MICROBIOLOGY OF NATURAL MINERAL WATERS Pr Henri LECLERC Professeur Emérite des Universités Spécialiste en Microbiologie de l'environnement¨, Laboratoire de Microbiologie, Faculté de Médecine 1 place de Verdun, 59045 Lille Cedex, France Definition and characteristics Groundwater habitat Bacterial flora from source to bottle Microbial community Inhibitory effect of autochtonous flora Assessing health risk from autochtonous flora Source-water monitoring for cryptosporidium protection/a> References Definition and characteristics Natural mineral water is defined as microbiologically wholesome water, originating from an underground water table or deposit and emerging from a spring tapped at one or more natural, or borehole sources. Natural mineral water can be clearly distinguished from ordinary drinking water by its nature, which is characterized by the mineral content, trace element components or other constituents, perhaps by certain beneficial effects; and by the fact that it has not been treated, having preserved its original qualities of the underground origin of the water, which has been protected from pollution. These characteristics, which may make natural mineral water beneficial to health, must have been assessed for geological, hydrological, physical, chemical and physico-chemical, microbiological, and if necessary, pharmacological, physiological and clinical characteristics. The composition, temperature and other essential characteristics of natural mineral water must remain stable within the limits of natural fluctuations, to show that the aquifer is stable and is not affected by surface alterations. Natural mineral waters are bacteriologically wholesome and, like all subterranean waters, contain a natural bacterial flora. The presence of the normal flora of mineral water has given rise to a number of questions and discussions about their effect on health, primarily because this flora is not yet very well known. However, natural mineral waters cannot be subjected to any type of desinfection that modifies or eliminates their biological components and, therefore, always contain bacteria that are primarily a natural component of these waters. Groundwater habitat Before the 1970s the study of life in groundwater habitats was relatively limited. In 1970s, however, it became increasingly obvious that certain waste disposal practices were contaminating subsurface environments with effects on ground water quality, leading to a current interest in the study of these environments. There has also been an increasing interest in demonstrating that a variety of shallow (Balkwill and Ghiorse, 1985; Madsen and Ghiorse, 1993), and deep environments (Balkwill, 1989; Fredrickson and Phelps, 1997) contain substantial numbers of viable microorganisms and in using the ability of these microorganisms to degrade potential polluants, i.e., in bioremediation. Subsurface microbiological research has progressed with the development or aseptic sampling techniques to study microbial community structure, microbial activities and geochemical properties of ground water environments (Chapelle, 1993; Madsen and Ghiorse, 1993; Wilson, 1995; Fredrickson and Phelps, 1997). The most fundamental distinction between subsurface hydrological environments is the difference between the satured and unsatured zones. The unsatured zone is usually divided into three components (Figure 1). The first of these is the soil zone with the A, B and C horizons, generally 1 or 2 meters thick, which contains living roots and which support plant growth. The underlying material, is often referred to as the intermediate zone and consists of sediments or rocks that have not been exposed to extensive pedogenic processes. The boundary between the unsatured zone and the satured zone is termed the capillary fringe. This boundary will fluctuate according to seasons, rainfalls or pumping rates and this will induce chemical oxydation-reduction reactions. The unsatured zone is characterized by pore spaces that are incompletely filled with water. Pore space that is not filled with water is filled with gas and the capillary forces between water and sediment particles prevent water from flowing to wells. Thus, the bottom of wells completed in the unsatured zone does not yield appreciable quantities of water. Fig 1. The three components of the unsaturated zone In contrast to surface freshwater, which may be influenced by suspended particles and their attached biota, most groundwater is interstitial, i.e., remaining within a matrix of minerals with variable chemistry, porosity, and degree of saturation. Thus in groundwater habitat, all life forms larger than microorganisms are excluded. Madsen and Ghiorse (1993) have presented a generalized model for the relationship between geological stratigraphy and microbiological parameters (Figure 2). In going across the A and B soil horizons into the C horizon, the bacterial abundance decreases in direct proportion with nutrient levels. The C soil horizon marks the beginning of the unsaturated subsurface zone with considerably fewer bacteria than in the B soil horizon. Below the C horizon microbial abundance increases substantially at the water table and, just above, in the capillary fringe zone. It can be speculated that the interface zones between the unsaturated and saturated zones may be the site of oxygen transport and nutrients recharge, especially in shallow unconfined aquifers (Madsen and Ghiorse, 1993). Thus, depth per se, appears not to govern bacterial abundance and activity in the saturated zone. Rather, the abundance of bacteria appeears to be related to hydrological, physical and geochemical properties of each stratum. Fig 2. Generalized vertical profile showing relationships between microbiological parameters and geologic stratigraphy.