Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Selection of Indicator Organisms and Agents as Components of Microbiological Criteria Regulatory agencies and industrial quality assurance personnel regularly examine foods or ingredients for microorganisms or their metabolic prod- ucts that may indicate (1) the possible presence of a pathogen or harmful toxin, (2) the possibility that faulty practices occurred during production, processing, storage, and distribution, and/or (3) the suitability of a food or ingredient for a desired purpose. The indicator organisms and agents discussed in this chapter are divided into four categories: (1) those that assess numbers of microorganisms and/ or microbial activity, (2) indicators of potential human or fecal contam- ination or possible presence of pathogens, (3) indicators of post-heat pro- cessing contamination, and (4) metabolic products of pathogens that indicate a health hazard. The section on assessment of numbers of microorganisms and/or mi- crobial activity includes (1) the aerobic plate count, thermoduric, psy- chrotrophic, thermophilic, proteolytic, and lipolytic counts; the direct microscopic count; Howard mold count; rot fragment count; "machinery mold"; yeast and mold count; heat-resistant molds and thermophilic spore count; and (2) examination for metabolic products such as by organoleptic examination, dye or indicator reduction time, pH, trimethylamine (TMA), total volatile nitrogen (TVN), indole, ethanol, diacetyl, histamine, en- dotoxins (Limulus amoebocyte lysate test), extract release volume, and adenosine triphosphate. Indicators of potential human or fecal contamination or of possible presence of pathogens include staphylococci, Escherichia coli, fecal col- iforms, enterococci (fecal streptococci, group D streptococci), and Pseu- domonas aeruginosa. Indicators of post-heat processing contamination 104
SELECTION OF INDICATOR ORGANISMS AND AGENTS 105 include coliform bacteria and Enterobacteriaceae. Thermonuclease and U.V. light examination of grains are discussed in a section on metabolic products of pathogens that indicate a potential health hazard. The signif- icance of the presence of phosphatase in certain dairy products is also discussed. Metabolic products such as tested for by organoleptic evalu- ation, pH, TMA, TVN, indole, ethanol, and diacetyl are not commonly used as components of microbiological criteria as are microorganisms. They are included in this chapter because they result from microbial activity and are intended as an adjunct or substitute for microbiological testing for one or more reasons such as convenience, economy, or effec- tiveness. Each of these indicators (organisms, group of organisms, or agents) will be discussed in reference to (1) its relative importance, (2) status and limitations of method for detection or enumeration, and (3) suitability as part of a microbiological criterion. Recommended uses of these indicators as part of microbiological criteria for foods are given in Chapter 9. ASSESSMENT OF NUMBERS OF MICROORGANISMS AND/OR MICROBIAL ACTIVITY Estimating Numbers of Microorganisms Aerobic Plate Count The aerobic plate count (APC) is used as part of a microbiological standard for raw and pasteurized milk and many other dairy products, and for shellfish at the wholesale level (see Chapters 8 and 91. It is also used (1) to monitor a large number of other foods for compliance with standards or guidelines set by various regulatory agencies, (2) to monitor foods for compliance with purchase specifications, and (3) to monitor adherence to good manufacturing practices. The test is based on the assumption that each microbial cell, pair of cells, chain of cells, or clump or cluster of cells present in an analytical unit (which is mixed with agar containing appropriate nutrients) forms visible, separate colonies when incubated for a sufficient period of time in an aerobic atmosphere and at a temperature suitable for growth. While the APC is the most popular method for estimating the number of viable microorganisms, it is often misused. Contrary to some assump- tions, the APC does not measure the size of the total bacterial population in a sample of food. Instead it measures only that fraction of the microbial flora that is able to produce colonies in the medium used under the pre- vailing conditions of plate incubation. Alterations in medium or environ- mental conditions change the fraction of the organisms that can grow.
106 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA Hence, rigid adherence to standardized test conditions is required if the APC is to be part of a microbiological criterion. The value of total counts as indexes of sanitary quality, organoleptic quality, safety, and utility of foods was reviewed by Silliker (19631. The APC of refrigerated perishable foods such as milk, meat, poultry, and fish may reflect conditions such as the microbial content of the raw ma- terials and ingredients, the effectiveness of the processing procedureks), the sanitary condition of equipment and utensils, and the time-temperature profile of storage and distribution. Specific causes of unexpected high counts can be identified by examination of samples at critical control points and by plant inspection. However, even perishable foods prepared from wholesome raw materials and properly processed, packaged, and stored develop high counts, lose quality, and ultimately spoil if held for long periods. Under these conditions, the APC reflects not sanitary quality but continued growth of psychrotrophic bacteria that were already present in the food. Thus, the usefulness of the APC depends to a great extent on the point in the process or distribution at which the sample was taken. The APC of shelf-stable foods reflects the same conditions as listed for perishable foods except that certain attributes of the shelf-stable food (aw, pH, heat treatment, frozen state) prevent further growth of contaminating bacteria. In the use or interpretation of the APC, additional factors need to be considered: (1) the APC measures only living microbial cells. Pro- cessing procedures, for example a heat treatment, can mask high-count raw materials or insanitary conditions. In addition, continued storage of a food in the frozen or dried state or at low pH values results in death and thus in decreases in count; (2) the APC is of little value in assessing the existing organoleptic quality of foods. Perceptible changes in quality characteristics of foods because of microbial activity do not occur until very high viable counts (106-107 per g or ml) have developed. Bacteri- ological analysis will then only confirm the results of an organoleptic examination; (3) the APC does not differentiate types of bacteria. Thus, even if numbers are approximately the same, biochemical activities leading to changes in a food may be different because of differences in microbial types or in composition of the food. In addition, high counts resulting from microbial growth are more likely to cause defects in foods than are similar levels resulting from massive contamination. For some foods, the APC can be used successfully to estimate potential shelf-life. For example, a standard plate count (SPC) on freshly pasteurized milk that was subsequently held at 7°C (44.6°F) for 5 days can provide an estimate of the potential shelf-life of this food (APHA, 19781. This application of the APC requires an intimate knowledge of the association between count and shelf-life of a food.
