression, NF-κB signalling and cell cytotoxicity

Poultry-derived food is a common source of infection of human with the non-host-adapted salmonellae while fowl typhoid and pullorum disease are serious diseases in poultry. Development of novel immune-based control strategies against Salmonella infection necessitates a better understanding of the host-pathogen interactions at the cellular level. Intestinal epithelial cells are the first line of defence against enteric infections and the role of macrophages is crucial in Salmonella infection and pathogenesis. While gene expression following Salmonella infection has been investigated, a comparison between different serovars has not been, as yet, extensively studied in poultry. In this study, chicken macrophage-like cells (HD11) and chick kidney epithelial cells (CKC) were used to study and compare the immune responses and mechanisms that develop after infection with different Salmonella serotypes. Salmonella serovars Typhimurium, Enteritidis, Hadar and Infantis showed a greater level of invasion and/or uptake characters when compared with S. Pullorum or S. Gallinarum. Nitrate and reactive oxygen species were greater in Salmonella-infected HD11 cells with the expression of iNOS and nuclear factor-κB by chicken macrophages infected with both systemic and broad host range serovars. HD11 cells revealed higher mRNA gene expression for CXCLi2, IL-6 and iNOS genes in response to S. Enteritidis infection when compared to S. Pullorum-infected cells. S. Typhimurium- and S. Hadar-infected HD11 showed higher gene expression for CXCLi2 versus S. Pullorum-infected cells. Higher mRNA gene expression levels of pro-inflammatory cytokine IL-6, chemokines CXCLi1 and CXCLi2 and iNOS genes were detected in S. Typhimurium- and S. Enteritidis-infected CKC followed by S. Hadar and S. Infantis while no significant changes were observed in S. Pullorum or S. Gallinarum-infected CKC.     

Development of a complete ELISA using Salmonella lipopolysaccharides of various serogroups allowing to detect all infected pigs

Although poultry is recognized as the major source of food-poisoning caused by Salmonella, pork also contributes to human infections. This study was therefore undertaken in order to develop a reliable serological method for the evaluation of the Salmonella status of piglets. A complete ELISA was performed using lipopolysaccharides of Salmonella Typhimurium, Anatum, Hadar and Infantis because these serovars were representative of the serogroups isolated from 30 contaminated fattening farms. S. Enteritidis was also added because of its importance in human infection and to include the O:9 antigen. This method potentially detects 100% of infected pigs. A significant correlation was found between this serological method and the bacteriological data from mesenteric lymph nodes (p = 0.01). In addition, both sensitivity and specificity were high (97% and 94% respectively). The ELISA test was therefore used in a cross-sectional study on 4 farms to evaluate when pigs became contaminated: seropositive pigs were only found for the 20 week old finishing pigs. The antibody response to Salmonella in piglets was also investigated: maternal antibodies persisted until 7 weeks of age and post-Salmonella contamination seroconversion was detected from 8 weeks of age onwards.     

Longitudinal monitoring of two commercial layer flocks and their environments for Salmonella enterica serovar enteritidis and other Salmonellae

Between August 20, 2001, and September 17, 2002, 1429 samples including drag swabs, egg belt or egg rollout swabs, fan-blade swabs, rodent organ and intestinal pools, beetle (Alphitobius diaperinus) pools, housefly (Musca domestica) pools, chicken organ and intestinal pools, and egg pools were obtained for Salmonella culture from two flocks from two different commercial layer ranches. The two ranches were purposefully selected for the study based on their previous status of Salmonella Enteritidis isolation using environmental drag swabs in cooperation with practicing veterinarians. Salmonella sp. was isolated from 337 out of 979 (34.42%) non-egg samples. No Salmonella was isolated from 450 egg pools collected from either ranch. S. enteritidis was isolated from samples obtained from ranch 1 from manure drag swabs, 4/284 (1.4%); rodent organs, 1/24 (4.2%); and housefly pool cultures 1/21 (4.8%). Salmonella Enteritidis was isolated from ranch 2 from mouse organ and intestinal pool samples, 1/24 (4.2%). Salmonella group B was isolated from all sample types except the insects. There was a statistically significant difference in isolation rates among seven serogroups of Salmonella: groups B, C1, C2, D, E, K, and untypeable (Pearson chi-square 18.96, P = 0.002). Overall, statistically significant differences were observed with respect to Salmonella isolation among the types of samples taken (Pearson chi-square 118.54, P < 0.0001). Intensive monitoring for Salmonella Enteritidis can be used to optimize a Salmonella reduction program for an individual poultry biosecurity unit.     

