Enterobacteriaceae Associated with Meats and Meat Handling

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 1981, p /81/ $02.00/0 Vol. 41, No. 4 Enterobacteriaceae Associated with Meats and Meat Handling MICHAEL E. STILES* AND LAI-KING NG Faculty of Home Economics and Department ofmicrobiology, The University ofalberta, Edmonton, Alberta, Canada T6G 2M8 Received 21 November 1980/Accepted 5 January 1981 The source of Enterobacteriaceae on meats was shown to be associated with the meat-handling work surfaces in the two packing plants studied. A total of 2,343 Enterobacteriaceae were isolated and identified from meat samples and work surfaces at the packing plants and at the retail facilities. Escherichia coli biotype I and Serratia liquefaciens were detected at all stages of meat handling, indicating that they may be present in meats throughout the meat-handling system. Enterobacter agglomerans and S. liquefaciens were the predominant Enterobacteriaceae at the retail level, but they had limited indicator potential for sanitation and hygiene. Klebsiella pneumoniae was a frequent isolate among Enterobacteriaceae from meats and meat-handling surfaces in the packing plants but not at the retail level, indicating that this organism might signal unhygienic handling of meats at the retail level. Escherichia coli biotype I is one of the predominant Enterobacteriaceae in ground beef similar fecal origin, and hence public health cated in human infections (3, 12), and has a (11, 28). With good manufacturing practices, E. significance, as typical E. coli (4, 17). Other coli I contamination on meat is generally considered to come from the skin or hide of animals for example, Citrobacter freundii is found in Enterobacteriaceae are of less specific origin; during processing (1, 14, 30) and possibly represents both fecal and nonfecal contamination spp. are reported to occur only rarely in the feces but also in soil (4, 29), and Enterobacter (27). Furthermore, E. coli I in the stationary human intestine (5). growth phase survives well in frozen and nonfrozen meat (18, 32) and grows in meats at bacteriaceae in ground beef were found to be E. In an earlier study (28), the principal Entero- improper storage temperatures. This fact led coli type I, Enterobacter agglomerans, and Serratia liquefaciens. Other Enterobacteriaceae many workers to criticize the use of E. coli as an indicator of sanitation and hygiene in raw meats were also identified on selective media, including (13, 15, 20). In contrast, Mossel and co-workers C. freundii, Enterobacter aerogenes, E. cloacae, (5, 21, 24, 25) criticized the use of E. coli as an E. hafniae, and K. pneumoniae. The significance of these organisms in retail meats could indicator of food safety for dehydrated, frozen, and refrigerated foods, because they found that not be determined. Hence, our object was to E. coli does not survive well under such conditions. As a result, they recommended the use of the meat-processing chain to determine whether study Enterobacteriaceae at different stages of Enterobacteriaceae as indicators of food safety any associations could be made between bacterial isolates and stage in the meat-handling proc- (22, 23, 26). Greater concem for non-e. coli coliforms has also been proposed because of ess. their increasing involvement in diarrheal diseases (34). MATERIALS AND METHODS E. coli biotype I (IMViC [indole, methyl red, Sample collection. Samples were collected at Voges-Proskauer, citrate] ++-- and gas positive in EC broth at 450C) is generally recognized subsequent isolation and identification of Enterobac- packing plants and retail points of meat handling for as an indicator of direct or indirect fecal contamination of foods, and E. coli or colifonn bacteria collected on 18 occasions (7 times at plant A and 11 teriaceae. In the packing plant study, samples were in pasteurized milk and cooked foods times at generally plant B) from two federally inspected Canadian indicate postprocessing contamination packing plants, each operating beef and hog kill (16). floors and meat-processing departments. Samples Klebsiella pneumoniae is also a frequent but were taken to include (i) meat and nonmeat contact not exclusive inhabitant of the intestines of animals and humans (4, 5, 9, 19), has been impli- areas of the plants. For sampling, we used Rodac surfaces and (ii) the beef, hog, and processed-meat 867