SELECTION OF INDICATOR ORGANISMS AND AGENTS 107 The APC may or may not be useful as a measure of the utility of an ingredient for a food. In many cases, it may be necessary to analyze the ingredient for the presence of specific organisms known to be of impor- tance for the stability of the final product. APCs have little if any rela- tionship to safety of foods. Also, APCs of fermented or retorted foods are of little value. Thermoduric, Psychrotrophic, Thermophilic, Proteolytic, and Lipolytic Counts Minor modifications of the APC can be applied to enumerate specific groups of microorganisms such as thermoduric, psychrotrophic, thermo- philic, proteolytic, or lipolytic bacteria. The thermoduric or laboratory pasteurization count (APHA, 1978) is used in the milk industry to check individual producer milk in order to identify samples with excessive num- bers of these organisms. Thermoduric organisms survive pasteurization of milk and thus may contribute to high SPCs on pasteurized milk. High thermoduric counts are closely associated with persistent improper clean- ing and sanitizing of equipment at the producer farm or at the processing plant. From a quality control point of view, psychrotrophic microorganisms are of great significance in refrigerated perishable foods because they grow, although not at optimum rate, at refrigeration temperatures. Some of these psychrotrophs can cause serious off-tastes, off-odors, or sliminess of foods when numbers are allowed to increase to 106-108 cells per g or ml. Although the initial number of psychrotrophs in a refrigerated perishable food may be low, it can increase to large numbers during storage, the time depending upon factors such as the initial level of con- tamination and the temperature. Shelf-life of such foods will be severely impaired when initial numbers are high, particularly when foods are held at marginal refrigeration temperatures. Unless consumed prior to this event, high microbial counts will definitely occur in perishable foods even when these foods are prepared from wholesome materials and are properly pro- cessed and stored. Psychrotrophic bacteria are often sources of heat-re- sistant proteolytic and lipolytic enzymes that may affect adversely the quality of foods during long-term storage after heat processing (Griffiths et al., 19811. Proteolytic and lipolytic microorganisms can be responsible for a variety of odor and flavor problems in foods. Some of the common psychrotrophic bacteria are strongly proteolytic and/or lipolytic and cause serious defects in dairy, meat, poultry, and seafood products when high counts (106 per g or ml or above) are reached during refrigerated storage. The proteolytic
108 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA count (Chapter 8, Table 8-4) is part of a microbiological criterion for butter and whipped butter. The proteolytic count and lipolytic count of butter can be helpful in drawing attention to unsatisfactory manufacturing practices. In addition to the APC (AOAC, 1980) and SPC (APHA, 1978), there are alternative procedures for determining viable bacterial counts such as the surface or spread plate method, plate loop, oval tube or boKle culture, drop plate, membrane filter, and spiral plate (AOAC, 1980; APHA, 1976, 1978, 19841. Reliable interpretation of the APC of a food depends upon an intimate knowledge of the expected microbial population at the point in a process or distribution at which the sample was collected. Higher than expected counts can serve as an alerting mechanism to investigate the potential causers) of this event. The two most widely used APC methods are the AOAC method (AOAC, 1980) and the standard plate count (SPC) method as described in Standard Methods for the Examination of Dairy Products (APHA, 19781. These reliable methods are used routinely to determine APC as part of many microbiological criteria. With due consideration of the precautions and limitations discussed in this section, the APC can be a useful component of microbiological criteria of many foods. Specific applications of the APC in microbiological criteria for foods are given in Chapter 9. Direct Microscopic Count The direct microscopic count (DMC) is used as a component of mi- crobiological criteria (standards) of raw (non-Grade A) milk, dried milks, liquid and frozen eggs, and dried eggs. The DMC gives an estimate of the total number of microorganisms, viable and nonviable, in a sample. The DMC is rapid, requires a minimum amount of equipment, and pro- vides information about the morphological characteristics of the cells. The DMC is also used to count the number of leucocytes in raw milk and thereby detect mastitic milk. Disadvantages of the method are that it is suitable only for foods containing relatively large numbers of microor- ganisms (105-106 per ml) and that the small quantity of sample (0.01 ml) examined limits precision. The lower limit of precision for the DMC is dictated by the Microscopic Factor and the number of fields examined. Separation of viable and nonviable cells by this method is not possible, and consequently counts may exceed corresponding agar plate counts many times. For these reasons the use of the DMC as part of microbiological criteria is very limited and applicable only when fairly high microbial populations (viable, nonviable, or both) are expected and information on
SELECTION OF INDICATOR ORGANISMS AND AGENTS 109 the microbial history of the product (such as dried milk) can be useful in quality control programs. Rapid standardized methods (AOAC, 1980; APHA, 1976, 1978, 1984) are available and are satisfactory for use in microbiological criteria. Other applications of the DMC use counting chambers such as the Petroff-Hauser slide or the haemocytometer. These counting chambers are used to determine populations of yeasts and other microorganisms in foods such as fruit juice concentrates. Microscopic MoldF Counts Microscopic mold counts are used to assess the soundness of raw hor- ticultural products and the sanitary conditions of processing lines. The Howard mold count and rot fragment count are used for regulatory pur- poses (see Chapter 9~. Howard Mold Count. The Howard mold count, one of the oldest microbiological criteria, is widely used to detect introduction of moldy, decomposed fruits and vegetables into processed foods. It is used to mon- itor both raw products and sorting and trimming operations. The test was originally developed for tomato products. It is also applied to foods such as apple butter, drupelet berries, citrus and pineapple juices, cranberry sauce, strawberries, pureed infant food, and garlic powder and other spices. Microbiological criteria (see APHA, 1976, 1984) vary with the type of food being evaluated. Rot Fragment Count. The rot fragment count is applied mainly to tomato products such as catsup and sauce. This test, a count of tomato cellular materials that exhibit mold filaments, is used mainly to evaluate whether or not a food was prepared from sound raw produce. Practical methods for these mold counts are applicable for microbiological criteria (AOAC, 1980; APHA, 1976, 19841. There is a sufficiently close rela- tionship between the Howard mold and rot fragment counts and the amount of decayed tissue present for the counts to serve as useful microbiological criteria for certain fruits and vegetable products and to be an index of manufacturing practices. However, processing conditions such as milling and homogenization can affect microscopic mold counts. "Machinery Mold." Geotrichum candidum is the predominant mold on tomato processing equipment and can be responsible for the slime that accumulates on this equipment. It has been referred to as "machinery mold." Geotrichum filaments have also been recovered from a variety of