Determination of the incidence of Salmonella spp., Campylobacter jejuni, and Clostridium perfringens in wild birds near broiler chicken houses by sampling intestinal droppings

Several methods were evaluated for collecting fecal and intestinal samples from wild birds found near broiler chicken houses. A few intestinal samples and cloacal swabs were obtained from European starlings and house sparrows. Most of the samples collected consisted of wild bird droppings found on or near the houses. Samples were collected from each of four farms of a broiler integrator during a grow-out cycle: a cycle in the summer for farm A, fall for farm B, and spring, summer, fall, and winter for farms C and D. Of the 25 wild bird intestinal and fecal samples collected from a broiler house on farm A during a grow-out cycle in July-August 1997, 24% were positive for Salmonella spp., 4% for Campylobacter jejuni, and 28% for Clostridium perfringens. Of the nine fecal samples collected from broiler house B in a grow-out cycle in September-November 1997, 33% were positive for Salmonella spp., 11% for C. jejuni, and 22% for C. perfringens. For farms C and D, of the 23 samples collected in March-April 1998, 0 were positive for Salmonella spp., 11% for C. jejuni, and 52% for C. perfringens; of 27 samples collected in June-July 1998, 4% were positive for Salmonella spp., 0 for C. jejuni, and 13% for C. perfringens; of 24 samples collected in August-October 1998, 14% were positive for Salmonella spp., 5% for C. jejuni, and 4% for C. perfringens; of 14 samples collected December 1998-January 1999, 0 were positive for Salmonella, 50% for C. jejuni, and 14% for C. perfringens. The incidence of these bacterial enteropathogens in wild birds near the broiler chicken houses suggests that wild birds that gain entry to poultry grow-out houses have the potential to transmit these pathogens to poultry.     

Persistent fecal Salmonella shedding in five dairy herds

OBJECTIVE: To monitor patterns of Salmonella fecal shedding in naturally infected dairy herds, determine the association between fecal shedding and individual animal production measures, and evaluate potential risk factors for shedding of Salmonella organisms among cattle in dairy herds. DESIGN: Longitudinal study. SAMPLE POPULATION: 5 Ohio dairy herds. PROCEDURE: For 3 herds, fecal samples were collected from all mature cows and unweaned calves 7 times during an 18-month period. For the remaining 2 herds, fecal samples were collected from 50 lactating cows 6 times during a 12-month period. Individual animal production records for 3 herds were used to examine associations between individual fecal Salmonella shedding status and 305-day mature-equivalent milk production, somatic cell count, milk fat content, and milk protein content. Multivariable logistic regression was used to test for associations between fecal shedding status and breed, lactation status, lactation number, and duration of lactation. RESULTS: None of the adult animals had clinical signs of salmonellosis, but prevalence of fecal Salmonella shedding at individual collection times ranged from 0 to 99% for cows and from 0 to 67% for unweaned calves. Mature cows were more likely to be shedding Salmonella organisms than were unweaned calves. Within herds, lactation status and duration of lactation for individual animals were associated with Salmonella shedding status. Salmonella fecal shedding status was not associated with individual cow production measures. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that subclinical fecal Salmonella shedding can persist in dairy herds for up to 18 months with no measurable effects on health or production of individual cows.     

Prevalence of fecal shedding of Salmonella spp in dairy herds

OBJECTIVE: To estimate prevalence of Salmonella spp in Ohio dairy farms and to identify potential risk factors for fecal shedding of salmonellae. DESIGN: Cross-sectional study. SAMPLE POPULATION: 105 Ohio dairy farms. PROCEDURE: Individual fecal samples from all mature cows in study herds were tested for Salmonella spp by use of standard bacteriologic culture procedures. Herds were identified as infected if at least 1 cow was shedding Salmonella spp. Information regarding herd characteristics, management practices, and health history were collected. Potential risk factors for herd-level Salmonella infection were identified. RESULTS: In 31% of the study herds (95% confidence interval, 22 to 40%), at least 1 cow was shedding Salmonella spp. Six percent of 7,776 fecal samples contained Salmonella organisms; prevalence within infected herds ranged from < 1 to 97%. Herd size, use of free stalls for lactating and nonlactating cows, and use of straw bedding in nonlactating cows were significantly associated with fecal shedding of Salmonella spp, as determined by use of univariate analysis. By use of multivariate analysis, large herds were more likely to be infected than smaller herds; however, no other factors were associated with Salmonella infection after adjustment for herd size. CONCLUSIONS AND CLINICAL RELEVANCE: Subclinical shedding of Salmonella spp is common in Ohio dairy herds, although we could not identify specific interventions that may influence the prevalence of Salmonella spp on dairy farms. It appears that large herd size and intensive management may provide an environment conducive to Salmonella shedding and chronic dairy herd infection.     