(Madsen and Ghiorse, 1993) Nutrient limitations and starvation-survival In shallow or deeper aquifers, the supply of readily utilizable carbon and energy sources may be extremely small. Only the most refractory organic material will persist the long and complex pathways through subsurface sediments. Thus, organic carbon is the most limiting nutrient in these aquifer systems. Low-nutrient environments, termed oligotrophic environments, lack primarily organic matter for the growth of heterotrophic bacteria. Limitation or starvation for one or more nutrient is common in most bacteria in natural environments such as groundwater. Even bacteria such as the enterics, which colonize the gut, must be able to survive in the environment as they are transmitted from host to host. Thus, bacteria have evolved mechanisms to cope with a feast-fast type of existence. 2.2. The viable but non-culturable state Under certain conditions of metabolic stress such as starvation, bacterial cells may enter into a viable but non-culturable state. It has been realized for some time that plate counts can dramatically underestimate the total number of bacteria, determined by acridine orange, present in samples taken from natural environments. In the late 1970s, several non-cultural methods (Zimmerman et al., 1978; Kogure et al., 1979; Rodriguez et al., 1992) for determining cell viability. These studies led to the concept that some bacteria, in response to certain environmental stresses, may loose the ability to grow on media on which they are routinely cultured, while remaining viable. A bacterium in this viable but non culturable (VBNC) state, is defined by Oliver (1993) as "a cell which can be demonstrated to be metabolically active, while being incapable of undergoing the sustained cellular division required for growth in or on a medium normally supporting growth of that cell". Bacterial flora from source to bottle In Figure 3 taken from Bischofberger et al. (1990) and used here as an example, it can seen that the colony counts of the water from five springs and from the mixed water derived from these springs are less than 1-4 CFU ml-1, whereas in the storage tank and immediately after bottling they are on average only slightly higher. During storage at 20°C bacterial population increase in numbers reaching a peak more than 105 CFU ml-1 by the end of one week. During the next four weeks, the bacterial counts decrease slowly or remains fairly constant. At the end of the 2 years of storage, colony counts are still about 103 CFU ml-1. The results of most studies (Yurdusef and Ducluzeau, 1985; Gonzalez et al., 1987; Morais and da Costa, 1990; Vachée et al., 1997) are broadly in agreement with the one reported by Bischofberger et al. (1990). Some of the results on colony counts from bottle still mineral waters are also described in the Table 1. Table 1. Heterotrophic plate counts (HPC) at 22°C in bottled .non -carbonated mineral waters samples at retail outlets Source Number of samples examined Heterotrophic plate counts (CFU/ml) Reference     < 102 102-103 103-104 104-105 >105   UK 44 18 11 18 36 16 Hunter et al (1990) France Portugal Belgium 23 26 4 34 34 - Manaia et al (1990) France Italy Spain 50 2 8 56 30 4 Leclerc (1994) Fig 3. Colony counts of non-carbonated natural mineral water of the source A in the five different springs, mixed water, reservoir, glass and plastic bottles immediately after bottling and after 1, 2, 3, 4, 5 and 104 weeks of storage at 20°C. (Bischofberger et al, 1990) These bacteria, capable of growing on simple organic compounds (principally carbohydrates, amino acids and peptides) found in the culture medium such as "Plate count agar" or R2A agar (Reasoner and Geldreich, 1979, 1985; Reasoner, 1990), are heterotrophic. These bacteria are also psychrotrophic, because they can grow at temperatures as low as 5°C, and their maximum growth temperature is about 35°C (Mossel et al., 1995). Therefore, incubation at 20°C for 3 days has prevailed in the monitoring of plate counts of mineral waters. Furthermore, these bacteria do not have growth factor requirements such as vitamins, amino acids or nucleotides and are, therefore, prototrophic, in contrast to auxotrophic bacteria which require many of these growth factors. Growth or resuscitation Bacteria living in ground-water systems are subject to constantly changing, and frequently stressful, conditions such as temperature downshift, fluctuation in pH, osmotic pressure, oxidative shock and nutrient limitation. In nutrient-poor ground-water systems, autochtonous bacteria survive for prolonged periods under conditions designated starvation-survival by Morita (1982, 1987, 