One

-

~

Enteric Indicator Organisms in Foods

Introduction Faad micrabialagists have lang been faced with the prablem af detecting faad that has been handled unhygienically and cantaminated with pathagens af enteric arigin such as Salmonella Typhi and ather salmanellae, shigellae and pathagenic vibrias. Early in this century, specific testing far these arganisms in faad was impassible far practical purpases and it remains difficult far mast labarataries taday. Methads far the detectian af same faadbarne enteric pathagens, such as the shigellae, pathagenic Escherichia coli, and enteric viruses can still be unreliable, camplex, and unsuitable far rautine use. When present, pathagens are aften unevenly distributed in faad, accur in very small numbers, and are grassly autnumbered by ather bacteria, many af which have characteristics in camman with the pathagens (42, 126). Faad micrabialagists are also. faced with prablems in determining whether raw materials have been handled carrectly befare caming under their cantral, and assessing whether the pracessing and the cades af hygienic practice that they are manitaring have been effective in praducing faad af acceptable micrabialagical quality. In develaping testing pragrams to. abtain this infarmatian an apprapriate balance must be attained between extensive testing af a small number af samples, and applying mare simple tests to. a larger number af samples, within the limits impased by the resaurces available. It is aften impractical to. test faad far a wide range af pathagenic micraarganisms and taxic micrabial praducts. Therefare micrabialagists have turned to. the use af bacterial graups ar species which are relatively easily and ecanamically enumerated, a~ase presence in fa~e numbers indicateB-defieiencies iD.raw matmials, pracessing, handling.QL.starage

11

af easily detectable arganisms present in large numbers in the gut af man and animals, enteric pathagens wauld prabably also. be absent. The indicatar chasen ariginally was E. coli and a little later the whale graup af bacteria knawn as califarms came to. be used. This principle was quickly adapted to. the testing af faad, althaugh the tests and their interpretatian must be different in many respects. The histary af indicatar tests has been described by Massel (122). A variety af indicatar tests with many different functians is applied to. faads (179). The graups and species af enteric indicatar bacteria discussed in this chapter are the califarms and E. coli, the Enterabacteriaceae, and the enteracacci. Tests far the first two. indicatars mentianed abave are amang the mast widely used. The principal abjectives af tests far these enteric indicatars in faad fall in!,9 two. general categaries. Their rale may be~ to. indicate patential faecal cantaminatian ar the passible presence af pathagens, ar to. indicate past-pracessing cantaminatian af faads that have been pracessed to. destray vegetative cells af pathagens such as salmanellae (e.g. pasteurised faads). Micraarganisms that are used far these two. purpases are referred to. as index and indicatar organisms, respectively, by same authars (125, 126). Several ather micraarganisms assaciated with the gut af man and animals have been suggested as indicatars af faecal cantaminatian (109). Because af the resistance af its spares, Clostridium perfringens has been used by public health autharities to. indicate distant saurces of faecal pallutian af water and shellfish (106). C. perfringens usually has quite a different significance far the faad micrabialagist, as discussed in Chapter 15. Caliphages (84, 97) and phages af Bacteroides fragilis (105) have been that --.!!light 1'111 ow-undesirable._c_antaminl!ti.Qll..-(}r suggested as markers af the possible presence af enteric viruses and ather enteric pathogens. micrabi~gro~th. These graups ar species are knawn as indicatar arganisms and tests far their Bacteriaphages are believed to. have physical presence are af value to. bath regulatary agencies characteristics and enviranmental persistence properties which are mare similar to. thase of and quality cantral micrabialagists. Thraugh the use af tests far indicatar arganisms, faad praducts enteric viruses than bacteria (173). Tests far the presence af the different enteric can be cleared in as little as 24-48 h, campared indicatar graups in faad have different limitatians with the several days needed far many pathagen tests. Same excellent reviews have discussed the and advantages. Therefare the test chosen and its rale and applicatian af tests far the presence af interpretatian depend an the informatian indicatar arganisms in faad (3, 11, 30, 42, 90, 125, required and the nature, histary and intended use 132, 179). af the faad. The enteric indicatar arganisms will be discussed 'here in relatian to. the standard Indicatar arganisms were first used by public health micrabialagists interested in the safety af methads for examinatian af faad published by water supplies. They reasaned that in the absence Standards Australia (SA) (4). Methads far the

$

.