A case of canine salmonellosis due to Salmonella infantis

A 7-year-old male dog kept outdoors manifested severe watery diarrhea with generalized weakness. Salmonella Infantis was isolated from a fecal sample and the dog recovered soon after medication with ampicillin, to which the isolate was highly sensitive. The present case was diagnosed as S. Infantis infection. Due to the importance of Salmonella in public health, soil samples were collected from the garden where the dog was kept and were examined for Salmonella, Some of them were positive for S. Infantis, however, no Salmonella was isolated from any soil samples collected after thorough disinfection of the surrounded environment.     

Persistence of salmonella enteritidis in poultry units and poultry food

1. Studies on the survival of Salmonella enteritidis in poultry units and food were carried out over a two‐year period.

2. The organism persisted for at least one year in an empty trial house at the laboratory in which naturally‐infected broiler breeder birds had previously been housed. A similar survival period was seen in a building which had housed an infected layer breeder flock, although infection was not detected in a subsequent pullet flock.

3. Salmonella enteritidis was also frequently found surviving outside poultry houses in small pockets of litter and fan dust which had been left after cleansing and disinfection of the site. On some poultry units S. enteritidis was also found in wild bird droppings.

4. Salmonella contamination appeared to persist preferentially in association with dust particles swept from the floor and in food troughs and S. enteritidis survived at least 26 months in artificially contaminated poultry food.     

Specific adhesion and invasion of Salmonella Enteritidis in the vagina of laying hens

Salmonella Enteritidis is the predominant serovar associated with egg-borne salmonellosis in humans. The colonization of S. Enteritidis in the vagina may play a role in the production of S. Enteritidis-contaminated eggs. In the first experiment, the in vitro adhesion of S. Enteritidis in vaginal and follicular explants was compared with that of S. Typhimurium by bacteriological isolation methods. The mean number of S. Enteritidis associated with vaginal explants was significantly (P < 0.05) higher than S. Typhimurium associated with vaginal explants and both serovars associated with follicular explants. In the second experiment, the in vitro adhesion and invasion of S. Enteritidis strains in the vaginal epithelium was compared with that of several strains of S. Agona, S. Infantis, S. Hadar, S. Heidelberg, S. Montevideo and S. Typhimurium, by immunohistochemical methods. The mean number of Salmonella in the vaginal epithelium depended on their lipopolysaccharide (LPS) type, with the rank order as follows: LPS type O9 (S. Enteritidis) > LPS type O4 (S. Agona, S. Typhimurium and S. Heidelberg) > LPS type O7 (S. Montevideo and S. Infantis) and LPS type O8 (S. Hadar). This rank order of Salmonella invasiveness is in accordance with the frequency of Salmonella outbreaks involving contaminated eggs. These findings suggest that S. Enteritidis has a higher ability to colonize the vaginal epithelium than other serovars, and the Salmonella LPS type may play an essential role in tropism of the reproductive tract.     

Variation in Salmonella enteritidis RAPD-PCR patterns may not be due to genetic differences

Salmonella Enteritidis is a leading cause of gastroenteritis associated with consumption of contaminated poultry meat and eggs. Because pulsed-field gel electrophoresis (PFGE) has limited utility in distinguishing between clonal Salmonella Enteritidis isolates, random amplified polymorphic DNA (RAPD) PCR has been recommended as an alternative molecular fingerprinting tool. This study's objective was to determine whether increasing PCR stringency would improve the repeatability of RAPD DNA patterns based on assessment of target sites within the genome. An in silico PCR was performed to predict amplification products from an Salmonella Enteritidis genome sequence for three different RAPD primers (1247, 1283, and OPA4) and to determine whether any primer would be more likely to amplify variable regions within the genome. A comparison of within- and between-isolate similarities in RAPD patterns was performed using primer 1247, which was predicted by in silico analysis to yield a variable size range of amplicons. In order to reduce artifactual variability associated with the method, three different methods for template preparation were evaluated. All were found to provide comparable results with respect to the similarities observed with repeated analyses of the same Salmonella Enteritidis isolates (n = 18, P = 0.91). Although the median within-isolate similarity (76.0%) was significantly greater than the median between-isolate similarity (66.7%; P = 0.001), duplicate RAPD-PCR runs of the same Salmonella Enteritidis isolates produced DNA patterns that ranged in similarity between 61.5 and 100%. These results indicate that the repeatability of RAPD-PCR is insufficient to distinguish genetic differences among related and unrelated Salmonella Enteritidis isolates.