Preliminary Observations on the Occurrence of Campylobacter jejuni at Four California Chicken Ranches

Transcription

1 9 Journal of Food Protection, Vol. 47, No. 4 t Pages 9-98 (April 984) Copyright, International Association of Milk, Food, and Environmental Sanitarians Preliminary Observations on the Occurrence of Campylobacter jejuni at Four California Chicken Ranches R. E. SMITHERMAN, C. A. GENIGEORGIS* and T. B. FARVER Department of Epidemiology and Preventive Medicine, School of Veterinary Medicine, University of California,Davis, California9566 (Received for publication August 9, 98) ABSTRACT During a -month period from April to July, 98, three Campylobacter jejuni survey studies were done at four chicken ranches in Northern California. In Survey, 9 of 00 (4.5%) total cloacal swab and dropping samples collected from the 0 occupied houses on the four ranches were positive for C. jejuni. Positive samples were from two of the four ranches. On one of these ranches, both occupied houses were positive. However, on the other ranch, only one of six houses was positive. Follow-up feed, water, litter and dropping samples were collected from three houses on this latter ranch during Survey. Again, positive samples were from only one house with 6 of 0 (86.7%) droppings positive for C. jejuni and of 0 (5%) water samples positive. No feed or litter samples were positive. During Survey, cloacal swabs or bird dropping samples were collected from two houses on each of three ranches at approximately weekly intervals from the time of arrival of new flocks of chickens. All six houses ultimately became positive. The first positive samples were collected from one house when chickens in that house were d old. This house had contained C. jejuni-positive chickens during Survey, had old litter, and had not been very thoroughly cleaned and disinfected. First positive samples in each of the other five houses were collected when chickens were between 40 and 46 d old. Two of these houses had previously been positive for C. jejuni but had old litter replaced with new and had been thoroughly cleaned and disinfected. The three other houses had been negative for C. jejuni during Survey. Two of these houses had new litter and had been well cleaned. The other house contained old litter and had not been well cleaned. When each house became positive, virtually all samples from that house were positive within a week indicating that C. jejuni likely spreads rapidly among birds in the house. Over the past decade, Campylobacter jejuni has been recognized as an important cause of human enteritis with isolation rates from diarrheic patients often equaling or exceeding the rates for Salmonella spp. or Shigella spp. (,,,5,). Animals are the main reservoir for C. jejuni (5,), and an important means of transmission from animals to man is via foods of animal origin (,). Prominent among these foods are various poultry products. Poultry has been incriminated as being responsible for several foodborne outbreaks of Campylobacter enteritis (4,4,0,). Furthermore, other associations have been demonstrated between Campylobacter enteritis and poultry. Kist (6) and Norkrans and Svedhem () found that significantly more Campylobacter-afflicted patients than controls had consumed poultry. Severin (9) demonstrated that Campylobacterafflicted patients more often used shorter cooking times for preparing chickens than did control subjects. A number of studies have shown that C. jejuni is commonly present in retail poultry products (6,5,,6,0). Other studies have demonstrated that C. jejuni is often present in poultry arriving at slaughterhouses and that commonly some C. jejuni survive routine slaughtering and processing procedures (6,,7,9,,0,,6). Similarly, C. jejuni has been found in large proportion of poultry at the farm level (,4,7,8). However, as indicated in data collected by Wempe et al. (6), the carriage rates of C. jejuni vary considerably, from lots in which no C. jejuni is found to lots in which C. jejuni is found in all birds sampled. Despite this, little research has been done to identify what factors at the farm level are associated with different prevalence rates, and particularly with the presence of C. jejuni in some groups of birds and not in others. Knowledge about and manipulation of such factors would potentially allow the production of poultry with lower levels of C. jejuni and, therefore, provide the consumer with a safer food product. This paper reports preliminary observations on the occurrence of C. jejuni on four