165 ~.

....

,Ill II!

",! i! /,

Heather M. Craven, Michae/ J. Ey/es and Jo A. Davey \ detection of Enterobacteriaceae and enterococci have not been recommended by SA, but reference will be made to methods published by other authorities. Alternative methods not defined as standard, but in routine use in many food laboratories, will also be considered. There has been much argument over the usefulness and proper roles of tests for the various enteric indicator organisms (30). These arguments will be resolved only by a better understanding of the ecological principles on which such testing must be based, greater knowledge of the distribution and physiology of indicators and pathogens, and by the adoption of improved, standardised methodology. Recent research on the survival of bacteria in aquatic environments may render invalid many of the assumptions upon which tests for indicators and pathogens are based. There is evidence that enteric pathogens and indicators can exist in aquatic environments for very long periods in a state whereby they are viable but not culturable when traditional methods are used (74, 87, 158). These organisms may retain their virulence in the non-culturable state, and while they cannot be cultured in artificial media, may cause disease when reintroduced into their host (141). These findings may lead to extensive changes in some areas of water and food microbiology when this phenomenon and its implications are understood more fully.

JI'

11:

Definitions

and

taxonomy

Coliforms and E. coli The coliforms are a diverse group of bacteria which are defined as much by the tests used for their isolation, as by taxonomic criteria. They are Gram negative, aerobic or facultatively anaerobic, non-sporeforming rods which ferment lactose with acid and gas production within 48 h at temperatures between 30°C and 37°C; depending on the test. The bacteria detected by coliform tests are members of several genera within the family Enterobacteriaceae, including Escherichia, Enterobacter, Klebsiella and Citrobacter. Bacteria that do not belong to this family may also give positive results, depending on the method used (10). Many of the coliforms are inhabitants of the human or animal intestine. Pathogenic strains of E. coli are the most noted coliforms that are significant agents of foodborne disease. Another coliform, Enterobacter sakazakii, has been implicated in a severe form of neonatal meningitis. Infant formulae containing this organism have been implicated in outbreak and sporadic cases of

JI' I'

j;,

Ih

the disease (135). The other coliforms are usually regarded either as non-pathogenic saprophytes or as occasionally pathogenic, for example Klebsiella pneumoniae can cause severe pneumonia. The pathogenic strains of E. coli are discussed in Chapter 9. \ There are many methods for the detection of coliforms in foods, but most follow the same general principles. The sample is inoculated into a medium containing lactose, which is usually made selective by the addition of ~ile salts, other surface active agents, or dyes. Presumptive coliforms, i.e. organi~~s fermenting lactose under the prescribed conditions, are confirmed as coliforms by further tests. Presumptive coliforms may also be inoculated into a selective medium incubated at an elevated tempernture" usually between ~mrJ4h hOC9rganisms that ferment lactose under these conditions are termed faecal

~

~ermoto

er ant

coliforms

~

s~ . aecal coliforms may be tested to determine whether E. coli, the coliform most closely associated with faecal sources, is present. Various modifications to this general scheme are in use, e.g. some tests seek E. coli directly and do not provide a coliform count.

Enterobacteriaceae The family Enterobacteriaceae is a large biochemically and genetically related group of bacteria that is heterogeneous in ecology and pathogenicity. Members of the family are Gram negative, non-sporeforming bacilli which grow in the presence and absence of oxygen, ferment glucose, give a negative reaction in the oxidase test, reduce nitrates to nitrites, and grow on laboratory media containing bile salts. The family includes many bacteria that are found in the human or animal intestinal tract, as commensals or pathogens. It is divided into over 20 genera, including those containing coliforms mentioned above, Salmonella, Shigella and Yersinia which are discussed in later chapters, and Proteus which is a resident of the gut of warm-blooded animals. The family also contains Hafnia alvei. Some strains of this species have been recognised as a cause of diarrhoea (156). Among the members of the family are plant pathogens in the genus Erwinia, species of Edwardsiella that are pathogenic for fish, and the genus Serratia, which is found in a variety.of habitats including soil and vegetable matter. These examples illustrate the diverse range of habitats and characteristics of the members of the family. The t~xonomy and properties of the Enterobacteriaceae are discussed in detail elsewhere (19, 52).