110 EVALUATION OF THE ROLE OF MlCROBlOLOGICAL CRITERIA fruits and vegetables at different stages of processing. Careful cleaning of processing equipment results in a marked reduction in the Geotrichum count. Methods to determine G. candidum for purposes of microbiological criteria are adequate (AOAC, 1980; APHA, 1976, 1984~. G. candidum is used as a microbiological criterion to check sanitation of equipment in fruit and vegetable processing plants and in bottling works. Yeast and Mold Counts Yeast and mold counts are used as part of microbiological standards of various dairy products, for example cottage cheese (Chapter 8, Table 8- 4) and sugar (National Soft Drink Association, 19751. Yeasts and molds are widely distributed in the environment and can enter foods through inadequately sanitized equipment or as airborne con- taminants. Yeasts and molds frequently become predominant on foods when conditions for bacterial growth are less favorable, e.g., foods with low aw, low pH, high salt, or high sugar content. Therefore, they become a problem in fermented dairy products, fruits, fruit beverages and soft drinks. They manifest themselves by producing undesirable odors, flavors, appearance (discoloration), gas, and sediment. Satisfactory methods are available for yeast and mold counts that are applicable for purposes of microbiological criteria. These methods use either acidified agars or agar media with added antibiotics to inhibit bacterial growth (APHA, 1976, 1978, 19841. Methods for osmophilic yeasts are available (APHA, 1984~. Heat-resistant MoldEs A few molds such as Byssochlamys fulva and Aspergillus fisher) produce ascospores that are sufficiently heat resistant to survive the thermal pro- cesses applied to fruit and fruit products. Food processors sometimes set limits for these molds in purchase specifications for ingredients such as fruit concentrates and tapioca starch. Satisfactory methods for isolation and enumeration of these molds applicable for microbiological criteria are available (APHA, 1976, 19841. Thermophilic Spore Count Various types of sporeformers are used as part of the microbiological criteria, primarily specifications, by the canning industry to monitor the quality of ingredients such as sugar, starch, flour, spices, mushrooms, nonfat dry milk, and cereals that are intended for low-acid heat processed foods. National Food Processors Association criteria (NCA, 1968) for
SELECTION OF INDICATOR ORGANISMS AND AGENTS 111 sugar and starch to be used for this purpose include limits for total ther- mophilic spore count, flat sour spores, thermophilic anaerobic spores, and sulfide spoilage spores. Concern for thermophilic spores in these ingredients is related to their high heat resistance and their ability to cause defects in foods held at elevated temperatures because of inadequate cooling and/or storage at too high temperatures. Examination of equipment surfaces, product in process, or finished product is in some cases useful for location of foci of spore buildup. Methods to determine these spores (AOAC, 1980; APHA, 1976, 1984; NCA, 1968) are adequate for application in microbiological criteria. Microbiological criteria involving thermophilic spore counts have been useful as purchase specifications to specify quality of or check ingredients intended for use in low-acid heat processed foods (see sections of Chapter 9 that apply). Measuring Metabolic Products Metabolic products produced during growth of microorganisms have been used to estimate bacterial populations and to express quantitatively the effect of microbial activity on the quality of foods. Organoleptic Examination Organoleptic examinations of foods are used extensively for evaluation of quality attributes such as taste, odor, body and texture, color, and appearance. The food industry uses these examinations to classify certain foods into quality grades. For basic information about organoleptic eval- uations of foods, see Amerine et al. (1965) and Larmond (19774. Use of organoleptic examinations of foods to evaluate microbial activity is limited, as the effect of microbes on organoleptic attributes varies. Certain microorganisms act upon the constituents of a food and produce marked changes in perceptible characteristics of the food; others are rel- atively inert biochemically and produce little change. The effect of certain levels and/or types of microorganisms on perceptible characteristics in foods varies with differences in food composition and physical-chemical characteristics. For example, objectionable odors in a perishable food may result from extensive microbial activity; on the other hand, large microbial populations do not necessarily cause perceptible odor problems. In ad- dition, alterations in packaging and distribution methods can interfere with normal expected changes in quality characteristics of a perishable food (see Chapter 11. Microbial types on meat, poultry, and fish stored under refrigeration in vacuum packages or in modified gaseous atmospheres
112 EVALUATION OF TlIE ROLE OF MICROBIOLOGICAL CRITERIA differ considerably from those on products exposed to air. Nevertheless, organoleptic examinations of certain foods can be useful in evaluation of microbial activity. For example, organoleptic examina- tions are used to evaluate the amount of acid and/or aroma compounds produced by various lactic acid bacteria in cultured dairy products. The organoleptic evaluation of cheese for typical flavor and body and texture, to a large extent, reflects the activities of desirable bacteria. Organoleptic examination is also used with some foods to evaluate the activity of spoilage bacteria that may lead to quality deterioration and subsequent spoilage of the food. In these cases, an attempt is made to relate the organoleptic trait, usually odor, to the level of activity of spoilage bacteria. Although organoleptic examinations for this purpose are used for many commodities such as raw milk, meat, and poultry, they are used exten- sively for fish and other seafoods (see Chapter 91. In general, seafoods are more perishable than are other high-protein meat foods, because many seafoods contain relatively large amounts of nonprotein nitrogen com- pounds that are readily available to support microbial growth. Also, sea- foods harvested from cold waters contain microbes that are not as effectively inhibited by refrigeration. The information in Table 5-1 is an example of organoleptic changes in relation to chemical changes and bacterial counts of fish. In recent years, FDA has used organoleptic examination to determine the degree of decomposition of imported shrimp. The shrimp are placed in one of three classes: Class 1, passable: This class includes fishery products that range from very fresh to those that contain fishy or other odors characteristic of the product, but not definitely identifiable as odors of decomposition. Class 2, decomposed (slight but definite): This is the first stage of definitely identifiable decomposition. An odor is present that is not really intense, but is persistent and readily perceptible to th experienced ex aminer as that of decomposition. Class 3, decomposed (advanced): This class of products possesses a strong odor of decomposition that is persistent, distinct, and unmistakable. Limits of acceptability of a lot are based on the number of shrimp in a sample that are placed in the three classes (Anonymous, 1979~. In general, trained personnel could consistently place shrimp correctly in the appropriate class. The presence of objectionable odors in raw seafood usually indicates that extensive microbial activity has taken place. When such changes are perceptible, bacterial numbers exceed 106 per gram. Successful application of organoleptic examination of fish and other foods to determine the degree of decomposition by microbial and tissue
SELECTION OF INDICATOR ORGANISMS AND AGENTS 113 enzyme activity requires well-trained personnel. The organoleptic ex- amination of specific foods including milk, meat, and poultry, but par- ticularly raw fish and other seafoods, can be a most valuable tool to determine the degree of quality deterioration by microbial activity. If off- odors of microbial origin are perceptible, then counts are high and the product should be rejected. Dye or Indicator Reduction Time Attempts have been made to use oxidation-reduction indicators (dyes) for estimating the microbial quality of various foods such as milk, meat, poultry, seafoods, and vegetables (Bush, 19701. Present use of these dyes is restricted to the milk industry where they are used for classifying certain manufacturing milks into grades (APHA, 1978) and in rapid tests to estimate the potential shelf-life of freshly pasteurized milk (Broitman et al., 1958; Parmelee, 19741. The application of dye reduction tests to foods other than milk usually has not been successful because naturally present reducing enzymes and agents interfere with the dye reduction pattern. Comparisons of reduction times in milk with estimates of bacterial pop- ulations by the agar plate method and other tests have demonstrated that reduction times are, in general, inversely related to the initial bacterial content of the sample. Dye reduction tests provide only estimates of bacterial activity, as reduction times are influenced not only by bacterial numbers but also by the growth phase of the bacteria at time of testing and by the types present. With improvements in milk quality resulting in lower bacterial counts, reduction tests have become less applicable for quality evaluation of raw milk. Results of oxidation-reduction tests on pasteurized milk can be used by the processor as guidelines for estimation of the potential shelf-life of freshly pasteurized milk. Simple, practical dye reduction methods are available (APHA, 19784. pN The pH of some foods changes as a result of microbial activity; con- sequently, pH measurements have been applied to determine the degree of quality deterioration. The pH of fish and shrimp increases during storage because of the production of ammonia and amines by microbial and tissue enzyme activity. The pH of molluscs, on the other hand, decreases during spoilage. Microbial activity results in the production of basic substances in refrigerated ground meat stored under aerobic conditions (Shelef and Jay, 19701. However, changes in pH because of microbial activity vary with conditions such as nature of the food (available substrate for microbial
