Salmonella in Wastes Produced at Commercial

Transcription

1 APuLED MICROU10LOOY, Nov. 1969, p Copyright 1969 American Society for Microbiology Vol. 18, No. 5 Printed in U.S.A. Salmonella in Wastes Produced at Commercial Poultry Farms1 D. J. KRAFIT, CAROLYN OLECHOWSKI-GERHARDT,' J. BERKOWITZ, AND M. S. FINSTEIN Department of Environmental Sciences, College of Agriculture and Environmental Science, Rutgers, The State University, New Brunswick, New Jersey Received for publication 19 August 1969 Composite samples of freshly voided excreta from 91 poultry houses were tested qualitatively for Salmonella; 26 (29%) were positive. The houses were located on 36 farms, 18 of which (50%) yielded one or more positive samples. In a separate, quantitative study, Salmonella densities ranged from less than I to over 34,000 per g of excreta (dry weight). High densities were noted in waste from cage houses, but not in waste from floor houses (litter or wire floors). Salmonella-shedding chickens were located in only one small area of the row of cages examined in detail. A total of 15 Salmonella serotypes were identified during the study. Recent studies have shown that Salmonella can be detected with relative ease in many surface waters (1, 4, 7; J. Clemente and R. G. Christensen, Proc. 10th Conf. Great Lake Res. 1:11, 1967), including one that was otherwise of very high quality (2). The current high level of interest in salmonellosis and rapid advances in Salmonella methodology (8, 12) insure the extension of such observations to additional waters, possibly leading to increased efforts to control environmental contamination with this pathogen. Because most salmonellae are non-hostspecific, and some species are widely distributed, they can be introduced into waters from numerous sources. As a result of a comprehensive study of an estuary in England, McCoy (7) tentatively concluded that nearby food and feed processing industries were major contributors of Salmonella. In rural areas, animal-raising industries might be among the more important contributors to the Salmonella burden of waters receiving runoff from feed lots (10) and manured fields. The purpose of this investigation was to evaluate the potential of poultry excreta and manure from commerial farms to contaminate the environment with Salmonella. This waste is of special interest because the domestic fowl is one of the main reservoirs of Salmonella (6). X Paper in the Journal Series, New Jersey Agricultural Experiment Station, Rutgers, The State University, Department of Environmental Sciences, New Brunswick, NJ. I Present address: Laboratory of Applied Microbiology, Division of Engineering and Applied Physics, Harvard University, Cambridge, Mass MATERIAlS AND METHODS Sampling. Freshly voided droppings and old wastes were obtained from commercial farms located in central and southern New Jersey. Old waste was removed from the bottom-most levels of litter or excreta piles underneath caged birds. Fresh droppings were obtained from the surface of litter or excreta piles. With the exception of one flock of turkeys, the samples came from White Leghorn laying hens. Subsamples were thoroughly mixed to provide composites which usually represented a single house, but occasionally an area of a large house. Generally, no more than 4 hr elapsed between sampling and the start of the enrichment. Qualitative survey. One to 3 g (wet weight) of a composite sample (consisting of approximately 10 fresh droppings) was placed into tetrathionate enrichment broth supplemented with 1.0 mg of brilliant green/ liter and 1.25 mg of sulfathiazole/liter (5). Bacteriological products were Difco brand, unless noted otherwise. Except where indicated otherwise, incubations were at 37 C. After 24 and 48 hr of incubation, the enrichments were streaked onto brilliant green agar, and, 18 to 24 hr later, suspect colonies were transferred to triple sugar iron agar (TSI) slants. Cultures showing Salmonella-like reactions after 18 to 24 hr of incubation on SI (acid butt, alkaline slant, H2Si) were tested for urease production (13) and for agglutination with polyvalent-o-antiserum. Ureasenegative cultures which gave rapid and extensive agglutination were considered to be Salmonella. The qualitative survey was carried out during the summer of Quantitative survey. This was accomplished largely during the summer of To provide representative fresh samples for enumerating Salmonella, between 50 and 100 individual fresh droppings were composited. Composite samples of old waste consisted of 50 sub- 703