1993). They may loose the ability to grow on media on which they are routinely cultured, while remaining viable (VBNC). As mentioned earlier, it is now well established that plate counts of still mineral waters at the source, or immediately after bottling range between about 1 to 10 ml-1 increasing to 104-105 ml-1 within 3-7 days after bottling. This population remains high and is generally stable for months. It remains unclear whether the ultimate large population of culturable bacteria in mineral water is due to resuscitation of a large number of non culturable dormant cells (VBNC) present in the water source, or in the bottling system, or is the result of cell-division and growth of a few culturable cells initially present. Preservation of biological constituents through bottling Mineral water ecosystems, as other aquifers, exhibit a high degree of phenotypic and genetic microbial diversity that cannot always be supported by species identification (Chapelle, 1993). Phenotypic characteristic, which rely on physiological activities, have been shown to be 1ess important for estimating bacterial diversity than genetic characteristics, because many metabolic traits may be induced or repressed by different environmental conditions. Restriction fragment length polymorphism patterns (RFLP) of rDNA regions (ribotyping), therefore, constitute a more reliable method for assessing genetic diversity within autochtonous bacterial associations of mineral water. In a recent investigation on the fate of the bacterial flora at source, before bottling and upon bottling, Vachée et al. (1997) isolated 890 strains from five springs and observed 378 distinct ribotypes. RFLP analysis detected a large number of polymorphisms combined with unequivocal band resolution in all groups, but particularly high in a set of isolates producing a fluorescent pigment (72 patterns for 174 strains). Mineral water samples from the five springs were analysed during one year in order to assess whether the isolates specific for each spring could be repeatedly isolated throughout the period of storage. In the case of spring A, 155 isolates provided 75 RFLP patterns. Indistinguishable or closely related isolates were found 114 times in the samples examined and, among these isolates 103, or 90% of the isolates had been obtained from samples before and after bottling (Table 2). The results from four springs are presented in Table 3. These data suggest that, within the ground-mineral water bacterial community, a high percentage of indistinguishable or closely related isolates (identical ribopatterns) is preserved through the bottling system and storage. Table 2. Ribopatterns of indistinguishable or closely related isolates found in samples during storage (Guillot and Leclerc, 1993) All samples Samples   I II III   Number of indistinguishable isolates 5 - 2 3 3 1 1 1 3 1 1 1 4 2 - 2 2 1 - 1 17 9 5 3 10 5 3 2 3 1 1 1 2 - 1 1 5 1 3 1 2 1 - 1 2 1 - 1 2 1 - 1 2 1 - 1 3 2 - 1 3 2 - 1 3 1 - 1 3 1 - 1 2 - 1 1 2 - 1 1 3 1 1 1 4 3 - 1   Number of closely related isolates 4 2 1 1 2 - 1 1 3 1 1 1 3 1 1 1 2 1 - 1 3 1 1 1 Table 3. Indistinguishable isolates (identical ribopatterns) recovered during storage (%) (Guillot and Leclerc, 1993)   A B C D   155 133 141 173 Indistinguishable isolates recovered more than one time 114 (74%) 96 (72%) 93 (70%) 94 (55%)   Indistinguishable isolates recovered before and after bottling 103 (90%) 50 (52%) 62 (65%) 61 (66%) Microbial community For natural mineral waters all of the data has been obtained, thus far, by culture methods. It must also be stressed that the study of the microbial populations of mineral water will undergo important developments as new molecular biological and other highly technical approaches are likely to be used to study this environment. Heterotrophic bacteria Because photosynthesis is not possible in ground-water environments, the food chain in ground water is primarily heterotrophic, depending either on DOC from the hydrological flow path, or on organic compounds of sedimentary origin. The composition of the bacterial flora that can be recovered depends largely on the culture techniques used and on the physico-chemical parameters of the aquifers. The heterotrophic plate count (HPC) introduced by the Standard Methods of the USA (Anon, 1985). This method enumerates aerobic and facultative aerobic bacteria found in potable water that are capable of growth on simple organic compounds found in the culture medium, for a specific incubation period and at specific temperatures. This method has been successfully applied to bottled non-carbonated mineral water, with some modifications (Bischofberger et al., 1990). The vast majority of the heterotrophic bacteria isolated from natural mineral waters can be classified in a few phylogenetic divisions. Prosthecate bacteria belonging to the alpha subclass of the Proteobacteria, Pseudomonads of the alpha, beta and gamma subclass of the Proteobacteria, members of the Cytophaga - Flavobacterium - Bacteroides (CFB) phylum, Gram-positive bacteria of the Actinomycetes subclass are the most common bacteria isolated from bottled mineral water. 