114 au 0 v: cat · _ · _ u, cd Cal cat ct .e s ct . ~ a' o to O cat ._ 04 ct v, of ~ .- ~ oo Vl Vl Cal ret 0 -1 0 0 ~ ~ ~ ~ 0 ~ 0 cacat ._._ ~ ° Ea ~ E a ~ ~ 00 , ~ ~04 ~ ~ - ° ~ vg - o 1 to Vl m - 00 of 1 to O ~ . E Y E g s c a s O _ 3 E-~ E 0 _ 3 3 E K o Cal C<, y u B 8, Y Y y c . a 1 ~ ~ ~ u ~ E u c ~ c D 3 y , ~ s . c~ ~ =^ ~ I ~ ~ _ ~ ~_ a' ° ~ ~ I ~ ~ ~r, c~ ct · _ c<' 0 - _
115 aid =- A 0 0 \~) O A A A Al.s:: rat C) - o hi: Cal ~ ~ o O D ~ ~ ~ O ~ ~ ~ . ~ ~ O ca D ~ :>, _ ~ .= O ~ Cal Cal m ca 0 Cal x ._ C) Cal - Ct o Ct o . . . Cal ED to o - 0 Ct CD x Ct D ~ O 4 - 0 O _ O m ,,, E~ ~ ~ r X C~ ca C) t4 O ~ ~ O ~ 2 O ~ > ~ y C ~ a ~ E oa a' ° ._ .. Ct ^> ~ .^ c E~ . . O 0 ~ 0.e a' C~ a' .= 0 00 V~ . r o 00 _4 V, _. . . o
116 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA activity), the type of microbial flora, and packaging conditions (gaseous atmosphere). Acidity as measured by pH has limited but important ap- plication in the monitoring of certain critical control points in the pro- duction of some foods. For example, voluntary guidelines were recommended (AMI, 1982) for the production of dry and semidry fermented sausages in which time-temperature relationships are given for reaching a meat mixture pH of 5.3 or less by the action of lactic acid producing bacteria. The goal of these voluntary guidelines is to provide an understanding of the critical control points in the fermentation stage of production of dry and semidry fermented sausages in order to minimize the opportunity for Staphylococcus aureus to develop. In the-production of cultured milk products and cheeses, measurement of pH of the inoculated milk is also of great importance as it monitors the rate of development of lactic acid bacteria. Lack of acid production (starter failure) may result in the buildup of S. aureus and in the presence of toxins in the cheese (see Chapter 9, Part A). Monitoring the pH of mayonnaise and salad dressings (Chapter 9, Part N) and of acidified canned foods (Chapter 9, Part K) is important in assuring safety of these foods. Reliable routine methods for pH measure- ment are available (APHA, 19841. Tr~methy~amine and Total Volatile Nitrogen Trimethylamine (TMA) and total volatile nitrogen (TVN) have been used for evaluation of the quality of fish and seafoods. Volatile amines, primarily ammonia, are produced in fish and other seafoods from amino acids by bacteria. TMA results from the reduction of trimethylamine oxide (TMAO) by bacteria particularly in gadoid fish such as cod, haddock, and hake. According to Montgomery et al. (1970), for example, 5 mg TMA- N/100 g and 30 mg TVN-N/100 g are limits for acceptability for shrimp in Australia and Japan. Stansby (1963) lists TVN values for different degrees of freshness of fish as follows: fresh fish, 12 mg or less; slight decomposition, but edible, 12-20 ma; borderline edible, 20-25 ma; badly decomposed and inedible, >25 mg TVN-N/100 g. These grading scales are at the present mainly applied for research purposes. Little detailed information is available on the TMA and TVN contents of seafoods in United States wholesale and retail markets. Information obtained with gadoid fish may not be applicable to other types of fish and other seafood. For example, TMA could not be detected in ice-stored shrimp until 8 days had elapsed, at which time the APC was approximately 108 per gram and the shrimp was judged of poor quality (Chang et al., 19831. During this storage period, TVN values decreased, probably because of the washing action of melting ice (Cobb et al., 19771. Cobb and coworkers (1977) also suggested that low TMA values in Gulf
SELECTION OF INDICATOR ORGANISMS AND AGENTS 117 shrimp could have resulted from lack of substrate (TMAO). In another study (Oberlender et al., 1983), TMA values of swordfish steaks stored in CO2-enriched atmospheres did not exceed recommended values for acceptability even when the steaks had spoiled. In this case, the lack of relationship between TMA content and sensory characteristics may have resulted from a shift in the composition of the microflora. Lactic acid bacteria were predominant on the fish stored under modified gaseous atmospheres, whereas gram-negative, aerobic rods usually predominated on samples exposed to air. Reliable methods are available to determine TMA and TVN (AOAC, 1980; Cobb et al., 19731. Quality deterioration and subsequent spoilage of fish and seafoods is a very complex process and is influenced by a great many variables, such as (1) fish species, (2) area of catch, (3) method of catch, (4) handling on board of vessel, and (5) processing techniques. For a limited number of marine species under certain conditions of handling, a relationship may exist between indicators, such as TMA and/or TVN, and microbial activity that influences acceptability of a product. If such indicators or criteria are applied, caution should be exercised not to include either other species for which this relationship has not yet been established or the same species when they are handled under different conditions. Because of these severe limitations, TMA and TVN by themselves should not be applied broadly in criteria to reflect microbial activity in fish and seafoods. lndfole Indole has been proposed as an indicator of decomposition for shrimp and oysters since the 1940s (Beacham, 1946; Duggan, 1948; King and Flynn, 19451. According to McClennan (1952) and Salwin (1964), indole is a good index for differentiating between acceptable and unacceptable shrimp. Indole is formed by the action of certain bacteria (such as Pro- teus spp., E. coli, and others) on tryptophan present in shrimp and oyster tissue. A study conducted by FDA (Duggan and Strasburger, 1946) reported a good correlation between indole concentration and sensory evaluation; the indole content increased as decomposition advanced. Fresh shrimp does not contain indole in excess of 1 ~g/100 g. The FDA established a defect action level of 25 ~g/100 g for indole in imported canned and cooked/frozen shrimp (FDA, 1981~. Shrimp at or exceeding this indole level is not permitted to enter the United States. This action resulted from concern with Salmonella, decomposition, and filth in frozen raw imported shrimp. Subsequent canning or cooking masks the odor of decomposition. Duggan and Strasburger (1946), however, reported that when shrinkage
1 18 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA due to moisture loss was taken into consideration, the indole level in cooked shrimp was nearly identical to that of the original raw shrimp. Chang et al. (1983) showed that indole levels of shrimp of acceptable quality remained stable during a 5-minute boil. High indole levels indicate decomposition, but shrimp of poor or un- acceptable quality may not necessarily contain indole (Chang et al., 1983; Staruszkiewicz, 19741. When shrimp was stored at 4°C (39°F) until it was severely decomposed, the indole level was lower than 25 ~g/100 g, the defect action level by FDA (Chang et al., 19831. The APC of this shrimp exceeded 108/g and the TVN value had reached 56 mg TVN-N/100 g. At higher storage temperatures (12° and 22°C/53° and 72°F), rapid production of indole took place. This suggests that mesophilic bacteria are more significant in the production of indole in shrimp than are psychrotrophic species. Hence, indole alone is not a suitable index of acceptability for all types of fresh or frozen shrimp, but when used in conjunction with other quality tests it can be of value in assessing prior high-temperature abuse. Methods for the determination of indole for microbiological criteria (AOAC, 1980; Cheuk and Finne, 1981) are satisfactory. Ethanol Ethanol can be produced by various bacteria from carbohydrates through glycolysis and/or the deamination and decarboxylation of amino acids such as alanine. Present knowledge (Crosgrove, 1978; Hollingworth and Throm, 1982, 1983) indicates that ethanol alone might be a useful index of decomposition of canned salmon. Hollingworth and Throm (1982) reported a high correlation between the ethanol content (as determined by gas chromatography) of four species of canned salmon and the sensory classification of decomposition. Tentative ranges suggested were: class 1 (passable), 0-24 ppm ethanol; class 2 (decomposed, slight but definite), 25-74 ppm ethanol; and class 3 (decomposed, advanced), 75 ppm ethanol and above. If this relationship is confirmed in collaborative studies, the ethanol level should have practical application as part of a HACCP system to monitor the precanning and postcanning quality of this salmon. It may be applicable to other seafoods if a similar relationship between sensory characteristics and ethanol can be established. The present analytical method for the ethanol determination, although relatively simple and rapid, requires sophisticated laboratory equipment. D,ace~l The diacetyl test was developed in the citrus industry to indicate mi- crobial activity in the early stages of multiple-stage, low-temperature evap