2 868 STILES AND NG contact plates and violet red bile agar (VRBA). At the retail level, three stores that received supplies directly or indirectly from these packing plants were used to obtain samples that included: (i) vacuum-packaged beef primal cuts before they were opened at the retail store; (ii) vacuum-packaged beef trimnmings for use in ground beef manufacture; (iii) packaged pork loins (not vacuum packaged) before they were opened at the retail store; (iv) frozen and thawed pork sausage packaged by manufacturers; and (v) meats handled at the retail level, including beef steaks, pork chops, and ground beef. Sample preparation. Samples of integral meats were collected by a spray gun technique (7) with a 100-ml sterile 0.1% peptone water wash. Samples were collected aseptically in a graduated Erlenmeyer flask (approximately 20 ml per sampling point). Comminuted meat samples were refrigerated and taken to the laboratory for immediate analysis. Comminuted meats were sampled by blending 11 g of meat with 99 ml of sterile 0.1% peptone water for 5 min in a Colworth stomacher (model 400). Bacteriology. Appropriate dilutions were plated onto VRBA by the pour plate technique. All VRBA plates were overlaid with 5 ml of VRBA and incubated at 35 C for 18 to 24 h. After incubation, representative numbers of bile-precipitating and non-bile-precipitating colony types were randomly picked from the VRBA plates and purified by streaking first on MacConkey agar and then on nutrient agar. Biochemical tests were done on an isolated colony from the nutrient agar plates. Enterobacteriaceae were screened by Gram stain and oxidase test. The biochemical tests were an expanded Minitek (BBL, Becton-Dickinson and Co., Mississauga, Canada) identification system, described previously (31), involving: arabinose, citrate, dulcitol, H2S-indole, inositol, lysine, malonate, o-nitrophenyl-,f-d-galactopyranoside, ornithine, phenylalanine, raffinose, rhamnose, and urea disks. Isolates were also inoculated onto phenol red glucose and lactose (1%) broths, nitrate-motility agar, triple sugar iron agar, and methyl red-voges-proskauer medium. Supplementary tests were carried out as required. RESULTS Table 1 shows the incidence of Enterobacteriaceae on various packing-plant surfaces. The raw meat-handling surfaces, when in use, were the areas most heavily contaminated with Enterobacteriaceae. Rodac plates with >1,000 to too numerous to count Enterobacteriaceae per plate frequently involved the meat saw tables and stainless-steel work surfaces. In addition, the carcass-splitting saw and beef carcass shrouds were implicated on two occasions each. Cooked-meat contact surfaces were less contaminated with Enterobacteriaceae, and surfaces with no meat contact (walls, doors, etc.) were not apparent reservoirs of Enterobacteriaceae. A total of 2,343 Enterobacteriaceae isolates were identified, 1,372 from packing-plant samples and 971 from meats sampled at the retail APPL. ENVIRON. MICROBIOL. level. The distribution of Enterobacteriaceae types isolated at the two packing plants (Table 2) indicated that K. pneumoniae, E. coli I, and S. liquefaciens were the principal Enterobacteriaceae isolated, each accounting for 10 to greater than 20% of the isolates. C. freundii, Enterobacter agglomerans, and E. cloacae were of lesser importance, accounting for 5 to 12% of the isolates. Each of the other types