2 704 KRAr ET AL. APPL. MIcRoBioL. samples obtained from five areas in a house. For an intensive study of one cage house, individual fresh droppings were cultured to study the distribution of Salmonella-shedding chickens. Enumeration was accomplished by use of a modification of the most probable number (MPN) procedure developed by McCoy (8). A 10-g composite sample (wet weight) was combined with 90 ml of tetrathionate broth, dispersed with a sterile blender, and 10-fold serial dilutions were prepared in tetrathionate broth. At most, 15 min elapsed during preparation of a dilution series. In early trials the most concentrated suspension included in the test contained 0.1 g of excreta per tube; this was later raised to 1 g to increase the sensitivity. After incubation for selected intervals, enrichments were streaked onto the bismuth sulfite agar of McCoy (8), which was prepared with yeast autolysate purchased from the Albimi Corporation, Flushing, N.Y. Plates were viewed with transmitted oblique illumination (3) and suspect colonies were transferred to TSI slants. Culture material from the TSI slants was used for the test with 0-antiserum, and also to inoculate H-broth for a subsequent agglutination test with polyvalent-h-antiserum. Dilutions giving rise to cultures which agglutinated with the H-antiserum were scored positive for Salmonella, and the resulting code was used to enter an MPN table (9). Additional details concerning the development of a satisfactory enumeration procedure are given in the results section. RESULTS Qualitative survey. Composite samples of freshly voided excreta representative of 91 poultry houses were tested for Salmonella; 26 of these samples yielded the pathogen. The houses sampled were located on 36 farms, 18 of which provided at least 1 positive sample. At a few farms, Salmonella was detected in all of the houses tested, but more often this bacterium was found in some, but not all of the houses (Table 1). No pattern of geographic distribution of the farms yielding Salmonella was noted; the pathogen was detected on a portion of the farms located in all of the five counties extensively sampled. Feed from at least six different commercial sources was being used on the farms which yielded the pathogen. Where floor-housing was practiced (litter or wire floor), Salmonella was detected in 28% of the houses (16/58). The comparable value for cage housing was 31% (10/32). One house was not classified with respect to housing arrangement. at Quantitative methodology. Preliminary attempts to enumerate Salmonella native to chicken excreta by an MPN procedure utilizing streaking onto brilliant green agar as a component step were unsuccessful. Many suspect colonies picked from this medium proved to be non-salmonella upon furtheh testing. The occurrence of numerous false TABLE 1. Detection of Salmonella in different poultry houses on a given farm Ratio of houses positive to houses tested No. of farms 1/1 1 1/2 4 2/2 3 1/3 4 2/3 4 1/4 1 2/6 1 TABLE 2. Efficiency of enrichment temperatures for enumerating S. typhimurium inoculated into chicken excreta4 Etnrichmnt Enrichment temp. Inoculum c Excreta count Recovery coun C cclsig MPN/g % 37 2,090 1, , , , a Most probable number (MPN) procedure (eight tubes per dilution) started immediately after inoculating excreta. b Refers to tetrathionate broth enrichments; all other incubations were at 37 C. Based on nutrient agar plate counts; inoculum was grown on nutrient agar. positives was tolerable in the qualitative survey where relatively few enrichments were involved, but not in the quantitative survey which required the use of large numbers of enrichment tubes. Replacement of brilliant green agar with the bismuth sulfite agar described by McCoy (8) almost completely eliminated false positives, judged on the basis of agglutination with polyvalent-h-antiserum; 105 of 106 suspect colonies agglutinated. Polyvalent-O-antiserum proved less inclusive; only 85 of these suspect colonies agglutinated. One suspect colony was polyvalent-opositive but -H- negative. Approximately 20% of the polyvalent-h-positive cultures were sent to the New Jersey State Department of Health, Trenton, for serotyping, and all were assigned serotype identities. The polyvalent-o-positive, -H-negative culture was judged to be non- Salmonella. The MPN dilution tubes were scored on the basis of the test with polyvalent-hantiserum. The MPN procedure described by McCoy (8) calls for enrichment in tetrathionate broth at 37 C, but a recent report (12) indicates that 41.5 C might be a more favorable temperature for the enrichment step of the procedure. To provide a