SELECTION OF INDICATOR ORGANISMS AND AGENTS 119 orators. The low oxygen tension, moderate temperature, and 18-20% sugar content of the juice during early stages of evaporation are ideal conditions for the rapid growth of lactic acid bacteria (Lactobacillus and Leucon- ostoc). These microorganisms produce diacetyl and acetylmethylcarbinol resulting in an undesirable buttermilk-like flavor in the finished concentrate (Fields, 19641. Several varieties of oranges (particularly mature Valencias) contain acetylmethylcarbinol as a normal constituent. By determining the diacetyl and acetylmethylcarbinol content of juice prepared in the labo- ratory from normal fruit, the quantities generated by microbial growth in the processed product can be estimated. The diacetyl test is a simple calorimetric procedure that requires approximately 30 minutes. Diacetyl can be used as a parameter to monitor the condition of the fruit and the sanitary condition of orange and apple processing (APHA, 1984; Fields, 1964; Hill et al., 1954; Hill and Wenzel, 19571. Histamine The formation of toxic levels of histamine in foods such as certain types of fish and cheeses and the limits for tuna set by FDA have been discussed in Chapter 4. The presence of toxic levels of histamine in fish such as tuna indicates extensive growth of bacteria such as Proteus and Klebsiella spp., which are capable of producing histamine from histidine. The prin- cipal reason for this growth is lack of postharvest cooling. An official fluorometric method (AOAC, 1980) is available to measure histamine and is applicable for use in microbiological criteria (see Chapter 4~. Problems in detecting this type of spoilage at the plant level include the facts that (1) odor and appearance do not reliably indicate this type of spoilage, (23 a broad range of spoilage exists among individual fish in a lot, and (3) the AOAC fluorometric method (AOAC, 1980) is a lengthy procedure not adaptable to the plant level. Recently, Lerke and coworkers (1983) reported a rapid screening method than can detect histamine in both raw and heat-processed tuna and mahi-mahi. In this procedure, histaminase acts on histamine, forming hydrogen peroxide which is then broken down by peroxidase with simultaneous formation of crystal violet from the oxidation of its leuco form. If further studies confirm that this test has satisfactory sensitivity, then it could be a valuable tool to monitor incoming scombroid fish for toxic levels of histamine (resulting from extensive microbial growth) and thus reduce the frequency of scombroid poisoning from fish. Limulus Ameebocyte Lysate Test Endotoxins of gram-negative bacteria can be detected in small quantities by the Limulus amoebocyte lysate test (LLT). This test has mainly been
120 EVALUATION OF TtIE ROLE OF MICROBIOLOGICAL CRITERIA used in clinical and pharmaceutical microbiology, but has also been ex- perimentally used for detection of endotoxins in foods (Haska and Ny- strand, 1979; Jay, 1977; Sullivan et al., 19831. The LLT has been used experimentally as a simple, rapid (two hours) screening test to estimate the microbial quality of certain foods such as fresh beef and fish in which, under aerobic conditions of storage, gram-negative bacteria are responsible for quality deterioration. The LLT potentially could be used in the future as a test to estimate shelf-life of foods such as ground beef and fish. However, at present there is not enough information to judge whether or not the LLT may have practical application in criteria to evaluate the microbiological condition of foods. Extract Release Volume Several reports relate extract release volume (ERV) to degree of mi- crobial deterioration of fresh meats (Jay, 19781. As meat deteriorates, the amount of water retained increases and ERV decreases. According to Jay (1978) this technique is of value in determining incipient spoilage in meats as well as in predicting refrigerator shelf-life. At the present, ERV is used for research and investigational purposes only and has no application as a parameter to evaluate the microbiological condition of foods. Adenosine Triphosphate Determination of bacterial adenosine triphosphate (ATP) in foods is based on the fact that all terrestrial life forms contain ATP. Experimental data indicate that microorganisms can be detected in foods by this method (Sharpe, 1973; Wood and Gibbs, 19821. At the present no practical routine methods exist to apply ATP as a parameter to evaluate the microbiological condition of foods. ASSESSMENT OF INDICATORS Indicators of Potential Human or Fecal Contamination or Possible Presence of Pathogens Staphylococci Staphylococci originate from the nasal passages, skin, and lesions of man and other mammals. They are applicable as part of microbiological criteria for cooked foods, various other food products that are commonly handled after heat processing, and foods that are handled extensively in preparation.
SELECTION OF INDICATOR ORGANISMS AND AGENTS 121 Staphylococci are usually killed during heat processing. Their reap- pearance on heat-processed foods is due to handling by workers or, less commonly, by contact with contaminated equipment or air. Small numbers of S. aureus are to be expected in foods that have been exposed to or handled by food workers. Large numbers (the number depending on the type of product) usually result from growth and almost invariably occur in foods that have been heat processed to eliminate competing organisms. Temperature abuse then sets the stage for growth. The presence of S. aureus can indicate a potential health hazard (see Chapter 41. Large numbers of staphylococci may be indicative of the presence of toxins, although small numbers do not mean absence of toxin since large populations may have been reduced to smaller ones by a processing step, e.g., heating or fermentation. Methods for the detection and enumeration of S. aureus are adequate for their use in microbiological criteria (AOAC, 1980; APHA, 19841. The AOAC procedure was originally designed to increase the sensitivity of the plating procedure but unfortunately is in- hibitory to injured cells. The direct plating method employing Baird-Parker agar effectively recovers both injured and noninjured S. aureus (Rayman et al., 19784. S. aureus counts are useful as part of microbiological criteria for foods that are handled after heat processing. Escher~chia cold E. cold conforms to the definitions of Enterobacteriaceae, coliforms, and fecal coliforms, but is further identified by an IMViC (indole, methyl red, Voges-Proskauer, citrate utilization) pattern of + + - - or -. Its natural habitat is the intestines of vertebrate animals. Thus, the presence of E. cold in a food indicates the possibility that fecal contamination has occurred and that other microorganisms of fecal origin, including pathogens, may be present. At present, E. cold is the best in- dicator of fecal contamination among the commonly used fecal-indicator organisms. In evaluating foods for safety, the presence of E. cold signifies a more positive assumption of hazard than the presence of other coliform bacteria. The failure to detect E. cold in a food, however, does not assure the absence of enteric pathogens (Mossel, 1967; Silliker and Gabis, 19761. In many raw foods of animal origin, small numbers of E. cold can be expected because of the close association of these foods with the animal environment and the likelihood of contamination of carcasses from fecal material, hides, or feathers during slaughter-dressing procedures. These organisms are easily destroyed by heat processing of foods. Thus, the presence of E. cold in a heat-processed food means either process failure or, more commonly, postprocessing contamination from equipment or