166 -=-

I

j.

Enteric Indicator Org

')ds

~ Enterococci The indicator organisms termed enterococci are contained within the genus Enterococcus. This includes the species E. faecal is, E. faecium, E. avium, E. gallinarum, E. durans, and others (163). Some species were formerly referred to as faecal stre tococci, based on older and now re un ant c asslfication systems. The bacteria in the genus Enterococcus are Gram positive cocci occurring in pairs or chains in hqUld cultures. l'hey are catalase negatlve, nonsporing, facultatively anaerobic, attack sugars by . fermentation, and most species~ are non-motile. Enterococci can be distinguished from streptococci by a few simple physiological tests (i63). The former grow at lOoC and 45°C, in 6.5% NaCI, and at pH 9.6. The major habitat of most species is the gut of humans and/or animals, where they are part of the normal microflora. They react with Lancefield group D antisera and usually survive heating at 60°C for 30 min (20).

The enterococci are hardy and equipped for ~

c

growth outside the gut. ,E. faecwm and some E. faecalis strains grow at 4°C. Growth of E. faecium at 1°C has been reported. Lactococci grow at lOoC but not at 45°C (163). Enterococci are potential D_athogens; they can cause~ urinary tract infections and subacute bacterial endo~tisin humans. There has been considerabl~ntroversy over whether or not these organisms cause food poisoning (178). It appears that if they can cause food poisoning, then the conditions required for the illness to occur are encountered rarely, or enteropathogenic strains are unusual. Enterococci in food can form clinically significant amoun s 0 some amines including histamine and tyramine. Enterococci can also acquire genes that confer resistance to some importa~ibiotics used to treat human disease. These strains have become a cause of increased mortality in hospital acquired infections. The use of antibiotics in animal production as growth promoters is now recognised as a contributing factor in the emergence of antibiotic resistant enterococci in the food supply, including strains that are resistant to vancomycin (69). However, at this time vancomycin resistant enterococci from food have not been implicated in human infections, possibly because strains from animal sources are less virulent. The discovery of antibiotic resistant enterococci in the food supply has led to a new interest in the incidence and safety of these organisms in foods, particularly in Europe, where enterococci may be used as starters in traditional cheese.

Significance of enteric indi~ organisms in foods

'~ Coliforms Coliforms were originally used as indicators because their habitat includes the gut of man and other animals, they are easy to detect and they have characteristics broadly similar to those of pathogenic members of the Enterobacteriaceae. Coliforms, particularly E. coli, are consistently present in the faeces of man and other animals, although the numbers and types vary according to many factors such as the type of animal, its age, physiological condition and food supply. Some coliforms are not associated with enteric sources, occurring naturally in foods and environments from which pathogens are consistently absent. Coliforms can survive and grow in environments associated with food processing in which enteric pathogens die or are overgrown. Thus the role of coliform counts in assessments of the microbiological quality of foods is quite different to that originally envisaged for coliform counts in water. Their presence in food certainly does not imply faecal contamination. In foods which normally receive a process sufficient to kill all Enterobacteriaceae present, coliform counts can provide a measure of the adequacy of processing and general sanitation. A high coliform count is evidence of underprocessing or unsatisfactory post-process contamination, although a low count does not necessarily mean that these defects have not occurred. Coliforms in these products may have a variety of sources. Coliforms can persist and grow on improperly cleaned equipment. They can establish themselves as part of the resident flora of food processing factories, where they can be difficult to eliminate. Coliforms may also enter a product after processing from the hands or garments of operators, from raw materials if ~ross contamination occurs, or from sl\ndry other ~ sources such as i~s. The best example of the use of coliforms as an indicator of sanitation efficiency is in the dairy industry. Historically, coliform tests have been used extensively to monitor the efficiency of measures taken to minimise bacterial recontamination of milk and dairy products after pasteurisation (91, 179). Coliform tests are still used to monitor sanitation but now are of less practical sigRificance as a result of improved manufacturing practices. Tests for other Gram 167

~~

..

i

'1 ,

Heather M. Craven, Micb -0

1:

detection of Enty have not been :re:

will be mar authoritiec standar labor

@ f!'

