Multivariate logistic regression for predicting total culturable virus presence at the intake of a potable-water treatment plant: novel application of the atypical coliform/total coliform ratio

Predicting the presence of enteric viruses in surface waters is a complex modeling problem. Multiple water quality parameters that indicate the presence of human fecal material, the load of fecal material, and the amount of time fecal material has been in the environment are needed. This paper presents the results of a multiyear study of raw-water quality at the inlet of a potable-water plant that related 17 physical, chemical, and biological indices to the presence of enteric viruses as indicated by cytopathic changes in cell cultures. It was found that several simple, multivariate logistic regression models that could reliably identify observations of the presence or absence of total culturable virus could be fitted. The best models developed combined a fecal age indicator (the atypical coliform [AC]/total coliform [TC] ratio), the detectable presence of a human-associated sterol (epicoprostanol) to indicate the fecal source, and one of several fecal load indicators (the levels of Giardia species cysts, coliform bacteria, and coprostanol). The best fit to the data was found when the AC/TC ratio, the presence of epicoprostanol, and the density of fecal coliform bacteria were input into a simple, multivariate logistic regression equation, resulting in 84.5% and 78.6% accuracies for the identification of the presence and absence of total culturable virus, respectively. The AC/TC ratio was the most influential input variable in all of the models generated, but producing the best prediction required additional input related to the fecal source and the fecal load. The potential for replacing microbial indicators of fecal load with levels of coprostanol was proposed and evaluated by multivariate logistic regression modeling for the presence and absence of virus.     

Comparative Analysis of Modeling Techniques for Coliform Organisms in Streams

The use of models for predicting changes in water quality parameters is currently considered an integral part of river basin management. The application of modeling techniques to coliform organisms is in its infancy due to the complexities involved and the lack of definitive information on coliform populations in natural environments. The purpose of this study was to make a comparative analysis of the available models for coliform organisms in order to improve on the state of the art of this subject. The available coliform models may be classified into deterministic or statistical types. In this study, six different models, three of each type, were selected for analysis and were applied to coliform data available on the Leaf River. Results of comparing the models indicated that a deterministic model was best suited for total coliform and a statistical model was best suited for fecal coliform. Ultimate selection of a model for coliform organisms is dependent not only on the accuracy of the model but on ease of implementation. Current technology would probably dictate the use of a deterministic model because of the lack of a complete data base on which to base statistical models.     

Optimum Membrane Structures for Growth of Coliform and Fecal Coliform Organisms

The purpose of this study was to determine the optimum membrane filter structure and characteristics for recovery of coliform organisms. Additionally, other factors such as sterilization method and membrane composition were examined. Fecal coliform growth tests with varied samples indicated that the most critical factor in recovery was surface pore morphology and not other factors previously suspected. Fecal coliform counts showed a dramatic increase, with increasing surface opening sizes. Membrane structures with surface openings large enough to surround the entrapped bacteria are required for optimum growth of fecal coliform organisms. Maximum fecal coliform recoveries are obtained using membranes composed of mixed esters of cellulose exhibiting a surface opening diameter of 2.4 μm and a retention pore size of 0.7 μm.     

Virus occurrence in municipal groundwater sources in Quebec, Canada

A 1 year study was undertaken on groundwater that was a source of drinking water in the province of Quebec, Canada. Twelve municipal wells (raw water) were sampled monthly during a 1 year period, for a total of 160 samples. Using historic data, the 12 sites were categorized into 3 groups: group A (no known contamination), group B (sporadically contaminated by total coliforms), and group C (historic and continuous contamination by total coliforms and (or) fecal coliforms). Bacterial indicators (total coliform, Escherichia coli, enteroccoci), viral indicators (somatic and male-specific coliphages), total culturable human enteric viruses, and noroviruses were analyzed at every sampling site. Total coliforms were the best indicator of microbial degradation, and coliform bacteria were always present at the same time as human enteric viruses. Two samples contained human enteric viruses but no fecal pollution indicators (E. coli, enterococci, or coliphages), suggesting the limited value of these microorganisms in predicting the presence of human enteric viruses in groundwater. Our results underline the value of historic data in assessing the vulnerability of a well on the basis of raw water quality and in detecting degradation of the source. This project allowed us to characterize the microbiologic and virologic quality of groundwater used as municipal drinking water sources in Quebec.     