122 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA employees or from contact with contaminated raw foods. If the objective is to check for postprocessing contamination, then coliforms, rather than E. colt, should be the target organisms. Large numbers of E. cold in some foods (soft cheese, for example) may be the result of growth in the product or buildup on the equipment. Routine methods for the detection and enumeration of E. cold (AOAC, 1980; APHA, 1984) are applicable for use in microbiological criteria. However, the MEN procedures for the enumeration of E. cold in foods are time-consuming, costly, inhibitory to injured cells, and lacking in precision. A direct plating procedure is now available (Anderson and Baird-Parker, 1975; Holbrook et al., 1980~. A comparative study showed that the direct plating procedure was preferable to the MEN method for enumerating E. cold in meats, because of its lower variability, better re- covery of E. cold from frozen samples, rapidity (24 hours vs. 8 days), decreased requirements for media, and decreased cost of analysts' time (Rayman et al., 19791. In addition, Anderson and Baird-Parker (1975) considered indole production, the index in the direct plating method a more reliable characteristic of E. cold from foods than gas production from lactose. Microbiological criteria involving E. cold are useful in those cases where it is desirable to determine if fecal contamination may have occurred. Contamination of a food with E. cold implies a risk that one or more of a wide diversity of enteric pathogens may have gained access to the food and introduced a health hazard. Fecal Coliforms Fecal coliform bacteria!are used as a component of microbiological standards to monitor the Wholesomeness of shellfish and the quality of shellfish growing waters (USDHEW, 19651. The term fecal coliform has arisen from attempts to find rapid, dependable methods to establish the presence of E. cold without isolation and purification of cultures and ap- plication of the lengthy and costly IMViC tests. The fecal coliforms are a group of organisms selected by incubating an inoculum derived from a coliform enrichment broth at higher temperatures (44° to 45.5°C/111° to 114°F) than used for incubating coliforms. Such preparations usually con- tain a high proportion of E. cold and are thus useful for indication of a probable fecal source. The fecal coliforms have a higher probability of containing organisms of fecal origin and hence of indicating fecal con- tamination than do coliforms that have received no further differential tests. The fecal coliforms comprise a population presumed to contain a high
SELECTION OF INDICATOR ORGANISMS AND AGENTS 123 proportion of E. cold but without the actual proportion of E. cold being positively established. In many foods, most fecal coliforms are E. colt, but some strains of Enterobacter and Klebsiella are included. In some foods, frozen vegetables for example, the composition may be different. Thus, if fecal coliforms in a food are used as an index of E. cold and therefore as an index of fecal contamination, this population's proportions must be established first. Fecal coliforms can become established on equip- ment and utensils in the food processing and handling environment and contaminate processed foods. They are easily destroyed by heat and are injured sublethally or may die during freezing and frozen storage of foods. Routine MPN procedures applicable for use in microbiological criteria are well established (AOAC, 1980; APHA, 1976, 1981, 19841. These MPN procedures are subject to limitations similar to those mentioned for E. cold MPN methods. The fecal coliform test is a useful part of the microbiological standard to evaluate the quality of shellfish growing waters. Its purpose is to reduce the risk of harvesting shellfish from waters polluted with fecal material. Because rapid, direct plating methods for E. cold (Anderson and Baird- Parker, 1975; Holbrook et al., 1980), including provisions for resusci- tation of injured cells, are now available it may be advantageous to use E. cold rather than fecal coliforms as a component of microbiological criteria for foods. Enterococci Enterococci have certain features that make them unique as indicator organisms. Most are quite salt-tolerant (grow in the presence of 6.5% NaCl). All are facultatively anaerobic and grow well at 45°C (1 13°F). Except for Streptococcus bovis and Streptococcus equines, they can grow at 7-10°C (44.6°-50°F). Unlike E. colt, they are relatively resistant to freezing. Streptococcus faecalis and Streptococcus faecium, the most common en- terococci of foods, are relatively heat-resistant and may survive usual milk pasteurization procedures. S. faecium may survive and cause defects in marginally pasteurized canned cured meats. Enterococci may be of fecal origin from both warm-blooded and cold- blooded animals; they also can establish an epiphytic relationship with growing plants (Mundt, 1970~. They are often associated with insects and are thus a part of the natural microflora of many foods. In addition, they can establish themselves and persist in the food processing establishment long after their introduction. Many foods normally contain small to large numbers of enterococci, especially S. faecalis and S. faecium. For instance, certain cheeses and
124 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA some fermented sausages may contain more than 106 enterococci per gram. A wide variety of other foods such as raw meats, poultry, and raw veg- etables may have relatively low levels (10~-103 per gram). It is well es- tablished that enterococci counts of foods are not a reliable index of fecal contamination. A thorough understanding of the role and significance of enterococci in a food is required before any meaning can be attached to their presence and population numbers. Many media have been proposed (APHA, 1976, 1984; ICMSF, 1978) for the selective isolation and/or enumeration of enterococci. Present meth- ods have definite shortcomings relative to the degree of selectivity, quan- titative recovery or differential ability (APHA, 19841. Enterococci counts have little useful application in microbiological criteria for foods. If they are used in specific cases to identify poor manufacturing practices, it is necessary to first establish the normal population levels at different stages of processing and handling with a standardized test procedure. Pseudomonas aeruginosa P. aeruginosa is a frequent contaminant of the environment (including water) and enters the environment with fecal wastes of human or of animals associated with humans (Hoadley, 19761. It is an opportunistic pathogen and, in some countries, is of particular concern for infants who may become infected when contaminated water or equipment is used to make infant formulas. This organism has been used in Europe as an indicator of contamination of bottled mineral water. It is oligocarbo-tolerant; there- fore, after a period of adaptation, it can multiply in water of low nutrient level. Although there is not an acceptable standard method of detection of P. aeruginosa, methods are given in APHA (1981) and by Mossel et al. (1976~. Most other test procedures, except for incubation at 42°C (107.6°F), are of little value in detecting P. aeruginosa. At the present time, P. aeruginosa has no application in microbiological criteria for water in the United States. Indicators of Post-Heat Processing Contamination Coliform Bacteria Coliform bacteria are members of the Enterobacteriaceae that are ca- pable of fermenting lactose with the production of acid and gas within 48 hours at 35°C (95°F); for dairy products some workers specify 32°C (89.6°F). Some coliforms (E. coli) are common in feces of man and other animals, but others (Enterobacter, Klebsiella, Serratia, Erwinia, and
SELECTION OF INDICATOR ORGANISMS AND AGENTS Aeromonas) are commonly found in soil, water, and grains. 125 Small numbers of coliform bacteria are usually present in raw milk, vegetables, meat, poultry, and many other raw foods. Therefore, they are of little, if any, value in monitoring raw foods. These organisms are easily killed by heat; hence, their presence in heat-processed foods suggests post- heat contamination or, possibly, process failure. To locate the source and/ or mode of entry of coliform bacteria into a food may require the ex- amination of line samples. When large numbers are present in a heat- processed food, growth most likely occurred because of a lack of or improper refrigeration. Coliform bacteria are easily sublethally stressed by freezing, so coliform counts of frozen foods should be interpreted cautiously. Special resuscitation procedures are recommended (APHA, 1976, 1984) for their detection and enumeration from frozen foods. From their original fecal, soil, water, or plant environment, coliform bacteria can reach the food processing plant, food service establishment, or home environment, where they may be spread via equipment and utensil surfaces or by employees. The presence of coliforms in food does not necessarily mean that there was fecal contamination or that pathogens are present. Their significance in foods depends upon the circumstances to which the food has been exposed. The fecal connotation of E. colt, a member of the coliform group, is often inappropriately linked to foods in which coliform bacteria are found. Further examination (see sections on fecal coliform bacteria and E. colt, above) is required to establish potential fecal association. The use of coliform bacteria as an index of contamination after heat processing of foods requires a thorough understanding of the production, processing, and distribution practices to which a food is sub- jected and the effect of these practices on coliform bacteria. Methods for the enumeration of coliforms are adequate for purpose of use in micro- biological criteria (AOAC, 1980; APHA, 1978, 1984; ICMSF, 19781. Coliform bacteria are particularly useful as part of microbiological cri- teria to indicate postprocessing contamination of foods that have been processed (heating, irradiation, or chlorination) for safety. They are also useful indicators in guidelines at critical control points, particularly after heat processing. Enterobactertaceae Enterobacteriaceae counts are not a component of microbiological cri- teria for foods in the United States; they are, however, used in Europe for this purpose. The enrichment-plating procedure, which employs violet- red bile agar with 1% glucose, allows colony formation by a greater spectrum of members of the Enterobacteriaceae than the usual procedures,