and Jo A. Davey

r~

..rotrophic ssessment shelf life

~

such

as

Ams occurs in most Ar'Cethe coliform count of -" ,,,me or location remote from J 'bear little relation to the - .nygiene under which the food was .. Psychrotrophic coliforms can grow in """,,we/foods stored at temperatures that are in accordance with good hygienic practice (136). Coliforms die in acid foods. Therefore a low coliform count in an acidic food indicates that hygiene was satisfactory only if the analysis was performed soon after production. The history of a product must be assessed carefully when the significance of coliform counts is being considered. In situations in which coliform counts are valid indicators, the presence of abnormally large numbers of coliforms does not necessarily indicate an immediate health hazard. It gives a warning that unsatisfactory handling or processing might have occurred and that contamination causing loss of quality, spoilage, or possibly a health ha:zard might also have taken place. The history of the product must be examined closely before the precise nature of the problem can be determined and the fault rectified. In many foods, particularly raw foods or dishes containing raw foods, coliforms have little sanitary significance sil).ce their presence need not indicate unsatisfactory production, handling or storage conditions. Coliforms are part of the normal microflora of cereal crops and a variety of fresh vegetables such as celery, lettuce and carrots (175, 182). In one survey involving a large number of samples, celery containing coliforms was consistently negative for salmonellae (170). In another survey of a variety of raw uncut and cut vegetables, coliforms were found at levels up to 104 cfu/g. E. coli was not isolated (64). Similarly, in 27 lettuce samples with more than 104 cfu/g Enterobacteriacae present, no Salmonella, Shigella, Campylobacter, E. coli 0157:H7, Vibrio cholerae, Listeria monocytogenes and E. coli were detected in these or in 124 Enterobacteriaceae negative lettuces (104). Some plant pathogenic species of Erwinia have no sanitary significance yet yield positive re,jults in coliform tests (90). Some coliforms, not necessarily recently derived from faeces, can be isolated from soil where they are presumed to grow, although they do not form a dominant part of the soil microflora (183). In these cases it may be more appropriate to look for abnormally high levels of coliforms rather than assess their presence in a food as a deficiency in food handling.

r

~ 11

~ I

168

The presence of coliforms is unavoidable in raw meats and should not be used as an indication of poor hygiene (136, 168, 184). Coliforms on raw meats are derived from both the animal and the processing plant. Psychrotrophic coliforms of no public health significance are widely distributed in meatworks (136, 168). Total psychrotroph counts have been proposed as a better means of measuring post-slaughter hygiene than tests which enumerate coliforms (102). Where decontamination processes are applied to carcases (e.g. lactic acid treatment), numbers of coliforms and other Gram negative bacteria will be selectively reduced compared with Gram positive bacteria. This will extend the shelf life of the meat but may permit the growth of pathogens which survive decontamination or recontamin~te the surface of the meat during subsequent handling (185, 186). The significance of traditional indicator tests needs to be considered carefully where new steps are introduced to control microorganisms in food processing. In these circumstances, alternative indicators may need to be found or tests for specific pathogens which survive the process may need to be introduced. Coliforms form part of the natural flora of some fermented foods (e.g. brined olives) and may take part in the fermentation process. Enterobacteriaceae In the opinion of many microbiologists, tests for the presence of Enterobacteriaceae in foods processed for safety do much the same job as coliform counts, but do it far better. Foods processed for safety are those that are processed in such a way that non-sporeforming pathogens will be destroyed. Pasteurised or cooked foods such as dried baby formula or cooked smallgood sausage are examples. The general principles which form the basis for the use of coliform tests, as discussed above, also apply to Enterobacteriaceae tests. Enterobacteriaceae and coliform counts differ slightly in their interpretation and significance. The family Enterobacteriaceae includes important pathogens (e.g. Salmonella), in addition to indicators of faecal contamination (e.g. E. coli), and genera that are not associated with faecal sources and have no direct public health significance (e.g. Erwinia). The detection of Enterobacteriaceae in foods processed for safety demonstrates the presence of a group of organisms that can be present only because of undesirable post process contamination or inadequate processing, and which includes several dangerous pathogens. Their presence in excessive numbers does not indicate faecal contamination and does not imply the presence of pathogens (30).