accounted for less than 5% of the isolates. Some differences were observed in the order of the frequency with which different Enterobacteriaceae were isolated at each packing plant. E. coli I was the most common isolate in plant A, whereas K. pneumoniae was the most common at plant B. In general, the frequency of isolating Enterobacteriaceae at plant A was lower than that at plant B. A total of 86% of the plates with confluent growth were obtained at plant B, compared with only 14% at plant A. The sanitation practices in both plants virtually eliminated all Enterobacteriaceae from inanimate meat-handling surfaces (stainless steel, polyethylene cutting boards, and PVK or Neoprene rubber food belting). Occasional survivors included: E. aerogenes, K. ozaenae, K. pneumoniae, and S. Iiquefaciens. No E. coli I were isolated on equipment after cleanup. The data could also be examined on the basis of the Enterobacteriaceae associated with different areas and surfaces within the packing plants (Table 3). The number of isolates for each surface type was proportional to the number of organisms growing on the Rodac sampling plates. This finding confirmed the earlier indication that the raw-meat-handling surfaces were the most heavily contaminated with Enterobacteriaceae, cooked-meat contact surfaces were less contaminated, and surfaces with no meat contact were not apparent reservoirs of Enterobacteriaceae in the plants. On raw-meat contact surfaces, E. coli I, K. pneumoniae, and S. liquefaciens predominated, whereas E. agglomerans and K. pneumoniae were the main types identified on cooked-meat contact surfaces. Similar frequency distributions (Table 4) were calculated on the basis of the Enterobacteriaceae associated with the different product types (beef, pork, and cooked meat). K. pneumoniae, E. coli I, and C. freundii were the principal isolates from beef work surfaces; E. coli I, S. liquefaciens, and K. pneumoniae (in that order) were the principal isolates from pork; and S. liquefaciens, E. cloacae, K. pneumoniae, and E. agglomerans were the principal isolates from cooked-meat-handling surfaces. On cookedmeat work surfaces, E. coli I and C. freundii were relatively minor isolates (<10%). The incidence of E. coli I on equipment and

3 VOL. 41, 1981 ENTEROBACTERIACEAE IN MEAT HANDLING 869 TABLE 1. Levels of Enterobacteriaceae detected on VRBA Rodac plates from different surfaces in two packing plants Enterobacteriaceae colonies/rodac plate Surface None detectable c >500-TNTCa In contact with meats Beef area (15.4)b (41.4) (6.7) (13.5) (23.1) Hog area (36.3) (34.1) (8.1) (14.8) (6.7) Processed meats Raw (41.7) (29.2) (8.3) (10.4) (10.4) Cooked (62.5) (34.4) (3.1) (0.0) (0.0) With no product contact Beef area (71.7) (26.4) (0.0) (0.0) (1.9) Hog area (87.1) (10.0) (0.0) (1.4) (1.4) Processed meats Raw (94.4) (2.8) (2.8) (0.0) (0.0) Cooked (94.7) (5.3 (0.0) (0.0) (0.0) a Too numerous to count; very heavy to confluent growth. b Numbers in parentheses are percentages. TABLE 2. Frequency of isolation and distribution of Enterobacteriaceae types obtained from VRBA contact swabs from two packing plants No. of isolates Organism Plant A Plant B E. coli I 73(20.0)a 149(14.8) Other E. coli 6(1.6) 6(0.6) C. freundii 29(7.9) 127(12.6) E. aerogenes 9(2.5) 26(2.6) E. agglomerans 43(11.8) 65(6.5) E. cloacae 30(8.2) 88(8.8) E. hafniae 14(3.8) 39(3.9) K. ozaenae 12(3.3) 24(2.4) K. pneumoniae 57(15.6) 227(22.6) S. liquefaciens 51(13.9) 148(14.7) S. marcescens 3(0.8) 35(3.5) S. rubidaea 5(1.4) 8(0.8) Y. enterocolitica 6(1.6) 9(0.9) Other Enterobacteriaceae 28(7.6) 55(5.5) a Numbers in parentheses are percentages. meat-handling surfaces in use was determined for each meat type. In the beef-cutting areas of both plants, E. coli I was detected on meathandling surfaces in 14 of 40 samples. E. coli I was also detected on a hand wash sink on the beef kill floor of plant B in four