126 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA which select for the lactose-positive members only (Mossel, 1967; Mossel et al., 19631. The Enterobacteriaceae test, like the coliform test, is useful to indicate postprocessing contamination in foods, but the presence of either group does not imply fecal contamination. Whether or not the Enterobacteriaceae count should supplant the coliform count in micro- biological criteria is a debatable question. TESTS FOR OTHER COMPONENTS Tests for Metabolic Products of Pathogens That Indicate a Health Hazard In a number of circumstances, tests for metabolic products of pathogens are preferred to tests for pathogens (see Chapter 4) to indicate the presence of pathogens or their toxins. Thermonuclease Test for Evidence of Growth of Staphylococci and Presence of Enterotoxins When large numbers ('106/g or ml) of S. aureus are present or sus- pected in processed foods, actual testing for enterotoxins is desirable, but the procedure is time-consuming and expensive. Furthermore, not all food laboratories have the capability to perform the test (see Chapter 9, Part A). Low counts of S. aureus in reheated or fermented foods may be misleading because the number of viable cells of 5. aureus may be the remainder of larger populations, whereas enterotoxins may be present. S. aureus pro- duces thermostable deoxyribonuclease (TNase), which has been used as a rapid and inexpensive procedure for screening foods for indication of extensive staphylococcal growth and possible presence of enterotoxin (APHA, 19841. The method is adequate for application in microbiological criteria. The TNase test has been recommended for testing foods such as cheeses and sausages when conditions such as improper lactic starter performance may have been responsible for the presence of significant levels of coagulase-positive S. aureus (Emswiler-Rose et al., 1980; Todd et al., 1981. See also Chapter 9, Part A). TNase can be a useful indicator because it can almost always be detected in foods whenever enterotoxins can be detected (Tatini, 19811. TNase-positive samples, whenever pos- sible, should be tested for enterotoxin. Afiatoxin Detection by Ultraviolet Light Long-wave ultraviolet (black) light (U.V.) has been used to detect the presence of Aspergillus flavus and Aspergillus parasiticus in corn (Shot
SELECTION OF INDICATOR ORGANISMS AND AGENTS 127 well and Hesseltine, 19811. When corn viewed under U.V. light displays a bright greenish-yellow fluorescence (BGYF), it is probable that aflatoxin may be present. Detection procedures for aflatoxins are given in AOAC (1980~. Because of the association of BGYF with the potential occurrence of aflatoxins, the examination of corn and other grains with U.V. light as a rapid screening procedure has been adopted by industry. This ex- amination is utilized only as a presumptive test; the occurrence of BGYF is indicative that more sophisticated testing for the presence of aflatoxin, e.g., by high-pressure liquid chromatography, is needed. The use of U.V. examination as a component of microbiological criteria is limited to pur- chase specifications or guidelines. Test for Phosphatase The phosphatase test is part of a criterion for certain milk and milk products to determine whether the product in question was pasteurized properly and also to detect the possible addition of raw milk to pasteurized milk (APHA, 1978; USPHS/FDA 19781. Methods, limitations, and inter- pretation of test results are presented in detail in Standard Methods for the Examination of Dairy Products (APHA, 19781. Routine methods for the detection of phosphatase are well established and applicable to evaluate the acceptability of pasteurized milk and milk products. REFERENCES Amerine, M. A., R. M. Pangborn, and E. B. Roessler 1965 Principles of Sensory Evaluation of Food. New York: Academic Press. AMI (American Meat Institute) 1982 Good Manufacturing Practices. I. Voluntary Guidelines for the Production of Dry Fermented Sausage; II. Voluntary Guidelines for the Production of Semi-dry Fer- mented Sausage. Washington, D.C.: AMI. 10 pp. Anderson, J. M., and A. C. Baird-Parker 1975 A rapid and direct plate method for enumerating Escherichia cold biotype 1 in food. J. Appl. Bacteriol. 39:111-117. Anonymous 1979 Shrimp Decomposition Workshop. National Shrimp Breaders and Processors Asso- ciation, National Fisheries Institute and FDA, Tampa, Florida. AOAC (Association of Official Analytical Chemists) 1980 Official Methods of Analysis of the Association of Official Analytical Chemists. 13th Ed. W. Horwitz, ed. Washington, D.C.: AOAC. APHA (American Public Health Association) 1976 Compendium of Methods for the Microbiological Examination of Foods. M. L. Speck, ed. Washington, D.C.: APHA. 1978 Standard Methods for the Examination of Dairy Products. 14th Ed. E. H. Marth, ed. Washington, D.C.: APHA. 1981 Standard Methods for the Examination of Water and Wastewater. 15th Ed. A. E.
128 EVALUATION OF THE ROLE OF MICROBIOLOGICAL CRITERIA Greenburg, J. J. Conners, D. Jenkins, and M. A. H. Franson, eds. Washington, D.C.: APHA. 1984 Compendium of Methods for the Microbiological Examination of Foods. 2nd Ed. M. L. Speck, ed. Washington, D.C.: APHA. Beacham, L. M. 1946 A study of decomposition in canned oysters and clams. J. Assoc. Off. Agric. Chem. 29:89-99. Broitman, S., W. L. Mallmann, and G. M. Trout 1958 A simple test for detecting keeping quality for milk. J. Milk Food Technol. 21:280- 284. Bush, J. C. 1970 The Use of Oxidation-Reduction Dyes in the Determination of the Shelf Life of Meats. M.S. thesis. College Station: Texas A&M University. Chang, O., W. L. Cheuk, R. Nickelson, R. Martin, and G. Finne 1983 Indole in shrimp: effect of fresh storage temperature, freezing and boiling. J. Food Sci. 48:813-816. Cheuk, W. L., and G. Finne 1981 Modified calorimetric method for determining indole in shrimp. J. Assoc. Off. Anal. Chem. 64:783-785. Cobb, B. F., I. Alaniz, and C. A. Thompson, Jr. 1973 Biochemical and microbial studies on shrimp. Volatile nitrogen and amino nitrogen analysis. J. Food Sci. 38:431-436. Cobb, B. F., III., C. S. Yeh, F. Christopher, and Carl Vanderzant 1977 Organoleptic, bacterial and chemical characteristics of penaeid shrimp subjected to short-term high-temperature holding. J. Food Prot. 40:256-260. Crosgrove, D. M. 1978 A rapid method for estimating ethanol in canned salmon. J. Food Sci. 43:641-643. Duggan, R. E. 1948 Report on decomposition in shellfish indole in shrimp, oysters and crabmeat. J. Assoc. Off. Agric. Chem. 31:507-510. Duggan, R. E., and L. W. Strasburger 1946 Indole in shrimp. J. Assoc. Off. Agric. Chem. 29:177-188. Emswiler-Rose, B. S., R. W. Johnston, M. E. Harris, and W. H. Lee 1980 Rapid detection of staphylococcal thermonuclease on casings of naturally contaminated fermented sausages. Appl. Environ. Microbiol. 40:13-18. FDA (Food and Drug Administration) 1981 Defect action level for decomposition in imported canned and cooked frozen shrimp; availability of guide. Federal Register 46:39221. Fields, M. L. 1964 Acetylmethylcarbinol and diacetyl as chemical indexes of microbial quality of apple juice. Food Technol. 18: 1224-1238. Griffiths, M. W., J. D. Phillips, and D. D. Muir 1981 Thermostability of proteases and lipases from a number of species of psychrotrophic bacteria of dairy origin. J. Appl. Bacteriol. 50:289-303. Haska, G., and R. Nystrand 1979 Determination of endotoxins in sugar with the Limulus test. Appl. Environ. Microbiol. 38:1078-1080. Hill, E. C., and F. W. Wenzel 1957 The diacetyl test as an aid for quality control of citrus products. 1. Detection of bacterial growth in orange juice during concentration. Food Technol. 11:240-243.