1

Enteric Indicator Organisms in Foods

"

Testsfor Enterobacteriaceae cannot completely replacedirect testing for pathogens. The use of Enterobacteriaceae tests has becomepopular in Europe, but not in the USA or Australia. Industries with substantial experience in using and interpreting coliform counts have, for goodreasons, been reluctant to abandon much of this accumulated knowledge by adopting new testing procedures. The proponents of Enterobacteriaceae tests cite several advantages of these procedures. Most pathogenic members of the Enterobacteriaceae found in foods do not ferment lactose, or ferment it slowly or without the production of gas (e.g. Salmonella spp., Shigella spp., some pathogenic E. coli). These organisms are not detected by coliformtests but their presence has substantially more direct public health significance than the presence of coliforms. Some foods may have predominantly lactose-negative populations of Enterobacteriaceae, including the pathogens, and thus coliform counts will give a falsely reassuring indication of the conditions of hygiene under which the food has been prepared. An example was provided by an outbreak of foodborne disease caused by enteropathogenic E. coli in cheese. The organism responsible (E. coli 0124) was a slow lactose fermenter and thus could not be detected by coliform tests. However, it was detected by Enterobacteriaceae tests. Samples of incriminated cheese showed coliform counts as low as 103 cfu/g, but E. coli 0124 was present at levels of 105-107 cfu/g (124). Another example is the internal contamination of eggs with salmonellae due to infection of the hen. An infected egg can contaminate a batch of egg pulp with salmonellae during processing, without the occurrence of unduly high levels of E. coli or other indicators. Enterobacteriaceae tests have good sensitivity because there are usually greater numbers of this considerably larger group present in foodthan coliforms or E. coli. Enterobacteriaceae detection methods are not unusual or difficult to use. Many of the media used are similar to media used in isolating coliforms, but with the addition Qfelllcose~ methods may be adapted to allow differentiation of the isolates, giving an indication

\I There of the source and significance of contamination. are valid arguments against the use of such ~I since a taxonomically ill-defined group as the coliforms, the variability in the bacterial groups

I detected by different coliform tests may make L interpretation of results difficult. Statistical aspects of sampling and testing for Enterobacteriaceae have been discussed at length by Mossel and his associates (42, 43, 126). Mter consideration of the ratios of Salmonella to

Enterobacteriaceae in foods processed for safety, in conjunction with various other factors, they concluded that examination of certain foods for Enterobacteriaceae can theoretically confer the same degree of consumer protection as Salmonella testing at less cost to the industry. However experience suggests that such an objective is not easily achievable in practice (30). E. coli E. coli is considered to be a more specific indicator of potential contamination with pathogens of faecal origin than coliforms or Enterobacteriaceae. More than four decades ago Buttiaux and Mossel (25) specified the following properties for bacteria occurring in the intestine if they are to be suitable for use as faecal indicators: a. the bacteria selected should occur natura~ly only in intestinal environments; b. they should occur in high numbers in faeces so that they are still detectable even at high dilutions; c. they must have a high resistance to the extraenteral environment, the pollution of which is to be measured; and d. reliable methods suitable for detecting small numbers of the organisms must be available. These criteria are still a useful starting point in considering the validity and usefulness of E. coli as a faecal indicator. No species or group of bacteria, including E. coli, fulfils these conditions perfectly, but none is superior to E. coli in most circumstances. E. coli is well adapted to life in the intestine and its major reservoir is the gut of humans and other warm blooded animals. One study found E. coli in the faeces of 92.9% of humans, with an average count of about 109cfu/g (dry weight) (55). It does not usually persist for long in environments other than the intestine, e.g. it is gradually eliminated from soil not subject to continuous faecal pollution (183). Thus the presence of E. coli in foods is generally believed to indicate recent pollution of faecal origin. Recent pollution of faecal origin may be described as contamination directly by faeces or indirectly by faecally contaminated material or by substrates which have allowed growth of bacteria of faecal origin. Pathogens and E. coli are presumed to behave similarly out~ide the gut, since the growth and survival characteristics of E. coli are broadly comparable to those of many pathogenic Enterobacteriaceae species. E. coli is capable of growth in food, and on inadequately cleaned surfaces associatea with food processing, under conditions similar to those that permit growth of the pathogens. It can also become established in food processing plants. 169