3 VOL. 18, 1969 SALMONELLA IN POULTRY WASMES 705 TABLE 3. Counts of Salmonella in fresh and old chicken excreta Farm Houseb House type Sample Count 95% Confidence interval MPN/g. MPN/gc 1 a Floor Fresh b Floor Fresh a Floor Fresh a Cage Fresh d a, Cage Fresh a2 Cage Fresh > 10,6006 4d a2 Cage Fresh 11,900 3,600-39,000 a2 Cage Old b 1 Cage Fresh > 10,000 b2 Cage Fresh a Cage Fresh a Cage Old b Cage Fresh 34,400-10, ,000 b Cage Old a Five-tube MPN. 6 Subscripts refer to different areas in a common house. c MPN, most probable number; oven-dry weight. d Seven days elapsed between sampling times. e Minimum count (all tubes positive). basis for choosing between these alternatives, they were applied to chicken excreta freshly inoculated with known numbers of S. typhimurium. The selective inhibitors, sulfathiazole and brilliant green, were added to the tetrathionate broth intended for use at 37 C, but brilliant green was omitted from the 41.5 C broth because at this temperature the dye severely inhibited the growth of the five Salmonella species tested. The procedures differed in one additional detail. At the lower temperature, enrichments were streaked onto bismuth sulfite agar at 24-hr intervals for 4 days, or until a culture gave rise to Salmonellalike colonies, at which time it was presumed positive and streaking was discontinued. Where enrichment was at the elevated temperature, the interval between streakings was shortened to 18 hr, and streaking was continued for a maximum of 3 days. The results (Table 2) suggest that both enrichment temperatures are suitable since approximately 50% of the inoculated cells were accounted for with either 37 or 41.5 C as the enrichment temperature. The elevated temperature was chosen for subsequent use, however, because this allowed completion of the enrichment step of the enumeration procedure in 3 days as compared to 4 days at the lower temperature. Quantitative survey. To further evaluate chicken waste as a source of Salmonella, the distributions were enumerated in samples obtained from some of the farms which had yielded the pathogen during the qualitative survey. Twenty-eight samples (consisting of an equal number of fresh and old samples) came from floor houses, and 18 came from cage operations (12 fresh and 6 old samples). Salmonella was detected in 14 of the samples tested, but not in the other 32; where present, the numbers ranged from less than 1 to over 34,000 per g of dry excreta (Table 3). At farm 4, samples representing different areas of common houses varied markedly in Salmonella density. Samples a, and a2 represent two rows of cages which were separated by three intervening rows; similarly, samples bl and b2 represent separate areas of a house. A third cage house at farm 4 did not yield the pathogen. All the positive samples obtained from floor houses contained relatively few Salmonella (MPN values ranged from 0.8 to 8.9 per g). In contrast, high numbers of the pathogen were detected in some of the excreta samples from caged birds; 50% of the fresh composite samples from caged birds which contained Salmonella yielded MPN values of over 10,000 per g. Consistent with this observed association of high pathogen density with fresh excreta from caged birds is the finding that three old samples from cage houses were positive, whereas the pathogen was not detected in old materials from floor houses. Because there were 10-fold variations in the amount of excreta placed into the lowest dilutions of the enrichment tubes during the quantitative survey, the method was not uniformly sensitive throughout this phase of the investigation. These data, therefore, cannot be used to judge the rela-

4 706 KRAFT ET AL. APPL. MICRoBioL. TABLE 4. Distribution of individual droppings containing Salmonella in a single row of cages Dropping no. Count 95% Confidence interval MPN/gb MPN/gb 1-15 NDc , , , , NDC Composite sample Two-tube MPN (most probable number), except for the composite sample where five tubes per dilution were used. b Oven-dry weight. Not detected (sensitivity-limit, approximately 2/g). tive frequency of Salmonella occurrence in floor and cage houses. Distribution study. The distribution of Salmonella-shedding chickens, in house a2, farm 4, was investigated 5 months after it had been sampled in the quantitative survey. Aluminum foil was placed on top of the waste pile in six areas of a single row of cages to receive fresh excreta during a 45-min period and to avoid contact with old waste. Each sampling area represented two cages (10 to 13 birds) separated from the next area by approximately 10 cages. Five droppings from each sampling area were tested individually; all the remaining droppings were composited. The pathogen was detected in five of the droppings in numbers ranging from 2 to over 1,000 per gram (Table 4). All of the positive droppings came from a single sampling area, indicating a nonrandom distribution of Salmonella-excreting chickens. Serotypes. The following serotypes were isolated: qualitative survey-s. anatum, S. braenderup, S. infantis [2], S. livingstone [2], S. montevideo, S. saintpaul, S. tennessee, S. typhimurium, S. typhimurium var Copenhagen, S. Worthington [2]; quantitative survey-s. anatum, S. blockley [3], S. memphis, S. Montevideo [4], S. siegburg, S. thompson [2], S. typhimurium, S. typhimurium var Copenhagen; distribution study-s. eimsbuettel. The number of isolations is given in brackets where there was more than one. No farm yielded the same serotype during both surveys. Similarly, the serotype from the distribution study differed from that isolated in the quantitative survey. The quantitative survey list