of nine samples. In contrast, E. coli I was found on only 1 of 14 tests on walls in the beef kill area. In the beefcutting area for vacuum-packaged primal cuts, E. coli I was detected on meat-handling surfaces in 26 of 102 samplings. Similar tests for E. coli I in the pork prepa- TABLE 3. Distribution of Enterobacteriaceae among different surfaces, excluding meats, within the packing plants No. of isolates on surfaces in contact with: Organism Raw meat Cooked No meat meat E. coli I 153(20.0)a 4(3.0) 5(19.2) Other E. coli 7(0.9) 0(0.0) 1(3.9) C. freundii 71(9.3) 8(7.1) 4(15.4) E. aerogenes 8(1.0) 8(7.1) 0(0.0) E. agglomerans 54(7.0) 26(23.2) 3(11.5) E. cloacae 62(8.1) 10(8.9) 1(3.8) E. hafniae 17(2.2) 5(4.5) 0(0.0) K. ozaenae 21(2.7) 9(8.0) 0(0.0) K. pneumoniae 144(18.8) 17(15.2) 3(11.5) S. liquefaciens 138(18.0) 11(9.8) 2(7.7) S. marcescens 23(3.0) 0(0.0) 3(11.5) S. rubidaea 8(1.0) 0(0.0) 2(7.7) Y. enterocolitica 7(0.9) 5(4.5) 0(0.0) Other 54(7.0) 9(8.0) 2(7.7) Enterobacteriaceae a Numbers in parentheses are percentages. ration areas resulted in fewer E. coli I isolates on the kill floor equipment (5 positive of 34 tests) than on the beef kill floor. However, there was a marked increase in the frequency of E. coli I detected in the pork-cutting areas (51 positive of 130 tests). For example, stainless-steel chutes for pork cuts had 8 of 18 samplings that were E. coli I positive; plastic cutting boards, 7 of 20; rubber belting, 7 of 19; and stainless-steel work tables, 13 of 22. The emulsion preparation area for cooked-

4 870 STILES AND NG APPL. ENVIRON. MICROBIOL. TABLE 4. Distribution of Enterobacteriaceae isolated from surfaces and meats in packingplants associated with different meat types No. of isolates Organism Beef area Hog area Raw Processed meat Cooked E. coli I 86(14.4)a 100(24.9) 31(13.2) 5(3.6) Other E. coli 3(0.5) 7(1.8) 2(0.8) 0(0.0) C. freundii 87(14.6) 37(9.2) 22(9.4) 10(7.3) E. aerogenes 19(3.2) 3(0.8) 4(1.7) 9(6.5) E. agglomerans 43(7.2) 18(4.5) 17(7.3) 30(21.7) E. cloacae 33(5.6) 27(6.7) 46(19.7) 12(8.7) E. hafniae 32(5.5) 4(1.0) 12(5.1) 5(3.6) K. ozaenae 9(1.5) 15(3.7) 3(1.3) 9(6.5) K. pneumoniae 174(29.1) 54(13.4) 35(15.0) 21(15.2) S. liquefaciens 56(9.4) 83(20.6) 45(19.2) 15(10.9) S. marcescens 15(2.5) 20(5.0) 3(1.3) 0(0.0) S. rubidaea 5(0.8) 5(1.2) 3(1.3) 0(0.0) Y. enterocolitica 4(0.7) 2(0.5) 2(0.9) 7(5.1) Other Enterobacteriaceae 32(5.4) 27(6.7) 9(3.8) 15(10.9) a Numbers in parentheses are percentages. meat products gave many E. coli I-positive samplings (18 of 86 tests). The incidence of E. coli I in this area of plant B was higher than at plant A. However, in the cooked-product handling areas of the two plants, the incidence of E. coli I was markedly reduced (only 4 positive of 123 samplings), and all of these positive samples were from the wiener-handling equipment at plant A (4 of 28 tests). Throughout the two plants, surfaces with no meat contact (walls, doors, etc.) were tested on 102 occasions. Only one swab gave a positive E. coli I test. The incidence of Enterobacteriaceae in general on these surfaces was low (Table 1). The distribution of different types of Enterobacteriaceae (Table 5) on vacuum-packaged beef revealed S. liquefaciens as the predominant isolate (29.7%) and E. coli I as the next most important isolate. In contrast, for ground beef prepared from supplies that had been centrally prepared but not vacuum packaged, the predominating Enterobacteriaceae were E. agglomerans, S. liquefaciens, and E. coli I. Frozen pork sausage, packaged by the manufacturer, revealed that S. liquefaciens was the predominant Enterobacteriaceae; E. agglomerans and E. coli I were also major isolates. In contrast to the packing-plant samples, K. pneumoniae was an isolate of minor importance in samples collected at the retail level. Enterobacteriaceae were also isolated from retail store trim. Of 115 isolates, the principal