Heather M. Craven, Michael J. Eyles and Jo A. Davey

E. coli is a useful indicator of potentially hazardous contamination of raw foods. It does not appear to be a natural inhabitant of raw foods derived from plants, and it has been used as an indicator of faecal contamination of some uncooked meats (168, 179). E. coli tests are widely used in the meat industry in the United States and Australia for identifying post slaughter faecal contamination of meat carcases. E. coli is also the indicator

of choice

for

verifying

apple

cider

sanitation (101). E. coli counts are used to monitor raw bivalve shellfish for faecal contamination. The degree of contamination of oysters and other bivalves is closely related to the cleanliness of the water in which they are grown because of their manner of feeding. The significance of E. coli as an indicator in water samples is historically well established, thus E. coli counts generally provide a good indication of the cleanliness of bivalves. The compliance of levels of E. coli with regulatory standards may not however guarantee the absence of enteric viruses and pathogens (2). This was demonstrated in a gastroenteritis outbrea.k in Hawaii after ingestion of contaminated raw shellfish (79). Furthermore, absence of E. coli does not always indicate the absence of pathogenic vibrios (180) or Campylobacter (196) in bivalve molluscs. It has been proposed that environmental factors such as water temperature may play a role

'the relaive proportions of indicator organisms and pa ogens in this food group (23, 196). Therefore it may be necessary to apply direct tests for pathogens or tests for multiple indicators to gua ty (112). epuration .s sometimes used to reduce the possibl Ity of the presence of pathogens in shellfish, especially where the risk of faecal contaminat~on of water is high. However, the efficiency of this process varies with temperature of application. It also appears to be more effective for some microorganisms than others; e.g. it has been demonstrated that E. coli are more quickly eliminated than clostridial spores and these in turn more quickly than enterococci (110). Coliphages have been suggested as a better indicator than E. coli for depuration processes (111) but they may not necessarily reflect the elimination of viruses (145). Molecular methods have been applied for the detection of hepatitis and enterovirus in cockle and mussel samples. More samples were positive with these techniques than with cultural techniques for viruses, suggesting that viral particles in shellfish are not always infectious. A poor correlation was also found between bacterial and coliphage indicators and infectious viruses (13). The validity of E. coli as a faecal indicator has been questioned on the basis that E. coli is not 170

~

sufficiently hardy and may be eliminated from foods or environments in which pathogens survive. Direct comparisons have not been plentiful, but there is evidence that E. coli can be less resistant than some pathogenic Enterobacteriaceae to processes used to destroy microorganisms in food (37, 120, 181). E. coli may also disappear faster than some pathogens from refrigerated, frozen or dried foods. The results of comparative studies depend in part on the strains examined. Different strains of E. coli and the enteric pathogens vary significantly in their survival and growth characteristics. For example, there is no significant difference between the resistance of E. coli and salmonellae to spray drying when the groups are treated as a whole, but there are significant differences between strains in 'each group (118). Some salmonellae survive better than some E. coli, and vice versa. Similar variations have been observed in studies with other food processing procedures (120). It has also been suggested that E. coli cannot be assumed to indicate faecal contamination (30). E. coli can become established in environments in which there is no evidence of the presence of pathogens. There is very little evidence showing a relationship between the presence of E. coli and pathogens in factory environments (30). Thus, although E. coli is the best indicator of recent faecal pollution and a risk is implied by its presence, the absence of E. coli from a pro(hlct d~ not ensure the absence o.LID.teric.patllogens..:.... Conversely, its ~e may not always signify faecal contamination. --Faecal coli orm tests are wid 1 used as a a~Thetestisintermediate between coliform and E. coli counts in time taken, cost and specificity. Faecal coliforms can ferment lactose at a temperature between 44°C and 45.5°C, depending on the test. Most strains of E. coli produce gas from lactose at temperatures in this range, while the majority of other coliforms do not. Thus E. coli is usually predominant among faecal coliforms (193). Several authors have suggested that elevated temperature tests are a better means than biochemical differentiation of separating faecal and non-faecal types (67). Coliforms recently isolated from faecal sources tend to have higher temperature ranges for growth (24). Non-enteric, psychrotrophic coliforms common in some environments give negative results in elevated temperature tests. Although coliforms considered non-faecal in origin can give positive results in faecal coliform tests, the method is often useful as a monitoring procedure. The tests can be made more specific for ~E._cQliby including an indole test in parallel with the elevated temp~~ature test. Faecal colifor~-

