Application of the fluorescent-antibody technique to the study of a methanogenic bacterium in lake sediments.

Fluorescent antibody (FA) was prepared for a methanogenic bacterium isolated from Wintergreen Lake pelagic sediment. The isolate resembles Methanobacterium formicicum. The FA did not cross-react with 9 other methanogens, including M. formicicum strains, or 24 heterotrophs, 18 of which had been isolated from Wintergreen Lake sediment. FA-reacting methanogens were detected in heat-fixed smears of several different lake sediments and anaerobic sewage sludge. Pretreatment of all samples with either rhodamine-conjugated geletin or bovine serum albumin adequately controlled nonspecific absorption of the FA. Autofluorescent particles were observed in the sediment samples but, with experience, they could easily be distinguished from FA-reacting bacteria. FA direct counts of the specific methanogen in Wintergreen Lake sediments were made on four different sampling dates and compared with five-tube most-probable-number estimates of the total methanogenic population that was present in the same samples. The FA counts ranged from 3.1 X 10(6) to 1.4 X 10(7)/g of dry sediment. The highest most-probable-number estimates were at least an order ofmagnitude lower.     

Application of the fluorescent-antibody technique to the study of a methanogenic bacterium in lake sediments.

Fluorescent antibody (FA) was prepared for a methanogenic bacterium isolated from Wintergreen Lake pelagic sediment. The isolate resembles Methanobacterium formicicum. The FA did not cross-react with 9 other methanogens, including M. formicicum strains, or 24 heterotrophs, 18 of which had been isolated from Wintergreen Lake sediment. FA-reacting methanogens were detected in heat-fixed smears of several different lake sediments and anaerobic sewage sludge. Pretreatment of all samples with either rhodamine-conjugated geletin or bovine serum albumin adequately controlled nonspecific absorption of the FA. Autofluorescent particles were observed in the sediment samples but, with experience, they could easily be distinguished from FA-reacting bacteria. FA direct counts of the specific methanogen in Wintergreen Lake sediments were made on four different sampling dates and compared with five-tube most-probable-number estimates of the total methanogenic population that was present in the same samples. The FA counts ranged from 3.1 X 10(6) to 1.4 X 10(7)/g of dry sediment. The highest most-probable-number estimates were at least an order ofmagnitude lower.     

Culturable Bacteria in Subglacial Sediments and Ice from Two Southern Hemisphere Glaciers

Viable prokaryotes have been detected in basal sediments beneath the few Northern Hemisphere glaciers that have been sampled for microbial communities. However, parallel studies have not previously been conducted in the Southern Hemisphere, and subglacial environments in general are a new and underexplored niche for microbes. Unfrozen subglacial sediments and overlying glacier ice samples collected aseptically from the Fox Glacier and Franz Josef Glacier in the Southern Alps of New Zealand now have been shown to harbor viable microbial populations. Total direct counts of 2–7 × 10⁶ cells g^–1 dry weight sediment were observed, whereas culturable aerobic heterotrophs ranged from 6–9 × 10⁵ colony-forming units g^–1 dry weight. Viable counts in the glacier ice typically were 3–4 orders of magnitude smaller than in sediment. Nitrate-reducing and ferric iron–reducing bacteria were detected in sediment samples from both glaciers, but were few or below detection limits in the ice samples. Nitrogen-fixing bacteria were detected only in the Fox Glacier sediment. Restriction fragment analysis of 16S rDNA amplified from 37 pure cultures of aerobic heterotrophs capable of growth at 4°C yielded 23 distinct groups, of which 11 were identified as -Proteobacteria. 16S rDNA sequences from representatives of these 11 groups were analyzed phylogenetically and shown to cluster with bacteria such as Polaromonas vacuolata and Rhodoferax antarcticus, or with clones obtained from permanently cold environments. Chemical analysis of sediment and ice samples revealed a dilute environment for microbial life. Nevertheless, both the sediment samples and one ice sample demonstrated substantial aerobic mineralization of ¹⁴C-acetate at 8°C, indicating that sufficient nutrients and viable psychrotolerant microbes were present to support metabolism. Unfrozen subglacial sediments may represent a significant global reservoir of biological activity with the potential to influence glacier meltwater chemistry.     

Comparison of homogenizing, shaking, and blending on the recovery of microorganisms and endotoxins from fresh and frozen ground beef as assessed by plate counts and the Limulus amoebocyte lysate test.

Of three methods studied, brisk shaking of samples in dilution blanks by hand and homogenization by a stomacher were compared relative to their capacity to recover the endotoxins and viable bacteria; blending with a Waring blender was compared with these two methods only on the recovery of viable cells. Aerobic plate counts were essentially the same by the three methods for fresh meats, with the stomacher producing slightly higher aerobic plate counts and significantly higher gram-negative counts determined by violet red bile agar. The stomacher produced significantly higher aerobic plate counts and violet red bile agar results on frozen meats than did shaking. Endotoxins were determined by the Limulus amoebocyte lysate test; results by shaking and stomacher on 15 single samples of frozen meat were identical. Of Limulus amoebocyte lysate-negative beef which was spiked with known endotoxin, a higher percentage of recovery was obtained with the stomacher. Although both aerobic plate counts and violet red bile agar counts were found by shaking and stomacher to decrease significantly in frozen meats, endotoxin content was not significantly affected. The stomacher was found to be the better method overall, especially when meats are to be examined for their content of viable gram-negative bacteria, endotoxins, or both.     

Comparison of homogenizing, shaking, and blending on the recovery of microorganisms and endotoxins from fresh and frozen ground beef as assessed by plate counts and the Limulus amoebocyte lysate test.

Of three methods studied, brisk shaking of samples in dilution blanks by hand and homogenization by a stomacher were compared relative to their capacity to recover the endotoxins and viable bacteria; blending with a Waring blender was compared with these two methods only on the recovery of viable cells. Aerobic plate counts were essentially the same by the three methods for fresh meats, with the stomacher producing slightly higher aerobic plate counts and significantly higher gram-negative counts determined by violet red bile agar. The stomacher produced significantly higher aerobic plate counts and violet red bile agar results on frozen meats than did shaking. Endotoxins were determined by the Limulus amoebocyte lysate test; results by shaking and stomacher on 15 single samples of frozen meat were identical. Of Limulus amoebocyte lysate-negative beef which was spiked with known endotoxin, a higher percentage of recovery was obtained with the stomacher. Although both aerobic plate counts and violet red bile agar counts were found by shaking and stomacher to decrease significantly in frozen meats, endotoxin content was not significantly affected. The stomacher was found to be the better method overall, especially when meats are to be examined for their content of viable gram-negative bacteria, endotoxins, or both.     

Attached and unattached bacterial communities in a 120-meter corehole in an acidic, crystalline rock aquifer

The bacteria colonizing geologic core sections (attached) were contrasted with those found suspended in the groundwater (unattached) by examining the microbiology of 16 depth-paired core and groundwater samples using a suite of culture-independent and culture-dependent analyses. One hundred twenty-two meters was continuously cored from a buried chalcopyrite ore hosted in a biotite-quartz-monzonite porphyry at the Mineral Park Mine near Kingman, Ariz. Every fourth 1.5-m core was acquired using microbiologically defensible methods, and these core sections were aseptically processed for characterization of the attached bacteria. Groundwater samples containing unattached bacteria were collected from the uncased corehole at depth intervals corresponding to the individual cores using an inflatable straddle packer sampler. The groundwater was acidic (pH 2.8 to 5.0), with low levels of dissolved oxygen and high concentrations of sulfate and metals, including ferrous iron. Total numbers of attached cells were less than 10(5) cells g of core material(-1) while unattached cells numbered about 10(5) cells ml of groundwater(-1). Attached and unattached acidophilic heterotrophs were observed throughout the depth profile. In contrast, acidophilic chemolithotrophs were not found attached to the rock but were commonly observed in the groundwater. Attached communities were composed of low numbers (<40 CFU g(-1)) of neutrophilic heterotrophs that exhibited a high degree of morphologic diversity, while unattached communities contained higher numbers (ca. 10(3) CFU ml(-1)) of neutrophilic heterotrophs of limited diversity. Sulfate-reducing bacteria were restricted to the deepest samples of both core and groundwater. 16S ribosomal DNA sequence analysis of attached, acidophilic isolates indicated that organisms closely related to heterotrophic, acidophilic mesophiles such as Acidiphilium organovorum and, surprisingly, to the moderately thermophilic Alicyclobacillus acidocaldarius were present. The results indicate that viable (but possibly inactive) microorganisms were present in the buried ore and that there was substantial distinction in biomass and physiological capabilities between attached and unattached populations.     

Methane oxidation and the competition for oxygen in the rice rhizosphere

A mechanistic approach is presented to describe oxidation of the greenhouse gas methane in the rice rhizosphere of flooded paddies by obligate methanotrophic bacteria. In flooded rice paddies these methanotrophs compete for available O(2) with other types of bacteria. Soil incubation studies and most-probable-number (MPN) counts of oxygen consumers show that microbial oxygen consumption rates were dominated by heterotrophic and methanotrophic respiration. MPN counts of methanotrophs showed large spatial and temporal variability. The most abundant methanotrophs (a Methylocystis sp.) and heterotrophs (a Pseudomonas sp. and a Rhodococcus sp.) were isolated and characterized. Growth dynamics of these bacteria under carbon and oxygen limitations are presented. Theoretical calculations based on measured growth dynamics show that methanotrophs were only able to outcompete heterotrophs at low oxygen concentrations (frequently < 5 microM). The oxygen concentration at which methanotrophs won the competition from heterotrophs did not depend on methane concentration, but it was highly affected by organic carbon concentrations in the paddy soil. Methane oxidation was severely inhibited at high acetate concentrations. This is in accordance with competition experiments between Pseudomonas spp. and Methylocystis spp. carried out at different oxygen and carbon concentrations. Likely, methane oxidation mainly occurs at microaerophilic and low-acetate conditions and thus not directly at the root surface. Acetate and oxygen concentrations in the rice rhizosphere are in the critical range for methane oxidation, and a high variability in methane oxidation rates is thus expected.     

Application of laser scanning for the rapid and automated detection of bacteria in water samples

It is widely accepted that the heterotrophic plate count method may not support the growth of all viable bacteria which may be present within a water sample and that alternative procedures using 'viability markers' may yield additional information. In this study, ChemChrome B (CB), which is converted to a fluorescent product by esterase activity, was used to stain viable bacteria (captured by membrane filtration) from potable water samples. The labelled bacteria from each sample were subsequently enumerated using a novel laser scanning instrument (ChemScan). Analysis of 107 potable water samples using this procedure demonstrated the presence of a significantly greater number of bacteria than were detected by culture (z-test, P < 0.05). The mean number of bacteria isolated by culture on R2A agar incubated at 22 degrees C for 7 d was only 25.2% of the total number of viable bacteria detected using the CB/ChemScan viability assay. Further analysis of 81 water samples using a 5-cyano-2,3,4-tolyl-tetrazolium chloride (CTC) viability assay also demonstrated the presence of many viable bacteria which were not capable of growth under the culture conditions employed in this study. However, the results indicate that ChemChrome B has the ability to stain a significantly greater number of heterotrophs than CTC (z-test, P < 0.05). In contrast, six potable waters were identified in which the CTC viability assay resulted in counts greater than those obtained using CB. The ChemScan instrument was successfully used for rapid and accurate enumeration of labelled micro-organisms, allowing information on the total viable microbial load of a water sample to be determined within 1 h. Furthermore, the ChemScan system has the potential for use in detecting specific organisms labelled with fluorescently-labelled antibodies or nucleic acid probes.     

Evaluation of media and techniques to enumerate heterotrophic microbes from karst and sand aquifer springs

Several media and techniques were compared for their efficiency to enumerate viable heterotrophs from both a karst and sand aquifer spring. A medium designed to enumerate bacteria from nutrient-poor waters (HCFU) as well as R2A medium proved superior to tryptic soy agar; however, the difference was always less than one order of magnitude. Membrane filtration resulted in lower counts of microbes than the spread plate, multitube turbidity, or drop plate methods from samples of both sand and karst springs. The drop plate technique yielded higher viable counts from the sand spring and basin of the karst spring, with a precision of 21% (coefficient of variation) and a maximum plating efficiency of 3.4% (viable count/direct count × 100). Subsequently, 63% of isolates from drop plates were recovered on HCFU. Microcolonies were visible by epifluorescence microscopy, acridine orange staining, and subsequent examination of excised agar sections containing drops. Correspondence to: A.T. Mikell, Jr.     

The effect of cleaning and disinfecting the sampling well on the microbial communities of deep subsurface water samples

Our knowledge of the microbial characteristics of deep subsurface waters is currently very limited, mainly because of the methods used to collect representative microbial samples from such environments. In order to improve this procedure, a protocol designed to remove the unspecific, contaminant biofilm present on the walls of an approximately 800 m deep well is proposed. This procedure included extensive purges of the well, a mechanical cleaning of its wall, and three successive chlorine injections to disinfect the whole line before sampling. Total bacterial counts in water samples decreased from 2.5 x 10(5) to 1.0 x 10(4) per millilitre during the cleaning procedure. Culture experiments showed that the first samples were dominated by sulfate-reducers and heterotrophs, whereas the final sample was dominated by oligotrophic and hydrogenotrophic bacteria. Community structures established on the diversity of the 16S rRNA genes and data analysis revealed that the water sample collected, after a purge without removal of the biofilm, was characterized by numerous phyla which are not representative of the deep subsurface water. On the other hand, several bacterial phyla were only detected after the full cleaning of the well, and were considered as important components of the subsurface ecosystem which would have been missed in the absence of well cleaning.     

Attached and Unattached Bacterial Communities in a 120-Meter Corehole in an Acidic, Crystalline Rock Aquifer

The bacteria colonizing geologic core sections (attached) were contrasted with those found suspended in the groundwater (unattached) by examining the microbiology of 16 depth-paired core and groundwater samples using a suite of culture-independent and culture-dependent analyses. One hundred twenty-two meters was continuously cored from a buried chalcopyrite ore hosted in a biotite-quartz-monzonite porphyry at the Mineral Park Mine near Kingman, Ariz. Every fourth 1.5-m core was acquired using microbiologically defensible methods, and these core sections were aseptically processed for characterization of the attached bacteria. Groundwater samples containing unattached bacteria were collected from the uncased corehole at depth intervals corresponding to the individual cores using an inflatable straddle packer sampler. The groundwater was acidic (pH 2.8 to 5.0), with low levels of dissolved oxygen and high concentrations of sulfate and metals, including ferrous iron. Total numbers of attached cells were less than 10⁵ cells g of core material⁻¹ while unattached cells numbered about 10⁵ cells ml of groundwater⁻¹. Attached and unattached acidophilic heterotrophs were observed throughout the depth profile. In contrast, acidophilic chemolithotrophs were not found attached to the rock but were commonly observed in the groundwater. Attached communities were composed of low numbers (<40 CFU g⁻¹) of neutrophilic heterotrophs that exhibited a high degree of morphologic diversity, while unattached communities contained higher numbers (ca. 10³ CFU ml⁻¹) of neutrophilic heterotrophs of limited diversity. Sulfate-reducing bacteria were restricted to the deepest samples of both core and groundwater. 16S ribosomal DNA sequence analysis of attached, acidophilic isolates indicated that organisms closely related to heterotrophic, acidophilic mesophiles such as Acidiphilium organovorum and, surprisingly, to the moderately thermophilic Alicyclobacillus acidocaldarius were present. The results indicate that viable (but possibly inactive) microorganisms were present in the buried ore and that there was substantial distinction in biomass and physiological capabilities between attached and unattached populations.     

Sampling and Pooling Methods for Capturing Herd Level Antibiotic Resistance in Swine Feces using qPCR and CFU Approaches

The aim of this article was to define the sampling level and method combination that captures antibiotic resistance at pig herd level utilizing qPCR antibiotic resistance gene quantification and culture-based quantification of antibiotic resistant coliform indicator bacteria. Fourteen qPCR assays for commonly detected antibiotic resistance genes were developed, and used to quantify antibiotic resistance genes in total DNA from swine fecal samples that were obtained using different sampling and pooling methods. In parallel, the number of antibiotic resistant coliform indicator bacteria was determined in the same swine fecal samples. The results showed that the qPCR assays were capable of detecting differences in antibiotic resistance levels in individual animals that the coliform bacteria colony forming units (CFU) could not. Also, the qPCR assays more accurately quantified antibiotic resistance genes when comparing individual sampling and pooling methods. qPCR on pooled samples was found to be a good representative for the general resistance level in a pig herd compared to the coliform CFU counts. It had significantly reduced relative standard deviations compared to coliform CFU counts in the same samples, and therefore differences in antibiotic resistance levels between samples were more readily detected. To our knowledge, this is the first study to describe sampling and pooling methods for qPCR quantification of antibiotic resistance genes in total DNA extracted from swine feces.     

A Note: Comparison of different homogenization procedures for detecting Campylobacter spp. in sewage sludge

C. HÖLLER AND U. SCHOMAKERS-REVAKA. 1994. Crude sewage sludge contains Campylobacter spp. in a concentration of 10^-1–10³ cfu 100 ml^-1 on average. Because large variations in the number of bacteria are seen when samples are examined in parallel, we attempted to improve the detection method. Seeded sewage sludge samples were homogenized by a high-speed blender, ultrasonic bath and ultrasonic bar. Bacterial counts were determined by the MPN method in triplicate. The recovery rate was < 10%. Subsequently, sludge samples without artificial contamination were also examined. The bacterial counts varied considerably, as seen earlier. In order to enhance the detection rate of campylobacters homogenization times and frequencies were increased, samples were diluted prior to treatment and pre-enriched in non-selective broth or supplemented with detergent. None of the methods applied proved satisfactory. The bacterial counts achieved with all methods varied greatly, with minimum and maximum values lying at least two orders of magnitude apart.     

Effect of freezing of water samples on viable counts of environmental mycobacteria

The effect of prolonged storage on mycobacteria and other heterotrophic bacteria in brook water samples was examined by determination of viable counts from fresh samples and again after water concentrates had been stored in nutrient broth at —75°C for 15 months. The counts of mycobacteria were on average three times higher after storage (range of ratio 0·9–10·4). In contrast, the viable counts of other heterotrophic bacteria were reduced by 69%. The increase in mycobacterial counts was probably due to break-up of microcolonies or release of attached bacteria from particles. The possibility of cultivating mycobacteria from frozen samples is of practical help in large-scale field surveys.     

Use of fluorochromes for direct enumeration of total bacteria in environmental samples: past and present.

Understanding the role of bacteria in microbial food webs is intimately connected to the methods applied in the direct enumeration of bacteria. We have examined over 220 papers describing studies in which fluorochrome staining followed by epifluorescent microscopic direct counts was used to estimate total bacterial abundances. In this review, we summarize patterns in the use of 3,6-bis[dimethylamino]acridinium chloride (acridine orange) and 4',6-diamidino-2-phenylindole (DAPI), the two stains most frequently used in bacterial enumeration. The staining of samples with these fluorochromes, followed by filtration and direct counting of bacterial cells on filter surfaces, has become routine over the past 10 years. We examine trends in features of the standard direct count methods, such as sample preservation and preparation techniques, membrane filter types used, applied stain concentrations, duration of staining, and counting strategies, in relation to the types of samples being examined. The high variability in bacterial counts observed within similar sample types may be partially accounted for by differences in methods. Synthesizing review findings, we include a recommended method for the direct enumeration of bacteria in environmental samples.     

Methane Oxidation and the Competition for Oxygen in the Rice Rhizosphere

A mechanistic approach is presented to describe oxidation of the greenhouse gas methane in the rice rhizosphere of flooded paddies by obligate methanotrophic bacteria. In flooded rice paddies these methanotrophs compete for available O₂ with other types of bacteria. Soil incubation studies and most-probable-number (MPN) counts of oxygen consumers show that microbial oxygen consumption rates were dominated by heterotrophic and methanotrophic respiration. MPN counts of methanotrophs showed large spatial and temporal variability. The most abundant methanotrophs (a Methylocystis sp.) and heterotrophs (a Pseudomonas sp. and a Rhodococcus sp.) were isolated and characterized. Growth dynamics of these bacteria under carbon and oxygen limitations are presented. Theoretical calculations based on measured growth dynamics show that methanotrophs were only able to outcompete heterotrophs at low oxygen concentrations (frequently <5 μM). The oxygen concentration at which methanotrophs won the competition from heterotrophs did not depend on methane concentration, but it was highly affected by organic carbon concentrations in the paddy soil. Methane oxidation was severely inhibited at high acetate concentrations. This is in accordance with competition experiments between Pseudomonas spp. and Methylocystis spp. carried out at different oxygen and carbon concentrations. Likely, methane oxidation mainly occurs at microaerophilic and low-acetate conditions and thus not directly at the root surface. Acetate and oxygen concentrations in the rice rhizosphere are in the critical range for methane oxidation, and a high variability in methane oxidation rates is thus expected.     

Population Sizes and Growth Pressure Responses of Intestinal Microfloras of Deep-Sea Fish Retrieved from the Abyssal Zone

The intestinal floras of seven deep-sea fish retrieved at depths of from 3,200 to 5,900 m were examined for population sizes and growth responses to pressure. Large populations of culturable bacteria, ranging from 1.1 x 10(sup6) to 3.6 x 10(sup8) cells per ml of contents, were detected when samples were incubated at conditions characteristic of those of the deep sea. Culturable cell counts at in situ pressures were greater than those at atmospheric pressure in all samples. Most of the strains isolated by the spread-plating method at atmospheric pressure later proved barophilic. Barophilic bacteria were the predominant inhabitants of the abyssal fish intestines.     

Microbiological Comparisons within and across Contiguous Lacustrine, Paleosol, and Fluvial Subsurface Sediments

Twenty-six subsurface samples were collected from a borehole at depths of 173.3 to 196.8 m in the saturated zone at the Hanford Site in south-central Washington State. The sampling was performed throughout strata that included fine-grained lacustrine (lake) sediments, a paleosol (buried soil) sequence, and coarse-grained fluvial (river) sediments. A subcoring method and tracers were used to minimize and quantify contamination to obtain samples that were representative of subsurface strata. Sediment samples were tested for total organic carbon, inorganic carbon, total microorganisms by direct microscopic counts, culturable aerobic heterotrophs by plate counts, culturable anaerobes by most-probable-number enumeration, basal respiration rates, and mineralization of (sup14)C-labeled glucose and acetate. Total direct microscopic counts of microorganisms were low, ranging from below detection to 1.9 x 10(sup5) cells g (dry weight)(sup-1). Culturable aerobes and anaerobes were below minimum levels of detection in most samples. Direct microscopic counts, basal respiration rates, and (sup14)C-glucose mineralization were all positively correlated with total organic carbon and were highest in the lacustrine sediments. In contrast to previous subsurface studies, these saturated-zone samples did not have higher microbial abundance and activities than unsaturated sediments sampled from the same borehole, the fine-textured lacustrine sediment had higher microbial numbers and activities than the coarse-textured fluvial sands, and the paleosol samples did not have higher biomass and activities relative to the other sediments. The results of this study expand the subsurface microbiology database to include information from an environment very different from those previously studied.     

A flow cytometric technique for quantification and differentiation of bacteria in bulk tank milk

AIMS: The present study describes a flow cytometric technique for quantification and differentiation of bacteria in bulk tank milk according to the main cause of elevated counts. METHODS AND RESULTS: A total of 75 Danish bulk tank milk samples exceeding the grading level of 3.0 x 10(4) CFU ml(-1) were examined by both flow cytometry and traditional microbiological analyses. The correlation coefficient (r) between the two methods was 0.71. For the differential analyses of the dominant bacterial populations four different parameters were used to give a species-characteristic pattern. The four parameters were as follows: staining with Oregon Green conjugated wheat germ agglutinin that binds to the cell wall of bacteria, staining with hexidium iodide that binds to all bacterial DNA, the flow cytometric forward scatter and the flow cytometric side scatter. Three regions in the flow cytometric plot were defined: region 1 includes bacteria mainly associated with poor hygiene, region 2 includes psychrotrophic hygiene bacteria and region 3 includes bacteria mainly related to mastitis. The ability of the flow cytometric technique to predict the main cause of elevated bacterial counts on routine samples was examined. Comparing these results with results obtained by traditional microbiological analyses for identification showed that for 81% of the samples the two techniques agreed on the main cause of an elevated bacterial count. CONCLUSIONS: The ability of the presented flow cytometric technique to enumerate and differentiate bacteria in bulk tank milk according to the main cause of elevated counts was demonstrated. SIGNIFICANCE AND IMPACT OF THE STUDY: This study described the first step in development of a technique suitable for routine analyses of bulk tank milk samples. A technique indicating the main cause of an elevated count will enable the farmer to eliminate the contamination source within a short time limit.     

Comparison of m-Endo LES, MacConkey, and Teepol media for membrane filtration counting of total coliform bacteria in water.

Total coliform counts obtained by means of standard membrane filtration techniques, using MacConkey agar, m-Endo LES agar, Teepol agar, and pads saturated with Teepol broth as growth media, were compared. Various combinations of these media were used in tests on 490 samples of river water and city wastewater after different stages of conventional purification and reclamation processes including lime treatment, and filtration, active carbon treatment, ozonation, and chlorination. Endo agar yielded the highest average counts for all these samples. Teepol agar generally had higher counts then Teepol broth, whereas MacConkey agar had the lowest average counts. Identification of 871 positive isolates showed that Aeromonas hydrophila was the species most commonly detected. Species of Escherichia, Citrobacter, Klebsiella, and Enterobacter represented 55% of isolates which conformed to the definition of total coliforms on Endo agar, 54% on Teepol agar, and 45% on MacConkey agar. Selection for species on the media differed considerably. Evaluation of these data and literature on alternative tests, including most probable number methods, indicated that the technique of choice for routine analysis of total coliform bacteria in drinking water is membrane filtration using m-Endo LES agar as growth medium without enrichment procedures or a cytochrome oxidase restriction.