organisms were E. agglomerans (43.5%), E. coli I (20.9%), and S. liquefaciens (13.9%). Samples from beef steaks and pork chops contained far fewer Enterobacteriaceae, and as a result only 56 and 45 isolates were identified from these samples, respectively. E. coli I was not detected among the beef steak isolates and represented only 2.2% of the isolates from pork chops. The predominant Enterobacteriaceae on both meat types were S. liquefaciens, which accounted for 32.1% of beef steak and 46.7% of pork chop isolates; E. agglomerans accounted for 30.4 and 20.0%, respectively. The only other isolate of any significance was E. hafniae, which accounted for 16.1% of the isolates from beef steak and 8.9% of those from pork chops. Whereas E. coli isolates were principally biotype I, with IMViC reaction ++--, the K. pneumoniae isolates had heterogeneous IMViC biotypes. The 284 K. pneumoniae of packing-plant origin included 59 (20.8%) IMViC type --++, 112 (39.4%) +-++, and 39 (13.7%) In comparison, the 118 K. pneumoniae isolates from meats at the retail level included 43 (36.4%) IMViC type --++, 47 (39.8%) +-++, and 10 (8.5%) This represents a high percentage of indole-positive strains. DISCUSSION The distribution of Enterobacteriaceae in the packing plants was primarily on meat-handling surfaces. Surfaces that did not normally come in contact with meats were not implicated as reservoirs of Enterobacteriaceae, and any buildup of Enterobacteriaceae during the day's operations appeared to be controlled by sanitation practices. Nonetheless, some work surfaces developed excessively high contaminating loads of Enterobacteriaceae during the work day, which would act as continual sources of contamination for products coming in contact with them. From these results, we conclude that the Enterobac-

5 VOL. 41, 1981 ENTEROBACTERIACEAE IN MEAT HANDLING 871 TABLE 5. Distribution of Enterobacteriaceae isolated from VRBA plates from different meats sampled before handling at the retail level Vacuum-packaged Ground beef from packaged supplies Packaged pork Organism beef trim and pri- sausage mal cuts High fat Low fat sausage E. coli I 22(15.2)a 48(20.9) 28(14.8) 31(15.0) Other E. coli 3(2.1) 2(0.9) 4(2.1) 2(1.0) C. freundii 9(6.2) 15(6.5) 10(5.3) 4(1.9) E. aerogenes 8(5.5) 10(4.3) 5(2.6) 6(2.9) E. agglomerans 14(9.6) 59(25.6) 59(31.2) 32(15.5) E. cloacae 9(6.2) 10(4.3) 15(7.9) 11(5.3) E. h.afniae 17(11.7) 12(5.2) 8(4.2) 4(1.9) K. ozaenae 5(3.5) 3(1.3) 8(4.2) 11(5.3) K. pneumoniae 1(0.7) 6(2.6) 9(4.8) 13(6.7) S. liquefaciens 43(29.7) 53(23.0) 37(19.6) 74(35.7) S. rubidaea 6(5.1) 3(1.3) 1(0.5) 1(0.5) Y. enterocolitica 5(3.5) 1(0.4) 0(0.0) 1(0.5) Other Enterobacteriaceae 3(2.1) 8(3.5) 5(2.6) 1(0.9) a Numbers in parentheses are percentages. teriaceae on the meat contact surfaces are a buildup from meats and their ongoing handling. As such, the stage at which the surfaces can be considered unsanitary is difficult to decide, especially since buildup can be attributed, in part, to growth of the bacteria on these surfaces. E. coli I, K. pneumoniae, and S. liquefaciens were generally the dominant Enterobacteriaceae in the raw-meat-handling areas of the packing plants. In contrast, the incidence of Enterobacteriaceae on cooked (processed)-meat work surfaces and products was dramatically reduced, as would be expected with the heat treatment. However, E. cloacae, S. liquefaciens, K. pneumoniae, and E. coli I were the predominant isolates from surfaces before processed products were cooked, whereas E. agglomerans, K. pneumoniae, and S. liquefaciens were the most common isolates from cooked-meat work surfaces used after the products had been cooked. These data indicated that K. pneumoniae and E. coli, both of which could be of fecal origin (4, 9), were predominant Enterobacteriaceae on meat-handling surfaces in the packing plants. At the retail level, even on meats before handling by the