4.2. Pseudomonas - Acinetobacter - Alcaligenes The classification of Gram-negative aerobic rods is becoming more and more complex due to the creation of many new genera and the description of numberless species. The species of the genus Pseudomonas comprise a substantial proportion of the microflora of free-living saprophytes in soils, fresh water, groundwater, marine environments and many other natural habitats, especially plants. During the last decade there has been a considerable revision of the phylogenetic relationships of these organisms leading to the description of new groups within the Proleobacteria (Figure 4). The most frequently isolated organisms from natural mineral waters belong to the a-, b- and g-subclass of the Proteobacteria and especially to genus Pseudomonas belonging to rRNA group I (Kersters et al., 1997). The species of Acinetobacter are nonmotile and oxidase negative, while Alcaligenes spp. are motile by peritrichous flagella. All species are aerobic, having a strictly respiratory type of metabolism with oxygen as the terminal electron acceptor. Fig 4. Phylogenetic relationships of proteobacterial groups (stipped triangles) containing species currently or formerly assigned to the genus Pseudomonas (bold) and selected reference groups (Kersters et al, 1996) The organisms most widely isolated from mineral water and representing major groups are shown in Table 4. These results are obtained in extensive studies by Schwaller and Schmidt-Lorenz (1981a, 1981b), Bischofberger et al. (1990), Manaia et al. (1990), Guillot and Leclerc (1993), and Vachée et al. (1997). In some studies such as that of Guillot and Leclerc (1993) and Vachée et al. (1997) the unidentified isolates reached about 80%. These results are not surprising because of the large phenotypic and genotypic diversity of bacteria from groundwater and the large number of unclassified species in this environment. Furthermore, some common species are remarkably heterogeneous; this is the case with P. fluorescens that can be subdivided by various criteria into subspecies or biovars. Table 4. Major groups of bacteria isolated in natural mineral waters Species, genus or group Schwaller and Schmitz-Lorenz (1980) Bischofberger et al (1990) Manaia et al (1990) Leclerc et al (1993, 1997) Pseudomonas fluorescent ++ ++ ++ ++ non fluorescent ++ + ++ + Burkholderia cepacia - + + + Ralstonia pickettii + - + + Burkholderia-Ralstonia (P. lemoignei) - ++ - - Comamonas acidovorans + - ++ + Comamonas testosteroni + + - - Acidovorax delafieldii + + - - Stenotrophomonas maltophilia - + + + Brevundimonas diminuta - - - + Brevundimonas vesicularis - - - + Sphingomonas paucimobilis - - + + Acinetobacter ++ + + + Alcaligenes + + ++ + Cytophaga - Flexibacter - Flavobacterium ++ ++ ++ + Arthrobacter - Corynebacterium + ++ - + By linking universal probes (16S rRNA sequences) with fluorescent dyes, phylogenetic diversity could be analysed by epifluorescence microscopy and flow cytometry (Amann et al., 1990a, 1990b, 1995; Amann, 1995; Wallner et al., 1996). This new approach can reveal microbial community structures and it is clear that the use of 16S rRNA gene technology may bring new information in environmental microbiology and especially in the case of microbial communities of natural mineral waters (Wagner et al.,. 1993; Wallner et al., 1993, 1995). Inhibitory effect of autochtonous flora Natural mineral water is not subjected to antibacterial treatments of any kind and, after bottling, it is often stocked for several months before it is distributed and sold. To assess public health risks it is, therefore, important to know the survival capacity of pathogen and indicator bacteria. The available data on the survival of bacteria in surface waters cannot be completely extrapolated to bottled mineral waters. It is important for example, to take into account some specific factors such as the impact of drilling, and bottling stress, the selective attachment of some populations to solid surfaces, the fate of autochtonous populations which can reach very high numbers a few days after bottling, the effect of an enclosed environment (bottle effect) and the influence of PVC, PET or glass of the bottles. Ducluzeau was the first to study the survival of enterobacteria in mineral water to assess the influence of autochtonous bacteria on indicator bacteria (Delabroise and Ducluzeau, 1974; Ducluzeau et al., 1976a, 1976c; Lucas and Ducluzeau, 1990a, 1990b). In the most significant experiment, Escherichia coli was inoculated into sterile water at a concentration of 1.2 x 105 CFU ml-1. The plate counts of E. coli was reduced by less than one log over of a 3 month period, and more than 102 CFU ml-1 were still detected five months later. On the other hand, when this experiment was repeated with mineral water, that is in the presence of the autochtonous mineral flora (between 5.104 and 5.105 CFU ml-1), the complete loss of viability of E. coli took place between 35 and 55 days depending on the experiment. The same loss of viability of the test organism exerted by the whole autochtonous flora, was exhibited by some strains from the dominant flora. Therefore, it appears that these authors observed an antagonistic activity by autochtonous flora of the mineral water, and not an effect due to the physical or chemical properties, of