SELECTION OF INDICATOR ORGANISMS AND AGENTS 129 Hill, E. C., F. W. Wenzel, and A. Barreto 1954 Colorimetric method for detection of microbiological spoilage in citrus juices. Food Technol. 8: 168- 171. Hoadley, A. W. 1976 Potential health hazards associated with Pseudomonas aeruginosa in water. Pp. 80- 114 in Bacterial Indicators/Health Hazards Associated with Water. A. W. Hoadley and B. J. Dutka, eds. Philadelphia: American Society for Testing Materials. Holbrook, R., J. M. Anderson, and A. C. Baird-Parker 1980 Modified direct plate method for counting Escherichia cold in foods. Food Technol. Austral. 32:78-83. Hollingworth, T. A., Jr., and H. R. Throm 1982 Correlation of ethanol concentration with sensory classification of decomposition in canned salmon. J. Food Sci. 47:1315- 1317. 1983 A headspace gas chromatographic method for the rapid analysis of ethanol in canned salmon. J. Food Sci. 48:290-291. ICMSF (International Commission on Microbiological Specifications for Foods) 1978 Microorganisms in Foods. 1. Their significance and methods of enumeration. Toronto: University of Toronto Press. 1980 Fish and shellfish and their products. Pp. 567-605 in Microbial Ecology of Foods. Vol. 2. Food Commodities. New York: Academic Press. Jay, J. M. 1977 The Limulus lysate endotoxin assay as a test of microbial quality of ground beef. J. Appl. Bacteriol. 43:99-109. 1978 Modern Food Microbiology. 2nd Ed. New York: D. Van Nostrand. King, W. H., and F. F. Flynn 1945 Experimental studies on decomposition of oysters used for canning. J. Assoc. Off. Agric. Chem. 28:385-398. Larmond, E. 1977 Laboratory Methods for Sensory Evaluation of Food. Research Branch. Canada Dept. of Agric. Pub. No. 1637. Lerke, P. A., M. N. Porcuna, and H. B. Chin 1983 Screening test for histamine in fish. J. Food Sci. 48:155-157. McClellan, G. 1952 Report on chemical indices to decomposition in shellfish. J. Assoc. Off. Agric. Chem. 35:524-525. Montgomery, W. A., G. S. Sidhu, and G. L. Vale 1970 The Australian prawn industry. 1. Natural resources and quality aspects of whole cooked fresh prawns and frozen prawn meat. CSIRO Food Preserv. Quart. 30(2):21. Mossel, D. A. A. 1967 Ecological principles and methodological aspects of the examination of foods and feeds for indicator microorganisms. J. Assoc. Off. Agric. Chem. 50:91-104. Mossel, D. A. A., M. Visser, and A. M. R. Cornelissen 1963 The examination of foods for Enterobacteriaceae using a test of the type generally adopted for the detection of salmonellae. J. Appl. Bacteriol. 26:444-452. Mossel, D. A. A., H. DeVor, and I. Eelderink 1976 A further simplified procedure for the detection of Pseudomonas aeruginosa in con- taminated aqueous substrata. J. Appl. Bacteriol. 41:307-309. Mundt, J. O. 1970 Lactic acid bacteria associated with raw plant food materials. J. Milk Food Technol. 33:550-553.
130 EVALUATION OF TlIE ROLE OF MICROBIOLOGICAL CRITERIA National Soft Drink Association 1975 Quality Specifications and Test Procedures for "Bottler's Granulated and Liquid Sugar." Washington, D.C.: Natl. Soft Drink Assoc. NCA (National Canners Association Research Laboratories) 1968 Laboratory Manual for Food Canners and Processors. Vol. 1. Microbiology and Processing. Westport, Conn.: AVI Publishing. Oberlender, V., M. O. Hanna, R. Miget, C. Vanderzant, and G. Finne 1983 Storage characteristics of fresh swordfish steaks stored in carbon dioxide enriched controlled (flow-through) atmospheres. J. Food Prot. 46:434-440. Parmelee, C. E. 1974 Early detection of psychrotrophs in pasteurized milk. Dairy and Ice Cream Field 157(8):38. Rayman, M. K., J. J. Devoyod, U. Purvis, D. Kusch, J. Lanier, R. J. Gilbert, D. G. Till, and G. A. Jarvis 1978 ICMSF Methods Studies. X. An international comparative study of four media for the enumeration of Staphylococcus aureus in foods. Can. J. Microbiol. 24:274-281. Rayman, M. K., G. A. Jarvis, C. M. Davidson, S. Long, J. M. Allen, T. Tong, P. Dodsworth, S. McLaughlin, S. Greenburg, B. G. Shaw, H. J. Beckers, S. Qvist, P. M. Nottingham, and B. J. Stewart 1979 ICMSF Methods Studies. XIII. An international comparative study of the MEN pro- cedure and the Anderson-Baird-Parker direct plating method for the enumeration of Escherichia cold biotype 1 in raw meats. Can. J. Microbiol. 25:1321-1327. Salwin, H. 1964 Report on decomposition and filth in foods (Chemical indexes). J. Assoc. Off. Agric. Chem. 47:57-58. Sharpe, A. N. 1973 Automation and instrumentation developments for the bacteriology laboratory. Pp. 197- 232 in Sampling Microbiological Monitoring of Environments. R. G. Board and D. W. Lovelock, eds. London: Academic Press. Shelef, L. A., and J. M. Jay 1970 Use of a titrimetric method to assess the bacterial spoilage of fresh beef. Appl. Microbiol. 19:902-905. Shotwell, O. L., and C. W. Hesseltine 1981 Use of bright greenish-yellow fluorescence as a presumptive test for aflatoxin in corn. Cereal Chem. 58: 124- 127. Silliker, J. H. 1963 Total counts as indexes of food quality. Pp. 102-112 in Microbiological Quality of Foods. L. W. Slanetz, C. O. Chichester, A. R. Gaufin, and Z. J. Ordal, eds. New York: Academic Press. Silliker, J. H., and D. A. Gabis 1976 ICMSF Methods Studies. VII. Indicator tests as substitutes for direct testing of dried foods and feeds for Salmonella. Can. J. Microbiol. 22:971-974. Stansby, M. E. 1963 Analytical methods. Pp. 367-373 in Industrial Fishery Technology. M. E. Stansby, ed. Huntington, N.Y.: Robert E. Krieger. Staruszkiewicz, W. F. 1974 Collaborative study of the gas-liquid chromatographic determination of indole in shrimp. J. Assoc. Off. Anal. Chem. 57:813-818. Sullivan, J. D. Jr., P. C. Ellis, R. G. Lee, W. S. Combs, Jr., and S. W. Watson 1983 Comparison of the Limulus amoebocyte lysate test with plate counts and chemical
SELECTION OF INDICATOR ORGANISMS AND AGENTS 131 analyses for assessment of the quality of lean fish. Appl. Environ. Microbiol. 45:720- 722. Tatini, S. R. 1981 Thermonuclease as an indicator of staphylococcal enterotoxins in food. Pp. 53-75 in Antinutrients and Natural Toxicants in Foods. R. L. Ory, ed. Westport, Conn.: Food and Nutrition Press. Todd, E., R. Szabo, H. Robern, T. Gleeson, C. Park, and D. S. Clark 1981 Variation in counts, enterotoxin levels and TNase in Swiss-type cheese contaminated with Staphylococcus aureus. J. Food Prot. 44:839-848. USDHEW (U.S. Department of Health, Education and Welfare) 1965 National Shellfish Sanitation Program, Manual of Operations. Part 1. Sanitation of shellfish growing areas. Washington, D.C.: U.S. Government Printing Office. USPHS/FDA (U.S. Public Health Service/Food and Drug Administration) 1978 Grade A Pasteurized Milk Ordinance. 1978 Recommendations. USPHS/FDA Publ. No. 229. Washington, D.C.: U.S. Government Printing Office. Wood, J. M., and P. A. Gibbs 1982 New developments in the rapid estimation of microbial populations in foods. Pp. 183- 214 in Developments in Food Microbiology-l. R. Davies, ed. Englewood, N.J.: Applied Science.