California chicken ranches. The observations were made to identify possible factors associated with the presence of C. jejuni and provide direction for further research. MATERIALS AND METHODS Ranches Four chicken ranches participated in this study. These ranches are managed by a single company. This company obtains eggs from a number of different breeder flocks, but the eggs are hatched in a single hatchery. New chicks are distributed to the ranches generally the day after hatching. The birds are raised on litter consisting of rice hulls. Between raises, the houses are spray-disinfected with DC&R Spray Fumigant (Hess and Clark, Inc., Ashland, OH) which in an unpublished study done in our lab was found to kill 0 6 C. jejuni per ml of solution in less than min even when used at one-tenth the recommended concentration. Old litter is completely pushed

2 94 SMITHERMAN ET AL. out and new litter added once yearly, generally after raising five groups of chickens. Although most houses contain birds of both sexes, a few houses have either just pullets or cockerels. Collection of samples A variety of samples were collected from the ranches and analyzed for the presence of C. jejuni. Cloacal swab samples were taken with sterile cotton swabs moistened in nutrient broth. Each swab was placed in a sterile tube containing ml of fresh nutrient broth. Fresh droppings were collected from the floor of chicken houses and placed in sterile Whirl-Pak bags (Nasco, Fort Atkinson, WI). Water samples were collected from individual automatic waterers using sterile pipets and placed in sterile Whirl-Pak bags. Feed samples were collected from individual automatic feeders using sterile tongue blades and placed in sterile Whirl-Pak bags. Litter samples were collected using sterile tongue blades and placed in sterile paper cartons. Litter samples were selected so as not to contain obviously fresh droppings, so that cultures would reflect organisms in the litter only. Media The media were prepared as described by Wempe et al. (6). For nutrient broth, 0 g of polypeptone (BBL Microbiology Systems, Cockeysville, MD), g of yeast extract, 5 g of sodium chloride, and L of distilled water were mixed together. The broth was autoclaved and then allowed to cool to 50 C. Vancomycin (0 mg/l of broth), trimethoprim lactate (5 mg/l) and polymixin B (5000 IU/L) were added. The antibiotics were added as presterilized solutions. The plating medium was a selective medium consisting of 7 g of brain heart infusion, 5 g of Noble agar, 0.5 g of yeast extract and L of distilled water. Following autoclaving and cooling to 50 C, 50 ml of lysed bovine erythrocytes and the same amounts of the three antibiotics used in the broth were added. Sample preparation Tubes containing the cotton swabs were agitated, and the swabs then removed aseptically. Portions (0. and 0.0 ml) of broth from each tube were plated. One gram of each dropping, feed and litter sample was placed in a sterile tube. Nutrient broth was added, 4 ml to each dropping and litter sample and ml to each feed sample. Tubes were agitated for min and then centrifuged at loooxg for 0 min. A 0.- ml portion of supernatant fluid from each feed and litter sample was plated. Portions (0. and 0.00 ml) of supernatant fluid from each dropping sample were plated. Dropping samples collected during Survey were centrifuged and plated following h of holding at 5 C. A 0.-ml portion of each water sample was plated directly onto the selective medium. All plates were incubated at 4 C for 48 h in anaerobic jars containing 5% O :0% C0 :85% N. Identification of colonies Suspect colonies were examined by phase-contrast microscopy for characteristic morphology and darting motility. Suspect colonies were further confirmed as C. jejuni by the following biochemical tests: oxidase and catalase production, nitrate and sodium selenite reduction, H S production, tolerance for % glycine but failure to grow at.5% NaCl, and hippurate hydrolysis (,5). Statistical methods C. jejuni prevalence rates in houses were compared using binomial probability tables (0). Associations between high or low C. jejuni prevalence rates and flock size, age and sex were investigated using a logistic regression model (BMDPLR)(S). RESULTS AND DISCUSSION During the period 6 April to 6 July, 98, three C. jejuni survey studies were done at four chicken ranches in Northern California. Survey To get a general idea of the prevalence of C. jejuni at four chicken ranches, a prevalence survey was conducted in each of the twenty occupied houses. Five