the water itself. The antagonistic activity could possibly be related to an inhibitory substance accumulating in the water during the successive cycles of growth and death of the autochtonous bacteria population (Figure 5). Fig 5. Effect of previous exposure of Vittel water to autochthonous organisms on its antagonistic effect against E. coli (Ducluzeau et al, 1976) More recent studies by Moreira et al. (1994) with Escherichia coli and other coliform indicators such as Enterobacter cloacae and Klebsiella pneumoniae showed that the viable counts of the three test enterobacteria decreased under all experimental conditions, but the decrease depended on the organism, and the conditions in which they were examined (Figure 6). Fig 6. Survival determined by viable counts (mean of two independent experiments) on tryptic soy agar of Escherichia coli inoculated in mineral water in PVC bottles (Moreira et al, 1994) Pseudomonas aeruginosa is frequently isolated from surface water and is also a major concern in mineral water bottling plants because it can contaminate boreholes and bottling plants. Therefore, the antagonistic effect of the autochtonous flora on P. aeruginosa was examined in three types of natural mineral water (very low mineral content, low mineral content, rich in mineral content) with an inoculum that gave a final concentration of approximately one organism per ml in the bottled water (Vachée and Leclerc, 1,995). Four test strains were used; one obtained from a culture collection, one from a patient with septicemia and two from surface water. The test bacteria were inoculated immediately after sampling from the source of the mineral waters. Overall experimental conditions mimicked natural contamination before bottling. In the filter sterilized waters P. aeruginosa attained more than 104 CFU ml-1 a few days after inoculation, and remained almost constant during the nine months of the experiment. In mineral water with the autochtonous flora, the initial inoculum did not increase at all during the experiment (Figure 7). Fig 7. Survival or growth determined by viable count on selective medium of Pseudomonas aeruginosa (wild strain) inoculated in mineral water (spring A) maintained at room temperature (about 20°C),with and without autochtonous flora (Vachée and Leclerc, 1995) Assessing health risk from autochtonous flora There are several approaches to detect bacterial population such as those autochtonous to mineral waters that could have public health importance but are not known to be pathogenic. The methods available include animal model systems and epidemiological studies One other approach is to search for virulence factors from bacterial isolates. Inoculation of the digestive tract of axenic mice Axenic animals, constitute a first-choice for the goal of determining whether autochtonous bacteria of mineral water are able to adhere, penetrate and multiply in epithelial cells or produce toxins or other substances causing tissue damage. The most (stringent) experiments were devised to compare the transit of an inoculum of several autochtonous strains and that of the spores used as markers (Figure 8). In spite of an equivalent number of Pseudomonas (strain Pl) cells and of the inert marker in the inoculum, the maximum number of Pseudomonas in the feces was lower than that of the spores, and the former disappeared from, the feces more rapidly than the latter. Thus, a partial destruction of Pseudomonas Pl was shown during its transit through the digestive tract. Other assays, performed with other strains of Pseudomonas or Acinetobacter provided similar results. Fig 8. Transit of strictly thermophilic B. subtilis spores as compared to that of Pseudomonas strains P1 in the digestive tract of axenic mice (Ducluzeau et al, 1976) Randomized trials in infants In the past, mineral water conditioned in glass bottles was used. Since 1970 PVC conditioning has been used, and some people have wondered about the modifications in the microbial populations which may have resulted from, using water bottled in PVC, as well as effects on health of babies. In order to answer this question, a study was carried out from January 1984 to January 1986 by Leclerc (1980) with two groups, each with 30 babies. One group was fed with milk made up from mineral water in PVC bottles, the other group was fed with milk made up from the same pasteurized mineral water in glass bottles. The method used was double-blind. In no case was it possible to isolate mineral water-derived bacteria from, rhinopharyngeal samples analysed one or two hours after the drinking of bottled of milk. Nor was there evidence for digestive tract colonization from the analysis of the stool samples. From an epidemiological point of view, no difference could be found in the two groups. Of the 60 babies, 9 showed signs of diarrhea as defined in the protocol. Of these 9 babies, 5 received milk prepared with sterile mineral water and 4 were fed with milk conditioned with natural mineral water. In no case did the risks appear sufficient severe to justify the suspension of milk feeding. It may be concluded that there was no difference in the pathology observed in the two groups. Virulence characteristics of bacteria Only Edberg et