Statistical Analysis of the Recovery of Coliform Organisms on Gelman and Millipore Membrane Filters

The recovery of coliform organisms on Gelman and Millipore membranes was analyzed by using both a model I (which assumes no error in the x variable) and model II (which allows errors in both the variables) regression analysis. The two models afford estimates of the slope which agree within their 95% confidence limits. Using equations derived in this paper, the model II confidence limits on the intercept are obtained. This range does not include the model I intercept limits, thereby demonstrating the differences between results from an incorrect (model I) and correct (model II) approach. In addition, fecal coliform show no differences in response to the two membranes, whereas total coliform exhibit higher recoveries on Gelman membranes.     

Influence of Coliform Source on Evaluation of Membrane Filters

Four brands of membrane filters were examined for total and fecal coliform recovery performance by two experimental approaches. Using diluted EC broth cultures of water samples, Johns-Manville filters were superior to Sartorius filters for fecal coliform but equivalent for total coliform recovery. Using river water samples, Johns-Manville filters were superior to Sartorius filters for total coliform but equivalent for fecal coliform recovery. No differences were observed between Johns-Manville and Millipore or Millipore and Sartorius filters for total or fecal coliform recoveries using either approach, nor was any difference observed between Millipore and Gelman filters for fecal coliform recovery from river water samples. These results indicate that the source of the coliform bacteria has an important influence on the conclusions of membrane filter evaluation studies.     

Coliform Bacteria and Nitrogen Fixation in Pulp and Paper Mill Effluent Treatment Systems

The majority of pulp and paper mills now biotreat their combined effluents using activated sludge. On the assumption that their wood-based effluents have negligible fixed N, and that activated-sludge microorganisms will not fix significant N, these mills routinely spend large amounts adding ammonia or urea to their aeration tanks (bioreactors) to permit normal biomass growth. N₂ fixation in seven Eastern Canadian pulp and paper mill effluent treatment systems was analyzed using acetylene reduction assays, quantitative nitrogenase (nifH) gene probing, and bacterial isolations. In situ N₂ fixation was undetectable in all seven bioreactors but was present in six associated primary clarifiers. One primary clarifier was studied in greater detail. Approximately 50% of all culturable cells in the clarifier contained nifH, of which >90% were Klebsiella strains. All primary-clarifier coliform bacteria growing on MacConkey agar were identified as klebsiellas, and all those probed contained nifH. In contrast, analysis of 48 random coliform isolates from other mill water system locations showed that only 24 (50%) possessed the nifH gene, and only 13 (27%) showed inducible N₂-fixing activity. Thus, all the pulp and paper mill primary clarifiers tested appeared to be sites of active N₂ fixation (0.87 to 4.90 mg of N liter⁻¹ day⁻¹) and a microbial community strongly biased toward this activity. This may also explain why coliform bacteria, especially klebsiellas, are indigenous in pulp and paper mill water systems.     

Comparison of Fecal Coliform Agar and Violet Red Bile Lactose Agar for Fecal Coliform Enumeration in Foods