cloacal swabs and five fresh droppings were obtained from each house. In addition, the initial number of birds in each house, the sex of the birds, and their ages at the time of sampling were determined. The results of this prevalence survey are shown in Table. C. jejuni was found on two of the four ranches. On one of these ranches, Ranch A, both houses occupied at the time of the survey were positive for C. jejuni. However, on the second ranch, Ranch B, only one of six houses was positive. The six houses on this ranch were in a single row with less than 0 m between each house. The positive house, House, was the second house in the row. No C. jejuni was found on Ranches C and D. Overall, C. jejuni was found in only three of twenty (5%) houses sampled and in 9 of 00 (4.5%) total swab and dropping samples collected. The three houses that were positive were strongly positive, with either nine or ten positive samples out of ten taken. These proportions were significantly different from the zero positive out of ten samples found in all other houses (P<0.0). Previous research has demonstrated that C. jejuni is common in poultry at the farm level, although prevalence seems to vary considerably. This may be due to different isolation procedures used by the various investigators. In this research, some samples with low counts may have been missed because of the direct plating without preenrichment. Richardson and Koornhof (7) found that 6 of 0 (86.7%) samples of fowl feces randomly collected from coops in Sowto, South Africa, were positive for C. jejuni. A high prevalence was also reported by Mehle et al. (9) who found C. jejuni in 9 of 7 (70%) flocks of chickens 0 to 50 d old in Yugoslavia. On the other hand, Prescott and Bruin-Mosch (4) reported a C. jejuni carriage rate in normal chickens of.8%, and Rosef and Kapperud (8) isolated C. jejuni from only 0 of 00 (0%) chicken flocks in Norway on the basis of pooled fecal samples. The prevalence of C. jejuni on the four chicken ranches investigated during Survey is at the lower end of the range of prevalences previously reported. However, it should be noted that the four ranches surveyed were chosen arbitrarily and, therefore, cannot be assumed to be representative of other chicken ranches in the area. Because of the limited number of samples taken in each house, the C. jejuni isolation rates can be viewed as only very rough prevalence estimates. However, the isolation rates can be used to identify each house as having either a low or high C. jejuni prevalence. Logistic regression, therefore, was used to investigate association between low or high C. jejuni prevalence and flock size, age and sex. Stepwise logistic regression using the BMDPLR computer program (8) was used. With this program, the relation between a dichotomous or binary dependent variable and a set of independent variables, which may be categorical or interval-scaled variables, was investigated. The logistic regression model failed to demonstrate any association between C. jejuni prevalence and flock size or sex. The only independent variable that entered the logistic regression model was age. The coefficient for the age com-

3 C. JEJUNI AT CHICKEN RANCHES 95 TABLE. Flock size, sex, age, and Campylobacter jejuni isolation rates for houses on four California chicken ranches, Survey. Farm House Hock size" d Sex Male Age at sampling (d) b 70 6 Cloacal swabs 4/5 No. isolations/no. sampled Fresh droppings 0 Total samples 9/ D Male "Flock size is the initial number of birds placed in the house. b Some houses had birds hatched on two consecutive days. For these houses, the age given is the age of the older birds in the house. c Based on a minimum detection level of 50 CFU/g of dropping. 'tonly Houses and on Farm A were occupied at the time of sampling. ponent of the model was positive indicating an increasing probability of being positive for C. jejuni with increasing age. However, as illustrated by Table, two houses that contained older, positive birds were on the same ranch, Ranch A. Thus, the ranch may be a confounding variable, with the apparent association between C. jejuni prevalence and age actually the result of the older birds surveyed being on the same ranch. On the other hand, in a study of turkey poults in two brooder houses, Acuff et al. (7) also suggested that a higher C. jejuni isolation rate in one house might have been associated with the older age of the turkeys in that house. Similarly, data collected in Survey to be described in this paper also suggest a possible age