al. (1997) studied HPC bacteria isolated from bottled water. Health risks were estimated by the determination of cytotoxicity and invasiveness in a human enterocyte cell line. More than 95% of naturally occurring HPC bacteria showed low invasiveness and cytotoxicity. When either invasiveness or cytotoxicity was demonstrated, only a small number of cells from, the culture were positive. Overall experimental data show that autochtonous bacteria of natural mineral waters, have never brought about detectable pathological disorders in man and animal and, in vitro, are incapable to directly damaging human cells in tissue culture. Since the existence of European regulations dating from 1980, no outbreak or single cases of disease due to the consumption of natural mineral water has been recorded in the literature, or by the health authorities of the countries within the European Community. Source-water monitoring for cryptosporidium protection There are a number of critical differences between the way in which indicators are used for the monitoring of source water and the way indicators have been otherwise used for the monitoring of tapwater from distribution systems. Under no circumstances should results obtained from tap water monitoring be applied to source water monitoring (Edberg, Leclerc and Robertson, 1997). First, source water does not have a residual disinfectant. The disinfectant in tapwater may affect bacteria before it affects parasites and viruses. Therefore, depending on the frequency of the collection of samples and the timing of the collection of the samples, bacteria may be rendered injured or nonrecoverable. Second, the sampling of tapwater for regulatory purposes is based on a rigid, infrequent sampling protocol. Generally, there is one sample taken per one thousand persons per month. Under this infrequent sampling protocol, it would be unlikely that there would be a bacterial indicator occurrence at the same time as a parasite or virus occurrence. Third, in the sampling of tapwater there is no linkage between the physical and biological indicators used for sampling. For example, in recent Centers for Disease Control (CDC) survey, it was shown that by routine sampling the indicators turbidity, free chlorine, and coliform had moderate degrees of association with cryptosporidium occurrence (Edberg, 1994; CDC, 1996). However, when the three were combined, the association was 92%. Fourth, there is no need in the sampling of tapwater to examine parameters that would indicate the water was consistent. However, for sources claiming to be protected, it is inherent that the physical and chemical nature of that water would be constant over time. Therefore, simple measurements such a temperature, ionic strength, etc, have great meaning in sampling source water, whereas they would have no meaning when sampling tapwater. Source Water can be adequately considered to be sequestered and safe from parasite intrusion if the following monitoring parameters and protocol are followed and yield negative results on a routine basis over a long period of time. Physicochemical measurements An ongoing historical record must be created of the physical and chemical parameters of source water, temperature, and ionic strength. These parameters should not correlate with surface water-related events. Biological indicators of fecal pollution E. coli - E. coli is a specific indicator of fecal contamination. Its presence in a water sample indicates health risk from parasites. If E. coli is detected and the sample was collected under aseptic technique, the water source is polluted. The half-life of E. coli is conservatively estimated to be at least 8 d under groundwater conditions. Therefore, it would be expected that recovery of E. coli from an occurrence event should be possible for weeks after the contamination event. However, one must account for the large dilutional factor present in an aquifer. Therefore, it is recommended that at least two 250 ml samples for E. coli be taken during its average half-life (i.e., two samples at least every 8 d). Alternatively, four 100 ml samples may be obtained during this period. Obviously, obtaining samples at the midpoint of the half-life and at the half-life would be optimum. A method such as the Colilert test that has been proven to be refractory to HPC interference should be used. Biological indicators of water quality Coliforms - Coliforms are present in large numbers in soil and the Earth's vegetation. While the presence in a water sample does not indicate an imminent health threat, it does indicate that the water source or borehole is at risk. Accordingly, the presence of a confirmed coliform in the source or processing should result in prompt action to examine the source water and borehole to determine the origin of the coliform presence. Should coliforms be found, the individual borehole should be placed off line and not used until the situation is remediated or explained and any cause of fecal contamination has been eliminated. The sampling frequency should be the same as for E. coli. A method such as the Colilert test, which is refractory to HPC interference should be used. Enterococci - Fecal streptococci such as E. coli are present in the