butchers, the predominating Enterobacteriaceae had changed to S. liquefaciens, E. agglomerans, and E. coli I. K. pneumoniae had become a relatively infrequent and minor isolate. This suggested a change in the Enterobacteriaceae surviving on meats, growth of paychrotrophic Enterobacteriaceae (5) on meats during the marketing process, or both. E. coli I was a frequent Enterobacteriaceae isolate on all samples except processed meats sampled after cooking. As a result, E. coli I in a product such as raw ground beef could have no meaning for sanitation at the retail level. S. liquefaciens was also present as a major Enterobacteriaceae isolate throughout the meat-handling process. However, E. agglomerans was a minor isolate at the packing plants, but it frequently became a major isolate, with S. liquefaciens, in retail samples. The Enterobacteriaceae with greatest indicator potential in these samples was K. pneumoniae. This organism was frequently isolated as a dominant Enterobacteriaceae in meats and on meat-handling surfaces in the packing plant, but it was only infrequently found on meats at the retail level. This observation was confirmed by data of Cox and Mercuri (11) in their study of Enterobacteriaceae on retail meats, whereas Newton et al. (27) found that 21.5% of the Enterobacteriaceae from hides and meat at packing plants were K. pneumoniae. These results suggest that this organism has a short survival time in meats, so that its detection at the retail level could indicate recent contamination, which in turn could indicate unsanitary and unhygienic handling. Studies on the distribution of Klebsiella spp. in a hospital kitchen indicated that 6 of 13 raw meats contained Klebsiella, but that heaviest contamination (>103/g) was observed on salads (8). Evidence of contamination in the kitchen was presented. The use of E. coli I as an indicator organism has always been predicated on its restricted occurrence in nature compared with the ubiquitous nature of E. aerogenes (5). Differentiation of K. pneumoniae and E. aerogenes poses even greater difficulties because of the similarity in their biochemical characteristics (2, 10, 11, 22). K. pneumoniae can be selectively enumerated, based on morphology on methyl violet agar medium (6); if contaminants on vegetables can be related to fecal origin (3), this organism might indeed be a more meaning-

6 872 STILES AND NG ful Enterobacteriaceae to use as an indicator for sanitation and hygiene of meats at the retail level than E. coli I. ACKNOWLEDGMENTS This work was supported by a research contract with Agriculture Canada. In addition, we acknowledge the cooperation of the meat-packing plants and retail stores that made this study possible. LITERATURE CITED 1. Ayres, J. C Microbiological implications in the handling, slaughtering, and dressing of meat animals. Adv. Food Res. 6: Bascomb, S., S. P. Lapage, W. R. Willcox, and M. A. Curtis Numerical classification of the tribe Klebsielleae. J. Gen. Microbiol. 66: Brown, C., and R. J. Seidler Potential pathogens in the environment: Klebsiella pneumoniae, a taxonomic and ecological enigma. Appl. Microbiol. 25: Buttiaux, R The value of the association Escherichieae-Group D Streptococci in the diagnosis of contamination in foods. J. Appl. Bacteriol. 22: Buttiaux, R., and D. A. A. Mossel The significance of various organisms of faecal origin in foods and drinking water. J. Appl. Bacteriol. 24: Campbell, L. M., and L. L. Roth Methyl violet: a selective agent for differentiation of Klebsiella pneumoniae from Enterobacter aerogenes and other gramnegative organisms. Appl. Microbiol. 30: Clark, D. S Improvement of spray gun method of estimating bacterial populations on surfaces. Can. J. Microbiol. 