includes three isolations from old waste; these differed in serotype from the isolates found in nearby fresh excreta. DISCUSSION Our surveys were restricted to central and southern New Jersey; this area, however, cannot be considered unusual with respect to the association of Salmonella with domestic fowl. Salmonella is widely distributed, and a high proportion of the isolates reported originate in poultry and poultry products (6). It seems probable that studies similar to the present one, if carried out in other poultry raising areas, would yield similar results. Although samples from floor and caged birds were comparable with respect to the incidence of Salmonella, as shown in the qualitative survey, there were distinct quantitative differences. High densities of this pathogen were found in fresh excreta from caged but not floor birds. Similarly, the pathogen was detected in some samples of old waste from cage houses but not floor houses. These data are not sufficiently extensive to justify generalizations; however, they are suggestive that cage housing may promote the shedding of Salmonella. The possible association between cage housing and enhanced pathogen shedding deserves further attention since this type of management is becoming increasingly popular for economic reasons. Should this association prove generally prevalent, anticipated improvements in Salmonella control resulting from a reduction of this bacterium in animal feeds (6) could be partially offset by the tendency towards cage housing. Because tests of the MPN procedure used showed it to account for about 50% of the cells inoculated into excreta, the quantitative data reported can be considered as conservative estimates of the true values. This reinforces the conclusion that the spreading of some of these wastes could have disseminated substantial numbers of Salmonella onto soil, possibly leading to the contamination of water via surface runoff. Routine tests are now available for the detection of Salmonella in waters, and these are rapidly being improved (8, 12). The application of such methods to surface waters for the purpose of identifying and controlling important sources of Salmonella has been advocated as a part of the effort to combat salmonellosis (11). Acceptance of this approach might, in some situations, bring into question traditional manuring practices. In this context, the recently developed plow-furrowcover method of manure application (C. H. Reed, Proc. Nat. Symp. Animal Waste Management, p , 1966.) is advantageous in burying the waste, thereby preventing its transport by surface runoff. ACKNOWLEDGMENTS Serotypes were detenmined by the Department of Healthr State of New Jersey, Trenton, for which we thank M. Goldfield,

5 VOL. 18, 1969 SALMONELLA IN POULTRY WASTES 707 Director of Laboratories. We are grateful to S. E. Katz for helpful discussions and for critically reviewing the manuscript. We thank Margaret R. Bitzky for excellent technical assistance. This investigation was supported by Public Health Service grant UI from the National Center for Urban and Industrial Health, Consumer Protection and Environmental Health Service. LITERATURE CRTD 1. Brezenski, F. T., and R. Russomanno The detection and use of Salmonellae in studying polluted tidal estuaries. J. Water Pollut. Contr. Fed. 41: Fair, J. F., and S. M. Morrison Recovery of bacterial pathogens from high quality surface water. Water Resour. Res. 3: Finkelstein, R. A., and K. Punyashthiti Colonial recognition, a "new" approach for rapid diagnostic enteric bacteriology. J. Bacteriol. 93: Gallagher, T. P., and D. F. Spino The significance of numbers of coliform bacteria as an indicator of eateric pathogens. Water Res. 2: Galton, M. M., J. E. Scatterday, and A. V. Hardy Salmonellosis in dogs; bacteriology, epidemiology, and clinical considerations. J. Infect. Dis. 91: Galton, M. M., J. H. Steele, and K. W. Newell Epidemiology of salmonelloss in the United States, p In The world problem of saimonello W. Junk, Publisher, The Hague, Netherlands. 7. McCoy, J. H Salmonellac in crude sewage, sewage effluent and sewage-polluted natural waters, p In B. A. Southgate (ed.), Advances in water pollution research, vol. 1. Pergamon Press, Ltd., Oxford, England. 8. McCoy, J. H The isolation of Salmonellae. J. Appl. Bacteriol. 25: Meynell, G. G., and E. Meynell Theory and practice in experimental bacteriology. Cambridge Univ. Press, Cambridge, England. 10. Miner, J. R., L. R. Fina, and C. Piatt Salmonella infands in cattle feedlot runoff. AppL Microbiol. 15: Newell, K. W Possibilities for investigation and control of salmonello for this decade. Amer. J. Public Health Nat. Health 57: Spino, D. F Elevated temperature technique for the isolation of SalmoneUa from streams. AppL MicrobioL 14: Stuart, C. A., E. Van Stratum, and R. Rustigian Further studies on urease production by proteus and related organisms. J. Bacteriol. 49: Downloaded from on December 17, 2018 by guest