colon of warm-blooded mammals. However, some species may be found in soil and may not be fecal specific. Should enterococci be found, the borehole should be placed off line and not used until the situation is remediated or explained and any cause of fecal contamination has been eliminated. The sampling frequency should be the same as for E. coli. Clostridium perfringens - There is little actual field data that demonstrate the usefulness of C. perfringens spores as a useful, fecal-specific indicator of groundwater. Spores can be found in soil without evidence of recent fecal contamination. However, the need for a long-lived indicator to accomplish extraordinary protection can be useful but the status of Cl. perfringens as an indicator needs careful assessment before being totally accepted. References 1. Amann, R.I., Krumholz, L. and Stahl, D.A (1990a) Fluorescent-oligonucleotide probing of whole cells for determinative, phylogenetic, and environmental studies in microbiology. Journal of Bacteriology 172, 762-770. 2. Amann, R.I., Binder, B.J., Olson, R.J., Chisbolm, S.W., Devereux, R. and Stahl, D.A. (1990b) Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixed microbial populations. Applied and Environmental Microbiology 56, 1919-1925. 3. Amann, R.I. (1995) Fluorescently labelled, rRNA-targeted oligonucleotide probes in the study of microbial ecology. Molecular Ecology 4, 543-553. 4. Anon (1985) Standard Methods for the Examination of Water and Wastewater, 16th edn, Washington DC: American Public Health Association, pp 853-917. 5. Balkwill, D.L., and Ghiorse, W.C. (1985) Characterization of subsurface bacteria associated with two shallow aquifers in Oklahoma. Applied and Environmental Microbiology 50, 580-588. 6. Balkwill, D.L. (1989) Numbers, diversity, and morphological characteristics of aerobic, chemoheterotrophic bacteria in deep subsurface sediments from a site in South Carolina. Geomicrobiology Journal 7, 33-52. 7. Bischofberger, T., Cha, s. K., Schmitt, R., König, B. and Schmidt-Lorenz, W. (1990) The bacterial flora of non-carbonated, natural mineral water from the springs to reservoir and glass and plastic bottles. International Journal of Food Microbiology 11, 51-72. 8. Centers for Disease Control (1996) Surveillance for waterborne-disease outbreaks United States, 1993-.1994. Morbid. Mortal. Week. Rep. 45(SS-1). 9. Chapelle, F.H. (1993) Ground-water microbiology & geochemistry, John Wiley & Sons, inc, New York. 10. Delabroise, A.M. and Ducluzeau, R. (1974) Le microbisme naturel de l'eau minérale, son développement, son innocuité sur l'organisme. Annales d'Hygiène, Médecine et Nutrition 10, 189-192. 11. Ducluzeau, R., Bochand, J.M. and Dufresne, S. (1976a) La microflore autochtone de l'eau minérale: nature, caractères physiologiques, signification hygiénique. Médecine et Nutrition 12 (2), 115-119. 12. Ducluzeau, R., Hudault, S. and Galpin, J.V. (1976c) Longevity of various bacterial strains of intestinal origin in gas-free mineral water. European Journal Applied Microbiology 3, 227-236. 13. Edberg, S.C. (1994) Do coliforms indicate the presence of protozoa? J. Ant. Water Works Assoc. 77, 4. 14. Edberg, S.C., Kops, S., Kontnick, C. and Escarzaga M. (1997) Analysis of cytotoxicity and invasiveness of heterotrophic plate count bacteria (HPC) isolated from drinking water on blood media. Journal of Applied Microbiology 82, 455-461. 15. Édberg, S.C., Leclerc H. and Robertson, J. (1997) Natural protection of spring and well drinking water against surface microbial contamination. II. Indicators and monitoring parameters for parasites. Critical Rev. Microbiol. 23, 179-206. 16. Fredrickson, J.K., Garland, T.R., Hicks, RJ., Thomas, J.M., Li, S.W. and McFadden S.M. (1989) Lithotrophic and heterotrophic bacteria in deep subsurface sediments and their relation to sediment properties. Geomicrobiology Journal 7, 53-66. 17. Fredrickson, J.K. and Phelps, T.J. (1997) Subsurface drilling and sampling, in Manual of environmental microbiology (eds C.J. Hurst and G.R. Knudsen), American Society for Microbiology, Washington, DC, pp 527-540. 18. Gonzalez, C., Gutierrez, C. and Grande, T. (1987) Bacterial flora in bottled uncarbonated mineral drinking water. Canadian Journal of Microbiology 33, 1120-1125. 19. Guillot, E. and Leclerc, H. (1993) Bacterial flora in natural mineral waters: characterization by ribosomal ribonucleic acid gene restriction patterns. Systematic and Applied Microbiology 16, 483-493. 20. Kersters, K., Ludwig, W., Vancanneyt, M., DeVos, P., Gillis, M. and Schleifer, K.H. (1997) Recent changes in the classification of the Pseudomonads: an overview. Systematic and Applied Microbiology 19, 465-477. 21. Kogure, K., Simidu, U., Taga, N. and Colwell, R.R. (1979) A tentative direct microscopic method for counting living marine bacteria. Canadian Journal Microbiology 25, 415-420. 22. Leclerc, H. (1980) Les qualités bactériologiques de l'eau minérale d'Evian, in De L'eau Minérale d'Evian Evian, France: S.A. des Eaux Minérales d'Evian, pp 24-27. 23. Lucas, F., and Ducluzeau, R. (1990a) Antagonistic role of various bacterial strains from the autochthonous flora of gas-free mineral water against Escherichia coli. Sciences des Aliments 10, 62-73. 24. Lucas, F., and Ducluzeau,. R..