association. The association between prevalence and age clearly needs further study. Finding C. jejuni in only one of six houses on Ranch B was somewhat surprising, especially considering that the ranch manager and other personnel move freely between the houses. Therefore, a second survey was conducted in which follow-up samples were taken at this ranch. Survey Follow-up samples were taken in each of three houses on Ranch B, i.e., the house positive for C. jejuni in Survey, a house next to the positive house, and the house farthest from the positive house. The birds in these houses were the same ones that had occupied the houses during the first survey. Follow-up samples included fresh droppings, water, feed and litter samples collected from a variety of locations in each of the three houses. The ranch manager was also interviewed about differences between the positive house and the other houses on the ranch. C. jejuni isolations from the samples are shown in Table. Again, C. jejuni was found only in House, with 6 of 0 (86.7%) droppings positive for C. jejuni and of 0 (5%) water samples positive. Counts of C. jejuni in dropping samples ranged from.5 x 0 5 CFU/g to greater than.5 x 0 6 CFU/g, with a geometric mean greater than TABLE. Campylobacter jejuni isolation rates from droppings, feed, litter and water samples from three houses on Ranch B, Survey. No. isolations/no. sampled House 6 Fresh droppings" 0/0 6/0 0/0 JOURNAL OF FOOD PROTECTION VOL. 47, APRIL 984 Feed samples b Litter samples 0 0/0 0/0 Water samples' 0/9 /0 "Based on a minimum detection level of 50 CFU/g. Dropping samples were taken when chickens in Houses and were 48 d old and chickens in House 6 were 44 d old. b Based on a minimum detection level of 5 CFU/g. c Based on a minimum detection level of 5 CFU/g. d Based on a minimum detection level of 5 CFU/ml. Minimum detection levels vary with type of sample because of variations in dilutions and amounts plated on agar media.

4 96 SMITHERMANETAL..8 X 0 6 CFU/g. C. jejuni counts from two of the water samples were 5 CFU/ml and 60 CFU/ml. The number in the third positive water sample could not be accurately counted because of confluence of colonies on the plating medium and masking by other bacteria. Results of the second survey provided further confirmation that only House on Ranch B was positive for C. jejuni. Considering that the houses on the ranch were fairly close together and workers move between the houses taking no special measures to disinfect clothing and shoes, these results suggest that C. jejuni may not be easily transmitted from house to house on the shoes and clothing of workers or via wind-blown dust. Within an infected house, transmission between birds is most likely due to contact with fresh droppings which contain large numbers of C. jejuni. However, as suggested by the three positive water samples, contaminated water may also be a means of transmission. In their study of two turkey brooder houses, Acuff et al. (/) also found C. jejuni in water samples. Interestingly, several outbreaks of human Campylobacter enteritis have been associated with contaminated water (). The water in all houses on Ranch B comes from a single well. Therefore, finding the only positive water samples in House indicates contamination of the water from within the house. Indeed, litter, dirt and apparent fecal material were fairly common in the waterers. The explanation for the presence of C. jejuni in only House is unclear. The chickens in House included birds from two breeder flocks not represented in the other houses on Ranch B. However, birds from these breeder flocks were also in houses on Ranch D from which no C. jejuni was isolated. The ranch manager of Ranch B noted that House was the house with the most flooding problems over the year. C. jejuni has been shown to be very sensitive to drying (7). One can speculate that the wetter litter environment of House might favor better survival of C. jejuni. Furthermore, because of wet litter in House, some litter had been removed and new litter added creating a different litter environment in this house. Fanelli et al. (9) demonstrated that survival of Salmonella infantis and Salmonella typhimurium differed in different litter environments. If C. jejuni behaves similarly, the different litter environment of House might explain the presence of C. jejuni in that house and not in the others. However, despite such speculation, no C. jejuni was found in any of the litter samples. Survey Having accumulated some data on the prevalence of C. jejuni in chickens of a variety of ages at these ranches, the decision was made to follow new flocks of chickens at approximately weekly intervals from their arrival at six houses on Ranches A, B and C. During this part of the study, samples initially taken were cloacal swabs of birds from several different areas in each house. Later, in four of the houses, fresh dropping samples were taken to determine the number of CFU/g dropping. The results of the third survey are shown in Figure. All six houses ultimately became positive for C. jejuni. The first to turn positive was House on Ranch B. The preceding group of chickens in this house had been positive for C. jejuni, the old litter had not been removed, and the cleaning and disinfection of this house had not been very thorough. This house turned positive in less than half the time required for any other house, suggesting that the previously contaminated environment allowed the new flock of chickens to be infected relatively early. Houses and on Ranch A had also previously been positive for C. jejuni, but had had old litter replaced with new and had been more thoroughly cleaned and disinfected. However, the new chickens in these houses did become positive, with the first positive samples being found at 40 and 4 d of age in House and House, respectively. Thus, despite new litter and cleaning and disinfecting, some C. jejuni may have persisted. The birds in House on Ranch A were hatched and placed on the ranch two weeks after the birds in House. Although the first positive samples were found in each house at virtually the same age, this occurred wk later in House. This supports the previous suggestion that C. jejuni may not be readily transmitted from house to house on the shoes and clothing of workers who move freely between the houses. RANCH A /a HOUSE J b RANCH B / ' HOUSE / RANCH C HOUSE / / a RANCH A J a HOUSE / b / a S / RANCH B / HOUSE RANCH C / ' HOUSE 4 / 'I'l' 5,, AGE IN DAYS Figure. Campylobacter jejuni isolations from droppings (points a only) and clocal swabs (all other points) from chickens raised in six houses on three California ranches, Survey. Data is based on 0 samples per sampling date except for sampling points b and preceding points which represent 0 samples. Both houses on Ranch A had been positive for C. jejuni the previous survey, and both houses on Ranch C had been negative. These four houses had litter replaced and had been cleaned and disinfected. House on Ranch B had been negative for C. jejuni and House positive. The old litter in these two houses had not been removed and the houses less thoroughly cleaned and disinfected.

5 C. JEJUNI AT CHICKEN RANCHES 97 Surprisingly, three previously negative houses, House on Ranch B and Houses and 4 on Ranch C, all became positive in this part of the study. Old litter had not been removed from House on Ranch B, and cleaning and disinfecting had not been thorough. However, old litter in Houses and 4 on Ranch C had been replaced with new and the houses thoroughly cleaned and disinfected. The chickens in Houses and 4 on Ranch C turned positive at older ages than the ages of the chickens in these houses sampled during the first survey. Thus, it could be argued that had the first survey been conducted later, these houses may have been positive for C. jejuni. However, the chickens in the houses during the first survey were on old litter and, therefore, might have been expected to turn positive at a younger age, such as the chickens in House on Ranch B. During Survey on Ranch B, chickens in House were cultured as late as 48 d of age and environmental samples taken as late as 60 d, with no samples positive for C. jejuni. Thus, this house appears to have been truly negative. The explanation for why previously negative houses turned positive may lie in the weather. During the time that these houses became positive, the region was experiencing unusually hot weather with temperatures some days exceeding 7.8 C (00 F). To increase ventilation, the doors to the house were opened. This, in turn, may have allowed wild birds greater access to the houses. Studies have shown that C. jejuni can commonly be isolated from a number of wild birds (8,4). Furthermore, heat stress by the birds may have made them more susceptible to C. jejuni infection. Such a stress effect has been demonstrated for other bacterial infections in poultry (5, C. C. Tsai, 97, MPVM Report, University of California, Davis). Interestingly, in North America and Europe, C. jejuni isolations from human patients with diarrhea show seasonal variation with peaks in the summer (,). A possible explanation suggested by our data is that the prevalence of C. jejuni in chickens may increase in the summer potentially leading to more contaminated retail poultry products. It is interesting that except for House on Ranch B all the houses turned positive when the chickens were around 6 wk old. One might have expected Houses and on Ranch A to become positive at similar ages because of probable similar levels of C. jejuni in these houses. Birds in the other three houses happened to be about 6 wk of age when the hotter weather appeared. Other explanations for the marked similarity in ages of these birds when they became positive are unclear. As Figure illustrates, when each house became positive, virtually all samples from that house were positive within wk. Thus, C. jejuni appears to spread rapidly among the birds in the house. This agrees with the observation by Mehle et al. (9) that, where flocks of chickens become colonized by C. jejuni, all the birds will likely be involved. Droppings were collected from four of the houses, and the number of CFU/g determined. Counts ranged from.5 x 0 CFU/g to greater than.5 x 0 6 CFU/g, with a geometric mean greater than.5 x 0 5 CFU/g. Clearly, this study is only a small step toward elucidating the epidemiology of C. jejuni in chickens. More questions are posed than answered. The ultimate source or sources of C. jejuni, the means of transmitting it between houses, the role of water in transmitting C. jejuni between birds, and conditions for survival for C. jejuni in litter all warrant further study. Possible age and weather effects on C. jejuni prevalence are suggested by our data. However, in Survey, these effects could not be isolated. Additional studies to more critically evaluate these effects are needed. ACKNOWLEDGMENTS This work was supported in part by funds provided by the Dean's Office as part of the State's allocation to the School of Veterinary Medicine for fiscal The senior author expresses his gratitude to the United States Air Force for the opportunity to pursue graduate studies during which this research was conducted. REFERENCES. Acuff, G. R., C. Vanderzant, F. A. Gardner, and F. A. Golan. 98. Examination of turkey eggs, poults, and brooder house facilities for Campylobacter jejuni. J. Food Prot. 45: Blaser, M. J. 98. Campylobacter jejuni and food. Food Technol. 6(): Blaser, M. J., and L. B. Reller. 98. Campylobacter enteritis. N. Engl. J. Med. 05: Brouwer, R., M. J. A. Mertens, T. H. Siem, and J. Katchaki An explosive outbreak of Campylobacter enteritis in soldiers. Antonie van Leeuwenhoek 45: Brownell, J. R., W. W. Sadler, and M. J. Fanelli Factors influencing the intestinal infection of chickens with Salmonella typhimurium. AvianDis. : Christopher, F. M., G. C. Smith, and C. Vanderzant. 98. Examination of poultry giblets, raw milk, and meat for Campylobacter fetus subsp. jejuni. J. Food Prot. 45: Doyle, M. P., and D. J. Roman. 98. Sensitivity of Campylobacter jejuni to drying. I. Food Prot. 45: Engleman, L. 98. Stepwise logistic regression, pp In W. J. Dixon (ed.) BMDP statistical software. University of California Press, Berkeley, CA. 9. Fanelli, M. J., W. W. Sadler, and J. R. Brownell Preliminary studies on persistence of salmonellae in poultry litter. Avian Dis. 4: Fleiss, J. L. 98. Statisical methods for rates and proportions. John Wiley and Sons, New York.. Hartog, B. J., and E. DeBoer. 98. Campylobacter jejuni in poultry products from retail outlets and in poultry slaughterhouses, p. 70. In D. G. Newell (ed.) Campylobacter: epidemiology, pathogenesis, and biochemistry. MTPPress, Lancaster.. Harvey, S. M Hippurate hydrolysis by Campylobacter fetus. J. Clin. Microbiol. : Hayek, L. J., and J. G. Cruickshank Campylobacter enteritis. Br. Med. J. :9. 4. Itoh,T.,K.Saito,Y.Yanagawa,S.Sakai,andM.Ohashi. 98.Campylobacter enteritis in Tokyo, pp. 5-. In D. G. Newell (ed.) Campylobacter: epidemiology, pathogenesis, and biochemistry. MTP Press, Lancaster. 5. Kinde, H., C. A. Genigeorgis, and M. Pappaioanou. 98. Prevalence of Campylobacter jejuni in chicken wings. Appl. Environ. Microbiol. 45: Kist, M. 98. Campylobacter enteritis: epidemiological and clinical data from recent isolations in the region of Freiburg, West Germany. pp In D. G. Newell (ed.) Campylobacter: epidemiology, pathogenesis, and biochemistry. MTP Press, Lancaster. 7. Luechtefeld, N. W., and W.-L. L. Wang. 98. Campylobacter fetus subsp. jejuni in a turkey processing plant. J. Clin. Microbiol. :66-68.

6 98 SMITHERMANETAL. 8. Luechtefeld, N. A. W., M. J. Blaser, L. B. Reller, and W.-L. L. Wang Isolation of Campylobacter fetus subspecies jejuni from migratory waterfowl. J. Clin. Microbiol. : Mehle, J., M. Gubina, andb. Gliha. 98. Contamination of chicken meat with Campylobacter jejuni during the process of industrial slaughter. pp In D. G. Newell (ed.) Campylobacter: epidemiology, pathogenesis, and biochemistry. MPT Press, Lancaster. 0. Mouton, R. P., J. J. Veltkamp, S. Lauwers, and J. P. Butzler. 98. Analysis of a small outbreak of Campylobacter infections with high morbidity, pp In D. G. Newell (ed.) Campylobacter: epidemiology, pathogenesis, and biochemistry. MTP Press, Lancaster.. Norkrans, G., and A. Svedhem. 98. Epidemiological aspects of Campylobacter jejuni enteritis. J. Hyg. 89: Oosterom, J., S. Notermans, H. Karman, andg. B. Engels. 98.Origin and prevalence of Campylobacter jejuni in poultry processing. J. Food Prot. 46: Park, C. E., Z. K. Stankiewicz, J. Lovett, and J. Hunt. 98. Incidence of Campylobacter jejuni in fresh eviscerated whole market chickens. Can. J. Microbiol. 7: Prescott, J. F., and C. W. Bruin-Mosch. 98. Carriage of Campylobacter jejuni in healthy and diarrheic animals. Am. J. Vet. Res. 4: Prescott, J. F., andd. L. Munroe. 98. Campylobacter jejuni enteritis in man and domestic animals. J. Am. Vet. Med. Assoc. 8: Rayes, H. M., C. A. Genigeorgis, andt. B. Farver. 98. Prevalence of Campylobacter jejuni on turkey wings at the supermarket level. J. Food Prot. 46: Richardson, N. J., and H. J. Koornhof Campylobacter infections in Soweto. S. Afr. Med. J. 55: Rosef, O., and G. Kapperud. 98. Isolation of Campylobacter fetus subsp. jejuni from faeces of Norwegian poultry. Acta Vet. Scand. : Severin, W. P. J. 98. Epidemiology of Campylobacter infection, pp In D. G. Newell (ed.) Campylobacter: epidemiology, pathogenesis, and biochemistry. MTP Press, Lancaster. 0. Simmons, N. A., and F. J. Gibbs Campylobacter spp. in ovenready poultry. J. Infect. : Skirrow, M. B. 98. Campylobacter enteritis-the first five years. J. Hyg. 89: Skirrow, M. B., R. G. Fidoe, andd. M. Jones. 98. An outbreak of presumptive food-borne Campylobacter enteritis. J. Infect. :4-6.. Smeltzer, T. I. 98. Isolation of Campylobacter jejuni from poultry carcasses. Austral. Vet. J. 57: Smibert, R. M Vibrio fetus var. intestinalis isolated from the intestinal contents of birds. Am. J. Vet. Res. 0: Ullmann, U Methods in Campylobacter, pp In T. Bergan and J. R. Morris (eds.) Methods in microbiology XIII. Academic Press, New York. 6. Wempe, J. M., C. A. Genigeorgis, T. B. Farver, and H. I. Yusufu. 98. Prevalence of Campylobacter jejuni in two California chicken processing plants. Appl. Environ. Microbiol. 45:55-59.