--

---

- -

- --

-

Enteric Indicator Organisms in Foods

tests should be validated for each application. Species other than E. coli that give positive results in faecal coliform tests occur in large numbers in some foods. Some of these organisms can also produce indole at elevated temperatures. Faecal coliform counts are of questionable significancein such circumstances. If unexpectedly high faecal coliform counts are obtained, the isolates should be characterised further. A similar approach has been suggested for Enterobacteriaceae counts. It was once considered that an incubation temperature of 42°C would permit Enterobacteriaceae strains of enteric origin to grow while suppressing psychrotrophic strains from non-enteric sources (131), but recent studies with raw poultry have shown that Hafnia alvei, Enterobacter cloacae, Klebsiella ozaenae and Citrobacterfreundii can grow at both 6 and 42°C (198). Thus, elevated temperature tests for faecal Enterobacteriaceae have little value in this context. Tests for E. coli were found to be more applicable as their numbers did not change significantly over six days at 6°C (up to the time of spoilage).

a

Enterococci The enterococci have been used as indicator organisms for many years by water microbiologists. Many of the procedures used to detect these organisms were developed originally for use with water samples. Both enterococci and the broader group once termed faecal streptococci have been used as indicators of unhygienic preparation and handling of food. The chief source of most species of enterococci is the animal intestine. They are consistently present in human and animal faeces. Certain species are associated with certain animal hosts. E. faecalis and E. faecium inhabit the gut of humans and many other animals. They are the dominant enterococci in human faeces. E. gallinarum and E. avium occur in poultry. The related species Streptococcus bovis and Strep. equinis have been found in bovine and equine faeces. The type of animal from which faecal contamination originated cannot be established with certainty from the species of enterococci present, nevertheless some authors consider that these associations help to identify the source of contamination in certain circumstances (116, 155). The enterococci are more hardy than the Enterobacteriaceae. Therefore counts of these cocci can be useful indicators of the efficiency of plant sanitation and hygiene of production of foods in which Enterobacteriaceae may die, e.g. frozen or acidic foods. They may also be useful with foods in which other indicator bacteria have been destroyed by heating or drying. Because they

are so robust, the elimination of significant numbers ofthese cocci froI1lfood havppn l1sed ~s an indication that the hardy hepatitis A virus will alBo have. been destroyed (125). For thIS reason . ' enterococcI have also been proposed as an indicator of the presence of L. monocytogenes conttmination of procesg-ed meat (90). The significance of enterococci in foods is often uncertain and the use of these organisms as indicators appears to be declining. They are rarely a part of microbiological criteria for foods (92, 102) and an expert group has stated that counts of enterococci should not be included in such criteria (179). Counts are useful for quality control purposes only after the significance and normal levels of these cocci in the food in question has been determined. The hardiness of enterococci is one cause of criticism of their use as indicators. They survive adverse conditions so well that their presence may have little significance in relation to any hazard from enteric pathogens. Even if organisms such as Salmonella and Shigf;lla are deposited at the same time as enterococci, they cannot survive in many instances where the cocci can (90). Enterococci are found in environments other than the animal intestine (38, 134). They are common inhabitants of plants and are part of the microflora of many insects. Enterococci form part of the microflora of many foods without necessarily indicating poor hygiene. Enterococci are found in many fermented foods, e.g. cheese and sausage, and may take part in the fermentation process (82) giving these foods their characteristic flavours. They are part of the normal flora of fresh meat as well as certain cured, dried and smoked meat products (168). In meat products which have received a severe heat process however, the presence of excessive numbers of enterococci in~cates unhygienic handling and/or faulty storage. Enterococci may survive the heat process given to some comminuted cured meat products, and even under ideal conditions those products may be recontaminated during subsequent slicing and packaging. Enterococci may also survive milk pasteurisation (65) and have not been widely a<:lopted as indicators of process hygiene in the dairy industry. In these cases the enterococci do not necessarily have any sanitary significance. Enterococci can become established in food processing plants-; growing in areas far removed from any source of faecal contamination. Enterococcus faecalis produces a bacteriocin which is inhibitory to Listaia and may~eful for~the control_of this organism in foods (162). 171

I I I