(1990b) Behavior of the autochtonous flora of mineral water from Vittel in the digestive tract of axenic mice, its antagonistic effect on Escherichia coli in vitro. Rivista Italiana d'Igiene 50, 383-393. 25. Madsen, EL. and Ghiorse, W.C. (1993) Groundwater microbiology: subsurface ecosystem processes, in Aquatic Microbiology: an ecological approach (ed T.E. Ford), Blackwell Scientific Publications, Boston, pp 167-213. 26. Manaia, C.M., Nunes, O.C., Morais, P.V. and Da Costa, M.S. (1990) Heterotrophic plate counts and the isolation of bacteria from mineral waters on selective and enrichment media. Journal of Applied Bacteriology 69, 871-876. 27. Morais, P.V. and Da Costa, M.S. (1990) Alterations in the major heterotrophic bacterial populations isolated from, a still bottled mineral water. Journal of Applied Bacteriology 69, 750-757. 28. Moreira, L., Agostinho, P., Morais, P.V. and Da Costa, M.S. (1994) Survival of allochthonous bacteria in still mineral water bottled in polyvinyl chloride (PVC) and glass. Journal of Applied Bacteriology 77, 334-339. 29. Morita, R.Y. (1982) Starvation-survival of heterotrophs in the marine environment. Advances in Microbiology and Ecology 6, 117-198. 30. Morita, R.Y. (1987) Starvation-survival: the normal mode of most bacteria. in the ocean, in Current Perspectives in Microbial Ecology (eds F. Megusar and M. Gantar), Slovene Soc. Microbiology, Ljubljana, Yugoslavia, pp 243-248. 31. Morita, R.Y. (1993) Bioavailability of energy and the starvation state, in Starvation in Bacteria (ed S. Kjelleberg), Plenum Press, New York and London, pp 1-23. 32. Mossel, D.A.A., Corry, LEL., Struijk, C.B. and Baird, R.M. (1995) Essentials of the microbiology of foods: a textbook advanced studies. John Wiley & Sons, Chichester. 33. Oliver, J.D. (1993) Formation of viable but nonculturable cells, in Starvation in Bacteria (ed S. Kjelleberg), Plenum Press, New York and London, pp 239-272. 34. .Reasoner, DJ. and GeIdreich, E.E. (1979) A new medium for the enumeration and subculture of bacteria from, potable water, in Abstract of the Annual Meeting of the American Society for Microbiology, American Society for Microbiology, Washington DC, pp 180. 35. Reasoner, D.J. and Geldreich, E.E. (1985) A new medium for the enumeration and subculture of bacteria from potable water. Applied and Environmental Microbiology 49, 1-7. 36. Reasoner, D.J. (1990) Monitoring heterotrophic bacteria in potable water, in Drinking water Microbiology (ed G.A. McFeters), Springer Verlag, New York, pp 452-477. 37. Rodriguez, G.G., Phipps, D., Ishiguro, K. and Ridgway, H.F. (1992) Use of a fluorescent redox probe for direct visualization of actively respiring bacteria. Applied and Environmental Microbiology 58, 1801--1808. 38. Schwaller, P. and Schmidt-Lorenz, W. (1981a) Flore microbienne de quatre eaux minérales non gazéifiées et mises en bouteilles. II. Les Pseudomonas et autres bactéries à Gram négatif - Composition fine de la flore. Zentralblatt fur Bakteriologie and Hygiene L Abt. Orig.C2, 179-196. 39. Schwaller, P. and Schmidt-Lorenz, W. (1981b) Dénombrements de colonies bactériennes dans les eaux minérales naturelles non gazéifiées et mises en bouteilles. 1re communication: Etude de méthodes spécifiques pour le dénombrement de colonies bactériennes. Brauerei-Rundschau 92, 281-285. 40. Vachée, A., and Leclerc, H. (1995) Propriétés antagonistes de la flore autochtone des eaux minérales naturelles vis-à-vis de Pseudomonas aeruginosa. Journal Européen d'Hydrologie 26, 327-338. 41. Vachée, A., Vincent P., Struijk, C.B., Mossel, D.A.A. and Leclerc, H. (1997) A study of the fate of the autochtonous bacterial flora of still mineral waters by analysis of restriction fragment length polymorphism of genes coding for rRNA. Systematic and Applied Microbiology (accepted for publication). 42. Wagner, M., Amann, R., Lemmer, H. and Schleifer, K.H. (1993) Probing activated sludge with oligonucleotide specific for proteobacteria: inadequacy of culture-dependent methods for describing microbial community structure. Applied and Environmental Microbiology 59, 1520-1525. 43. Wallner, G., Amann, R. and Beisker, W. (1993) Optimizing fluorescent in situ hybridization with rRNA-targeted oligonucleotide probes for flow cytometric identification of microorganisms. Cytometry 14, 136-143. 44. Wallner, G., Erhart, R. and Amann, R. (1995) Flow cytometric analysis of activated sludge with rRNA-targeted probes. Applied and Environmental Microbiology 61, 1859-1866. 45. Wallner, G., Steinmetz, I., Bitter-Suermann, D. and Amann, R. (1996) Combination of rRNA-targeted hybridization probes and immuno-probes for the identification of bacteria by flow cytometry. Systematic and Applied Microbiology 19,569-576. 46. Wilson, N. (1995) Soil Water and Ground Water Sampling. CRC Press, Inc., Boca Raton, Fla. 47. Yurdusev, N. and Ducluzeau, R. (1985) Qualitative and quantitative development of the bacterial flora of Vittel mineral water in glass or plastic bottles. Sciences des Aliments 5, 231-238. 48. Zimmerman, R., Iturriaga, R. and Becker-Birck, J. (1978) Simultaneous determination of the total number of aquatic bacteria and the number thereof involved in respiration. Applied and Environmental Microbiology 36, 926-935.