A 24-h direct plating method for fecal coliform enumeration with a resuscitation step (preincubation for 2 h at 37 ± 1°C and transfer to 44 ± 1°C for 22 h) using fecal coliform agar (FCA) was compared with the 24-h standardized violet red bile lactose agar (VRBL) method. FCA and VRBL have equivalent specificities and sensitivities, except for lactose-positive non-fecal coliforms such as Hafnia alvei, which could form typical colonies on FCA and VRBL. Recovery of cold-stressed Escherichia coli in mashed potatoes on FCA was about 1 log unit lower than that with VRBL. When the FCA method was compared with standard VRBL for enumeration of fecal coliforms, based on counting carried out on 170 different food samples, results were not significantly different (P > 0.05). Based on 203 typical identified colonies selected as found on VRBL and FCA, the latter medium appears to allow the enumeration of more true fecal coliforms and has higher performance in certain ways (specificity, sensitivity, and negative and positive predictive values) than VRBL. Most colonies clearly identified on both media were E. coli and H. alvei, a non-fecal coliform. Therefore, the replacement of fecal coliform enumeration by E. coli enumeration to estimate food sanitary quality should be recommended.     

Isolation of Fecal Coliform Bacteria from the Diamondback Terrapin (Malaclemys terrapin centrata)

Total and fecal coliform bacteria were isolated from the cloaca and feces of the estuarine diamondback terrapin. The majority of samples contained fecal coliforms. Escherichia coli was the predominant fecal coliform species isolated, and members of the genus Salmonella were isolated from 2 of 39 terrapins. Fecal coliform numbers are used to regulate shellfish harvests, and diamondback terrapins inhabit the brackish-water habitats where oyster beds are found; therefore, these findings have implications for the efficacy of current regulatory parameters in shellfishing waters.     

Comparison of Verification Procedures for the Membrane Filter Total Coliform Technique

Verification of membrane filter total coliform colonies from drinking water was increased 87% by testing for the presence of β-galactosidase and cytochrome oxidase, compared with verification by determination of gas production in lauryl tryptose broth. Over 90% of the coliforms verified by testing for β-galactosidase and cytochrome oxidase were representative of the typical coliform genera.     

Effect of environmental factors on the relationship between concentrations of coprostanol and fecal indicator bacteria in tropical (Mekong Delta) and temperate (Tokyo) freshwaters

A reliable assessment of microbial indicators of fecal pollution (total coliform, Escherichia coli, and fecal streptococcus) is critical in tropical environments. Therefore, we investigated the relationship between concentrations of indicator bacteria and a chemical indicator, coprostanol (5beta-cholestan-3beta-ol), in tropical and temperate regions. Water samples were collected from the Mekong Delta, Vietnam, during wet and dry seasons, and from Tokyo, Japan, during summer, the aftermath of a typhoon, and winter. During the wet season in the Mekong Delta, higher bacterial densities were observed in rivers, probably due to the higher bacterial inputs from soil particles with runoff. In Tokyo, higher bacterial densities were usually observed during summer, followed by those in the typhoon aftermath and winter. A strong logarithmic correlation between the concentrations of E. coli and coprostanol was demonstrated in all surveys. Distinctive seasonal fluctuations were observed, as concentrations of coprostanol corresponding to 1,000 CFU of E. coli/100 ml were at their lowest during the wet season in the Mekong Delta and the typhoon aftermath in Tokyo (30 ng/liter), followed by the dry season in the Mekong Delta and the summer in Tokyo (100 ng/liter), and they were much higher during the winter in Tokyo (400 ng/liter). These results suggested that E. coli is a specific indicator of fecal contamination in both tropical and temperate regions but that the densities are affected by elevated water temperature and input from runoff of soil particles. The concurrent determination of E. coli and coprostanol concentrations could provide a possible approach to assessing the reliability of fecal pollution monitoring data.     

Artificial neural network prediction of viruses in shellfish

A database was probed with artificial neural network (ANN) and multivariate logistic regression (MLR) models to investigate the efficacy of predicting PCR-identified human adenovirus (ADV), Norwalk-like virus (NLV), and enterovirus (EV) presence or absence in shellfish harvested from diverse countries in Europe (Spain, Sweden, Greece, and the United Kingdom). The relative importance of numerical and heuristic input variables to the ANN model for each country and for the combined data was analyzed with a newly defined relative strength effect, which illuminated the importance of bacteriophages as potential viral indicators. The results of this analysis showed that ANN models predicted all types of viral presence and absence in shellfish with better precision than MLR models for a multicountry database. For overall presence/absence classification accuracy, ANN modeling had a performance rate of 95.9%, 98.9%, and 95.7% versus 60.5%, 75.0%, and 64.6% for the MLR for ADV, NLV, and EV, respectively. The selectivity (prediction of viral negatives) was greater than the sensitivity (prediction of viral positives) for both models and with all virus types, with the ANN model performing with greater sensitivity than the MLR. ANN models were able to illuminate site-specific relationships between microbial indicators chosen as model inputs and human virus presence. A validation study on ADV demonstrated that the MLR and ANN models differed in sensitivity and selectivity, with the ANN model correctly identifying ADV presence with greater precision.     

Effect of Environmental Factors on the Relationship between Concentrations of Coprostanol and Fecal Indicator Bacteria in Tropical (Mekong Delta) and Temperate (Tokyo) Freshwaters

A reliable assessment of microbial indicators of fecal pollution (total coliform, Escherichia coli, and fecal streptococcus) is critical in tropical environments. Therefore, we investigated the relationship between concentrations of indicator bacteria and a chemical indicator, coprostanol (5β-cholestan-3β-ol), in tropical and temperate regions. Water samples were collected from the Mekong Delta, Vietnam, during wet and dry seasons, and from Tokyo, Japan, during summer, the aftermath of a typhoon, and winter. During the wet season in the Mekong Delta, higher bacterial densities were observed in rivers, probably due to the higher bacterial inputs from soil particles with runoff. In Tokyo, higher bacterial densities were usually observed during summer, followed by those in the typhoon aftermath and winter. A strong logarithmic correlation between the concentrations of E. coli and coprostanol was demonstrated in all surveys. Distinctive seasonal fluctuations were observed, as concentrations of coprostanol corresponding to 1,000 CFU of E. coli/100 ml were at their lowest during the wet season in the Mekong Delta and the typhoon aftermath in Tokyo (30 ng/liter), followed by the dry season in the Mekong Delta and the summer in Tokyo (100 ng/liter), and they were much higher during the winter in Tokyo (400 ng/liter). These results suggested that E. coli is a specific indicator of fecal contamination in both tropical and temperate regions but that the densities are affected by elevated water temperature and input from runoff of soil particles. The concurrent determination of E. coli and coprostanol concentrations could provide a possible approach to assessing the reliability of fecal pollution monitoring data.     

Identification and origins of neutral fecal sterols in adult Large White sows: occurrence of externally-secreted intestinal cholesterol

Cholesterol, the main neutral fecal sterol (54-84 p. 100) in adult Large White sows fed a controlled semi-purified diet containing 0.08 p. 100 cholesterol (500 g twice a day; 3 510 kcal/day), was partially converted into coprostanol (10-44 p. 100). Exceptionally, epicoprostanol was present, indicating a second pathway of bacterial cholesterol degradation. In this paper, the term "fecal cholesterol" is restricted to the sum of cholesterol + coprostanol. The contribution of fecal cholesterol to the bulk of neutral fecal sterols eliminated daily, averaged 97 +/- 1 p. 100. For a given dietary cholesterol intake of 80 mg per day, eliminated fecal cholesterol was estimated to be 392 +/- 47 mg/day and mean fecal cholesterol concentration 1.88 +/- 0.12 mg/g of stools. The various sources of fecal cholesterol were unabsorbed ingested cholesterol, cholesterol excreted from the plasma, and externally-secreted intestinal cholesterol, synthesized by the digestive tract, discharged into the lumen and not absorbed. The respective contributions of these different sources were as follows: unabsorbed dietary cholesterol 34 +/- 2 mg/day, excreted cholesterol 234 +/- 28 mg/day and externally-secreted cholesterol 125 +/- 23 mg/day.     

Real-Time PCR Detection of Pathogenic Microorganisms in Roof-Harvested Rainwater in Southeast Queensland, Australia

In this study, the microbiological quality of roof-harvested rainwater was assessed by monitoring the concentrations of Escherichia coli, enterococci, Clostridium perfringens, and Bacteroides spp. in rainwater obtained from tanks in Southeast Queensland, Australia. Samples were also tested using real-time PCR (with SYBR Green I dye) for the presence of potential pathogenic microorganisms. Of the 27 rainwater samples tested, 17 (63%), 21 (78%), 13 (48%), and 24 (89%) were positive for E. coli, enterococci, C. perfringens, and Bacteroides spp., respectively. Of the 27 samples, 11 (41%), 7 (26%), 4 (15%), 3 (11%), and 1 (4%) were PCR positive for the Campylobacter coli ceuE gene, the Legionella pneumophila mip gene, the Aeromonas hydrophila lip gene, the Salmonella invA gene, and the Campylobacter jejuni mapA gene. Of the 21 samples tested, 4 (19%) were positive for the Giardia lamblia β-giardin gene. The binary logistic regression model indicated a positive correlation (P < 0.02) between the presence/absence of enterococci and A. hydrophila. In contrast, the presence/absence of the remaining potential pathogens did not correlate with traditional fecal indicators. The poor correlation between fecal indicators and potential pathogens suggested that fecal indicators may not be adequate to assess the microbiological quality of rainwater and consequent health risk.     

Influence of Ileo-Caecal Cannulation and Oxytetracycline on Ileo-Caecal and Rectal Coliform Populations in Pigs

The effect of surgery (insertion of an ileo-caecal cannula) and a subsequent parenteral treatment with oxytetracycline on the ileo-caecal and rectal coliform populations in 7 Swedish Yorkshire castrates were studied. Samples were collected during surgery as well as 3, 7, 14 and 20 days post surgery. The diversity of the enteric coliform flora was initially high both in the ileo-caecal ostium and in rectum. No alteration in the diversity of the enteric coliform flora was observed following surgery and treatment with oxytetracycline. As the insertion of ileo-caecal cannulas did not affect the intestinal coliform flora this study gives support to the use of this technique to mirror processes in the small intestine of pigs. Further, the diversity of the enteric coliform flora was unaffected by the parenteral treatment with oxytetracycline.     

Comparison of Gelman and Millipore Membrane Filters for Enumerating Fecal Coliform Bacteria

Tests of two leading brands of membrane filters used for enumerating fecal coliform bacteria showed that Gelman GN-6 filters recovered statistically more colonies of bacteria than did Millipore HAWG 047SO filters from pure cultures incubated at either 35 C (the optimal growth temperature) or 44.5 C (the standard temperature for the fecal coliform test). Standard membrane filter procedures with M-FC broth base were used to enumerate the organisms. Densities of colonies incubated on Gelman filters at 44.5 C averaged 2.3 times greater than those on Millipore filters. Plate counts of the bacteria at both temperatures indicated that incubation at 44.5 C did not inhibit propagation of fecal coliform bacteria. For the pour plates, M-FC broth base plus 1.5% agar was used. This modified medium compared favorably to plate count agar for enumerating Escherichia coli. At 35 and 44.5 C, colony counts on Gelman filters agreed closely with plate counts prepared concurrently, but Millipore counts were consistently lower than plate counts, especially at 44.5 C. Comparative analyses of river water for fecal coliform bacteria by the membrane filter technique gave results comparable to those for the pure cultures.     

Molecular Detection and Characterization of Gastroenteritis Viruses Occurring Naturally in the Stream Waters of Manaus, Central Amazônia, Brazil

To assess the presence of the four main viruses responsible for human acute gastroenteritis in a hydrographic network impacted by a disordered urbanization process, a 1-year study was performed involving water sample collection from streams in the hydrographic basin surrounding the city of Manaus, Amazonas, Brazil. Thirteen surface water sample collection sites, including different areas of human settlement characterized as urban, rural, and primary forest, located in the Tarumã-Açu, São Raimundo, Educandos, and Puraquequara microbasins, were defined with a global positioning system. At least one virus was detected in 59.6% (31/52) of the water samples analyzed, and rotavirus was the most frequent (44.2%), followed by human adenovirus (30.8%), human astrovirus (15.4%), and norovirus (5.8%). The viral contamination observed mainly in the urban streams reflected the presence of a local high-density population and indicated the gastroenteritis burden from pathogenic viruses in the water, principally due to recreational activities such as bathing. The presence of viral genomes in areas where fecal contamination was not demonstrated by bacterial indicators suggests prolonged virus persistence in aquatic environments and emphasizes the enteric virus group as the most reliable for environmental monitoring.     

Using geographic information systems and regression analysis to evaluate relationships between land use and fecal coliform bacterial pollution

Geographic Information Systems (GIS) and regression modeling techniques were used to evaluate relationships between land use and fecal pollution in Murrells Inlet, a small, urbanized, high-salinity estuary located between Myrtle Beach and Georgetown, SC. GIS techniques were used to identify and calculate land use and spatial variables to be used in a regression model. The regression analysis was performed to identify specific land-use characteristics that may influence fecal coliform densities in the estuary. The regression modeling used land-use parameters to explain the variability of fecal coliform densities measured monthly at 21 locations in the estuary over the 10-year period from 1989 to 1998. Individual regression models were generated for each season, and for a combined data set. The results of the regression analyses indicate that proximity to areas with septic tanks, and rainfall runoff from urbanized areas are important predictors of fecal coliform densities in the estuary. Sampling sites closer to areas with high densities of active septic tanks or more urbanized land uses tended to have higher fecal coliform densities. Although these results may suggest that septic tanks are a substantial human source of fecal pollution, previous research has indicated that the fecal pollution in those areas is probably not from human sources. The areas of Murrells Inlet with higher septic tank densities also are located in areas of high housing density, are in close proximity to the land-water interface, and are in the extreme upper reaches of the estuary, where flushing and dilution effects may be reduced. The higher fecal coliform densities observed at these locations may be a coincidental result of these factors and fecal deposition from pets, and not the direct result of fecal pollution input from the septic tanks.The results also show seasonal differences in the dynamics of fecal coliform bacterial pollution in the estuary. The winter model included a rainfall interaction term, which indicates that ground saturation effects may be an important part of fecal deposition in winter months. The water quality management implications of this research include identification of strategies to reduce or intercept urban stormwater runoff, reduction of waste dumping from boats, and reduction of pet waste.     

Presence and Sources of Fecal Coliform Bacteria in Epilithic Periphyton Communities of Lake Superior

Epilithic periphyton communities were sampled at three sites on the Minnesota shoreline of Lake Superior from June 2004 to August 2005 to determine if fecal coliforms and Escherichia coli were present throughout the ice-free season. Fecal coliform densities increased up to 4 orders of magnitude in early summer, reached peaks of up to 1.4 × 10⁵ CFU cm⁻² by late July, and decreased during autumn. Horizontal, fluorophore-enhanced repetitive-PCR DNA fingerprint analyses indicated that the source for 2% to 44% of the E. coli bacteria isolated from these periphyton communities could be identified when compared with a library of E. coli fingerprints from animal hosts and sewage. Waterfowl were the major source (68 to 99%) of periphyton E. coli strains that could be identified. Several periphyton E. coli isolates were genotypically identical (≥92% similarity), repeatedly isolated over time, and unidentified when compared to the source library, suggesting that these strains were naturalized members of periphyton communities. If the unidentified E. coli strains from periphyton were added to the known source library, then 57% to 81% of E. coli strains from overlying waters could be identified, with waterfowl (15 to 67%), periphyton (6 to 28%), and sewage effluent (8 to 28%) being the major potential sources. Inoculated E. coli rapidly colonized natural periphyton in laboratory microcosms and persisted for several weeks, and some cells were released to the overlying water. Our results indicate that E. coli from periphyton released into waterways confounds the use of this bacterium as a reliable indicator of recent fecal pollution.