11: Cooke, E. M., T. Sazegare, A. S. Edmondson, J. C. Brayson, and D. Hall Klebsiella species in hospital food and kitchens: a source of organisms in the bowel of patients. J. Hyg. 84: Cowan, S. T Gram-negative facultatively anaerobic rods, p In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 10. Cowan, S. T., K. J. Steel, C. Shaw, and J. P. Duguid A classification of the Klebsiella group. J. Gen. Microbiol. 23: Cox, N. A., and A. J. Mercuri Comparison of two minikits (API and R-B) for identification of Enterobacteriaceae isolated from poultry and meat products. J. Food Prot. 41: Eickhoff, T. C., B. W. Steinhauer, and M. Finland The Klebsiella-Enterobacter-Serratia division: biochemical and serologic characteristics and susceptibility to antibiotics. Ann. Intern. Med. 65: Goepfert, J. M The aerobic plate count, coliform and Escherichia coli content of raw ground beef at the retail level. J. Milk Food Technol. 39: Hess, E Hygiene during meat production, p In B. C. Hobbs and J. H. B. Christian (ed.), The microbiological safety of food. Academic Press, Inc., New York. 15. Hill, W. M Microbiological standards-are they realistic? Food Prod. Dev. 9:38, 43, International Commission of Microbiological Speci- APPL. ENVIRON. MICROBIOL. fications for Foods Micro-organisms in foods. 1. Their significance and methods for enumeration. 2nd ed. University of Toronto Press, Toronto, Canada. 17. Knittel, M. D Occurrence of Klebsiella pneumoniae in surface waters. Appl. Microbiol. 29: Mackey, B. M., C. M. Derrick, and J. A. Thomas The recovery of sublethally injured Escherichia coli from frozen meat. J. Appl. Bacteriol. 48: Matsen, J. M., J. A. Spindler, and R. 0. Blosser Characterization of Klebsiella isolates from natural receiving waters and comparison with human isolates. Appl. Microbiol. 28: Miskimin, D. K., K. A. Berkowitz, M. Solberg, W. E. Riha, Jr., W. C. Franke, R. C. Buchanan, and V. O'Leary Relationships between indicator organisms and specific pathogens in potentially hazardous foods. J. Food Sci. 41: Mossel, D. A. A The presumptive enumeration of lactose negative as well as lactose positive Enterobacteriaceae in foods. Appl. Microbiol. 5: Mossel, D. A. A An ecological investigation on the usefulness of Eijkman's test as one of the methods used for the detection of faecal contamination of foods. J. Appl. Bacteriol. 25: Mossel, D. A. A., I. Eelderink, M. Koopmans, and F. van Rossem Influence of carbon source, bile salts and incubation temperature on recovery of Enterobacteriaceae from foods using MacConkey-type agars. J. Food Prot. 42: Mossel, D. A. A., W. H. J. Mengerink, and H. H. Scholts Use of a modified MacConkey agar medium for the selective growth and enumeration of Enterobacteriaceae. J. Bacteriol. 84: Mossel, D. A. A., M. Visser, and A. M. R. Cornelissen The examination of foods for Enterobacteriaceae using a test of the type generally adopted for the detection of salmonellae. J. Appl. Bacteriol. 26: Moussa, R. S., N. Keller, G. Curiat, and J. C. de Man Comparison of five media for the isolation of coliform organisms from dehydrated and deep frozen foods. J. Appl. Bacteriol. 36: Newton, K. G., J. C. L. Harrison, and K. M. Smith Coliforms from hides and meat. Appl. Environ. Microbiol. 33: Ng, L.-K., and M. E. Stiles Enterobacteriaceae in ground meats. Can. J. Microbiol. 24: Parr, L. W Coliform bacteria. Bacteriol. Rev. 3: Shooter, R. A., E. M. Cooke, S. O'Farrell, K. A. Bettelheim, M. E. Chandler, and F. M. Bushrod The isolation of Escherichia coli from a poultry packing station and an abattoir. J. Hyg. 73: Stiles, M. E., and L.-K. Ng Biochemical characteristics and identification of Enterobacteriaceae isolated from meats. Appl. Environ. Microbiol. 41: Stiles, M. E., and L.-K. Ng Estimation of Escherichia coli in raw ground beef. Appl Environ. Microbiol. 40: Thomas, S. B., R. G. Druce, and K. Elson An ecological study of the coli-aerogenes bacteria ofsurface soil. J. Appl. Bacteriol. 23: Twedt, R. M., and B. K. Boutin Potential public health significance of non-escherichia coli coliforms in food. J. Food Prot. 42: