Development of predictive models for determining enterococci levels at Gulf Coast beaches

The US EPA BEACH Act requires beach managers to issue swimming advisories when water quality standards are exceeded. While a number of methods/models have been proposed to meet the BEACH Act requirement, no systematic comparisons of different methods against the same data series are available in terms of relative performance of existing methods. This study presents and compares three models for nowcasting and forecasting enterococci levels at Gulf Coast beaches in Louisiana, USA. One was developed using the artificial neural network (ANN) in MATLAB Toolbox and the other two were based on the US EPA Virtual Beach (VB) Program. A total of 944 sets of environmental and bacteriological data were utilized. The data were collected and analyzed weekly during the swimming season (May-October) at six sites of the Holly Beach by Louisiana Beach Monitoring Program in the six year period of May 2005-October 2010. The ANN model includes 15 readily available environmental variables such as salinity, water temperature, wind speed and direction, tide level and type, weather type, and various combinations of antecedent rainfalls. The ANN model was trained, validated, and tested using 308, 103, and 103 data sets (collected in 2007, 2008, and 2009) with an average linear correlation coefficient (LCC) of 0.857 and a Root Mean Square Error (RMSE) of 0.336. The two VB models, including a linear transformation-based model and a nonlinear transformation-based model, were constructed using the same data sets. The linear VB model with 6 input variables achieved an LCC of 0.230 and an RMSE of 1.302 while the nonlinear VB model with 5 input variables produced an LCC of 0.337 and an RMSE of 1.205. In order to assess the predictive performance of the ANN and VB models, hindcasting was conducted using a total of 430 sets of independent environmental and bacteriological data collected at six Holly Beach sites in 2005, 2006, and 2010. The hindcasting results show that the ANN model is capable of predicting enterococci levels at the Holly Beach sites with an adjusted RMSE of 0.803 and LCC of 0.320 while the adjusted RMSE and LCC values are 1.815 and 0.354 for the linear VB model and 1.961and 0.521 for the nonlinear VB model. The results indicate that the ANN model with 15 parameters performs better than the VB models with 6 or 5 parameters in terms of RMSE while VB models perform better than the ANN model in terms of LCC. The predictive models (especially the ANN and the nonlinear VB models) developed in this study in combination with readily available real-time environmental and weather forecast data can be utilized to nowcast and forecast beach water quality, greatly reducing the potential risk of contaminated beach waters to human health and improving beach management. While the models were developed specifically for the Holly Beach, Louisiana, the methods used in this paper are generally applicable to other coastal beaches.     

Linking non-culturable (qPCR) and culturable enterococci densities with hydrometeorological conditions

Quantitative polymerase chain reaction (qPCR) measurement of enterococci has been proposed as a rapid technique for assessment of beach water quality, but the response of qPCR results to environmental conditions has not been fully explored. Culture-based E. coli and enterococci have been used in empirical predictive models to characterize their responses to environmental conditions and to increase monitoring frequency and efficiency. This approach has been attempted with qPCR results only in few studies. During the summer of 2006, water samples were collected from two southern Lake Michigan beaches and the nearby river outfall (Burns Ditch) and were analyzed for enterococci by culture-based and non-culture-based (i.e., qPCR) methods, as well as culture-based E. coli. Culturable enterococci densities (log CFU/100 ml) for the beaches were significantly correlated with enterococci qPCR cell equivalents (CE) (R = 0.650, P < 0.0001, N = 32). Enterococci CE and CFU densities were highest in Burns Ditch relative to the beach sites; however, only CFUs were significantly higher (P < 0.0001). Culturable enterococci densities at Burns Ditch and the beaches were significantly correlated (R = 0.565, P < 0.0001, N = 32). Culturable E. coli and enterococci densities were significantly correlated (R = 0.682, P < 0.0001, N = 32). Regression analyses suggested that enterococci CFU could be predicted by lake turbidity, Burns Ditch discharge, and wind direction (adjusted R² = 0.608); enterococci CE was best predicted by Burns Ditch discharge and log-transformed lake turbidity × wave height (adjusted R² = 0.40). In summary, our results show that analytically, the qPCR method compares well to the non-culture-based method for measuring enterococci densities in beach water and that both these approaches can be predicted by hydrometeorological conditions. Selected predictors and model results highlight the differences between the environmental responses of the two method endpoints and the potentially high variance in qPCR results.     

Rapid decay of host-specific fecal Bacteroidales cells in seawater as measured by quantitative PCR with propidium monoazide

We investigated the persistence of feces-derived Bacteroidales cells and their DNA in seawater under natural conditions using an optimized chemical method based on co-extraction of nucleic acids with propidium monoazide (PMA), which interferes with PCR amplification of molecular markers from extracellular DNA and dead bacterial cells. The previously validated Bacteroidales assays BacUni-UCD, BacHum-UCD, BacCow-UCD, and BacCan-UCD were utilized to determine concentrations of Bacteroidales genetic markers targeting all warm-blooded animals, humans, cows and dogs, specifically, over a period of 24 d. Microcosms containing mixed feces in dialysis tubing were exposed to seawater under flow-through conditions at ambient temperature in the presence and absence of sunlight. Using a two-stage plus linear decay model, the average T₉₉ (two-log reduction) of host-specific Bacteroidales cells was 28 h, whereas that of host-specific Bacteroidales DNA was 177 h. Natural sunlight did not affect the survival of uncultivable Bacteroidales cells and their DNA with the exception of the BacCow-UCD marker. Bacteroidales DNA, as measured by quantitative PCR (qPCR) without PMA, persisted for as long as 24 d at concentrations close to the limit of detection. Culturable Enterococcus cells were detected for only 70 h, whereas Enterococcus cells measured by qPCR with and without PMA persisted for 450 h. In conclusion, measuring Bacteroidales DNA without differentiating between intact and dead cells or extracellular DNA may misinform about the extent of recent fecal pollution events, particularly in the case of multiple sources of contamination with variable temporal and spatial scales due to the relatively long persistence of DNA in the environment. In contrast, applying qPCR with and without PMA can provide data on the fate and transport of fecal Bacteroidales in water, and help implement management practices to protect recreational water quality.     

Concentrations of host-specific and generic fecal markers measured by quantitative PCR in raw sewage and fresh animal feces

We measured the concentrations of four host-specific (human, dog, cow, and horse Bacteroidales), four generic fecal (16S total Bacteroidales and Escherichia coli, 23S Enterococcus and uidA E. coli,) and two universal bacterial (16S universal and rpoB universal) DNA targets by qPCR in raw sewage and pooled fecal samples from dogs, cows, horses, and Canada Geese. A spiking protocol using the non-fecal bacterium Pseudomonas syringae pph6 was developed to estimate the recovery of DNA from fecal and environmental samples. The measured fecal marker concentrations were used to calculate baseline ratios and variability of host-specific to generic indicators for each host type. The host-specific markers were found in high concentrations (8-9 log₁₀ copies/g dry wt.) in their respective hosts' samples, which were equal to or greater than the concentrations of generic E. coli and Enterococcus markers, lending support to the use of host-specific and generic Bacteroidales as sensitive indicators of fecal pollution. The host-specific markers formed a consistent percentage of total Bacteroidales in target host feces and raw sewage, with human-specific comprising 82%, dog-specific 6%, cow-specific 4% and horse-specific 2%. Based on this limited data set, the measurement of host-specific indicators by qPCR has several promising applications. These applications include determining the percentage of total Bacteroidales contributed by a specific host type, using the ratios of host-specific markers to E. coli or Enterococcus to estimate the contribution of each source to these regulated fecal indicator bacteria, and estimating the mass of feces from each host type in environmental samples.     

Sensitive detection of sample interference in environmental qPCR

Sample interference in environmental applications of quantitative PCR (qPCR) can prevent accurate estimations of molecular markers in the environment. We developed a spike-and-recovery approach using a mutant strain of Escherichia coli that contains a chromosomal insertion of a mutant GFP gene. The method was tested in water samples by separately reducing extraction efficiency or adding humic acids and ethanol, compounds that often contaminate environmental DNA extracts, and analyzing qPCR amplification of the spiked E. coli control and human fecal Bacteroides markers (HF183 and HF134). This approach, coupled with previously developed kinetic outlier detection (KOD) methods, allowed sensitive detection of PCR inhibition at much lower inhibitor concentrations than alternative approaches using Cq values or amplification efficiencies. Although HF183 was more sensitive to the effects of qPCR inhibitors than the E. coli control assay, KOD methods correctly identified inhibition of both control and HF183 assays in samples containing as little as 0.1 ng humic acids per reaction or 5% ethanol. Because sigmoidal modeling methods allow distinction of qPCR inhibition from poor DNA recovery, we were able to simultaneously identify qPCR-inhibited reactions and estimate recovery of nucleic acids in environmental samples using a single control assay. Since qPCR is currently used to estimate important water quality parameters that have serious economic and human health outcomes, these results are timely. While we demonstrate the methods in the context of water quality regulation, they will be useful in all areas of environmental research that use qPCR.     

Quantitative real-time PCR analysis of total and propidium monoazide-resistant fecal indicator bacteria in wastewater

A real-time quantitative PCR (qPCR) method and a modification of this method incorporating pretreatment of samples with propidium monoazide (PMA) were evaluated for respective analyses of total and presumptively viable Enterococcus and Bacteroidales fecal indicator bacteria. These methods were used in the analyses of wastewater samples to investigate their feasibility as alternatives to current fecal indicator bacteria culture methods for predicting the efficiency of viral pathogen removal by standard treatment processes. PMA treatment was effective in preventing qPCR detection of target sequences from non-viable cells. Concentrates of small volume, secondary-treated wastewater samples, collected from a publicly owned treatment works (POTW) under normal operating conditions, had little influence on this effectiveness. Higher levels of total suspended solids, such as those associated with normal primary treatment and all treatment stages during storm flow events, appeared to interfere with PMA effectiveness under the sample preparation conditions employed. During normal operating conditions at three different POTWs, greater reductions were observed in PMA-qPCR detectable target sequences of both Enterococcus and Bacteroidales than in total qPCR detectable sequences. These reductions were not as great as those observed for cultivable fecal indicator bacteria in response to wastewater disinfection. Reductions of PMA-qPCR as well as total qPCR detectable target sequences from enterococci and, to a lesser extent, Bacteroidales correlated well with reductions in infectious viruses during both normal and storm flow operating conditions and therefore may have predictive value in determining the efficiency at which these pathogens are removed.     

Using rapid indicators for Enterococcus to assess the risk of illness after exposure to urban runoff contaminated marine water

Background

Traditional fecal indicator bacteria (FIB) measurement is too slow (>18 h) for timely swimmer warnings.

Objectives

Assess relationship of rapid indicator methods (qPCR) to illness at a marine beach impacted by urban runoff.

Methods

We measured baseline and two-week health in 9525 individuals visiting Doheny Beach 2007-08. Illness rates were compared (swimmers vs. non-swimmers). FIB measured by traditional (Enterococcus spp. by EPA Method 1600 or Enterolert™, fecal coliforms, total coliforms) and three rapid qPCR assays for Enterococcus spp. (Taqman, Scorpion-1, Scorpion-2) were compared to health. Primary bacterial source was a creek flowing untreated into ocean; the creek did not reach the ocean when a sand berm formed. This provided a natural experiment for examining FIB-health relationships under varying conditions.

Results

We observed significant increases in diarrhea (OR 1.90, 95% CI 1.29-2.80 for swallowing water) and other outcomes in swimmers compared to non-swimmers. Exposure (body immersion, head immersion, swallowed water) was associated with increasing risk of gastrointestinal illness (GI). Daily GI incidence patterns were different: swimmers (2-day peak) and non-swimmers (no peak). With berm-open, we observed associations between GI and traditional and rapid methods for Enterococcus; fewer associations occurred when berm status was not considered.

Conclusions

We found increased risk of GI at this urban runoff beach. When FIB source flowed freely (berm-open), several traditional and rapid indicators were related to illness. When FIB source was weak (berm-closed) fewer illness associations were seen. These different relationships under different conditions at a single beach demonstrate the difficulties using these indicators to predict health risk.     

Rapid QPCR-based assay for fecal Bacteroides spp. as a tool for assessing fecal contamination in recreational waters

Concentrations of fecal indicator bacteria (FIB; e.g. Escherichia coli, and Enterococcus sp.) can only be used in limited ways for determining the source of fecal contamination in recreational waters because they cannot distinguish human from non-human fecal contamination. Several Bacteroides spp. have been suggested as potential alternative indicators. We have developed a rapid, culture-independent method for quantifying fecal Bacteroides spp. using quantitative PCR (QPCR) targeting the 16S rRNA gene. The assay specifically targets and quantifies the most common human Bacteroides spp. The details of the method are presented, including analyses of a wide range of fecal samples from different organisms. Specificity and performance of the QPCR assay were also tested via a laboratory experiment where human sewage and gull guano were inoculated into a range of environmental water samples. Concentrations of fecal Bacteroides spp., total Enterococcus sp., Enterococcus faecium, Enterococcus faecalis, and Enterococcus casseliflavus were measured using QPCR, and total Enterococcus sp. and E. coli were quantified by membrane filtration (MF). Samples spiked with gull guano were highly concentrated with total Enterococcus sp., E. coli, E. faecalis, and E. casseliflavus, demonstrating that these indicators are prominent in animal feces. On the other hand, fecal Bacteroides spp. concentrations were high in samples containing sewage and were relatively low in samples spiked with gull guano. Sensitivity and specificity results suggest that the rapid fecal Bacteroides spp. QPCR assay may be a useful tool to effectively predict the presence and concentration of human-specific fecal pollution.     

Spatial and temporal variation in indicator microbe sampling is influential in beach management decisions

Fecal indicator microbes, such as enterococci, are often used to assess potential health risks caused by pathogens at recreational beaches. Microbe levels often vary based on collection time and sampling location. The primary goal of this study was to assess how spatial and temporal variations in sample collection, which are driven by environmental parameters, impact enterococci measurements and beach management decisions. A secondary goal was to assess whether enterococci levels can be predictive of the presence of Staphylococcus aureus, a skin pathogen. Over a ten-day period, hydrometeorologic data, hydrodynamic data, bather densities, enterococci levels, and S. aureus levels including methicillin-resistant S. aureus (MRSA) were measured in both water and sand. Samples were collected hourly for both water and sediment at knee-depth, and every 6 h for water at waist-depth, supratidal sand, intertidal sand, and waterline sand. Results showed that solar radiation, tides, and rainfall events were major environmental factors that impacted enterococci levels. S. aureus levels were associated with bathing load, but did not correlate with enterococci levels or any other measured parameters. The results imply that frequencies of advisories depend heavily upon sample collection policies due to spatial and temporal variation of enterococci levels in response to environmental parameters. Thus, sampling at different times of the day and at different depths can significantly impact beach management decisions. Additionally, the lack of correlation between S. aureus and enterococci suggests that use of fecal indicators may not accurately assess risk for some pathogens.     

Evaluation of rapid methods and novel indicators for assessing microbiological beach water quality

A broad suite of new measurement methods and indicators based on molecular measurement technology have been developed to assess beach water quality, but they have generally been subjected to limited testing outside of the laboratory in which they were developed. Here we evaluated 29 assays targeting a variety of bacterial, viral, and chemical analytes by providing the method developers with twelve blind samples consisting of samples spiked with known concentration of sewage or gull guano and negative controls. Each method was evaluated with respect to its ability to detect the target organism, absence of signal in the negative controls and repeatability among replicates. Only six of the 30 methods detected their targets in at least 75% of the samples while consistently determining the absence of the target in the negative controls. Among quantitative methods, QPCR for Bacteroides thetaiotamicron and Enterococcus detected by Luminex reliably identified all but one sample containing human fecal material and produced no false positive results. Among non-quantitative methods, the Enterococcus esp gene, the Bacteroidales human specific marker and culture-based coliphage were the most reliable for identifying human fecal material. We also found that investigator-specific variations of methods targeting the same organism often produced different results.     

Large scale analysis of virulence genes in Escherichia coli strains isolated from Avalon Bay, CA

Contamination of recreational waters with Escherichia coli and Enterococcus sp. is a widespread problem resulting in beach closures and loss of recreational activity. While E. coli is frequently used as an indicator of fecal contamination, and has been extensively measured in waterways, few studies have examined the presence of potentially pathogenic E. coli strains in beach waters. In this study, a combination of high-throughput, robot-assisted colony hybridization and PCR-based analyses were used to determine the genomic composition and frequency of virulence genes present in E. coli isolated from beach water in Avalon Bay, Santa Catalina Island, CA. A total of 24,493 E. coli isolates were collected from two sites at a popular swimming beach between August through September 2007 and from July through August 2008. All isolates were examined for the presence of shiga-like toxins (stx1/stx2), intimin (eaeA), and enterotoxins (ST/LT). Of the 24,493 isolates examined, 3.6% contained the eaeA gene, indicating that these isolates were potential EPEC strains. On five dates, however, greater than 10% of the strains were potential EPEC, suggesting that incidence of virulence genes at this beach has a strong temporal component. No STEC or ETEC isolates were detected, and only eight (<1.0%) of the potential EPEC isolates were found to carry the EAF plasmid. The potential EPEC isolates mainly belonged to E. coli phylogenetic groups B1 or B2, and carried the β intimin subtype. DNA fingerprint analyses of the potential EPEC strains indicated that the isolates belonged to several genetically diverse groups, although clonal isolates were frequently detected. While the presence of virulence genes alone cannot be used to determine the pathogenicity of strains, results from this study show that potential EPEC strains can be found in marine beach water and their presence needs to be considered as one of the factors used in decisions concerning beach closures.     

A multi-beach study of Staphylococcus aureus, MRSA, and enterococci in seawater and beach sand

Incidences of Staphylococcus aureus and methicillin resistant S. aureus (MRSA) have risen worldwide prompting a need to better understand routes of human exposure and whether standard bacterial water quality monitoring practices adequately account for this potential threat. Beach water and sand samples were analyzed during summer months for S. aureus, enterococci, and MRSA at three southern California beaches (Avalon, Doheny, Malibu Surfrider). S. aureus frequently was detected in samples of seawater (59%, n = 328) and beach sand (53%, n = 358). MRSA sometimes was detected in seawater (1.6%, n = 366) and sand (2.7%, n = 366) at relatively low concentrations. Site specific differences were observed, with Avalon Beach presenting the highest concentrations of S. aureus and Malibu Surfrider the lowest in both seawater and sand. S. aureus concentrations in seawater and sand were correlated to each other and to a variety of other parameters. Multiple linear regression on the combined beach data indicated that significant explanatory variables for S. aureus in seawater were S. aureus in sand, water temperature, enterococci in seawater, and the number of swimmers. In sand, S. aureus concentrations were related to S. aureus in seawater, water temperature, enterococci in seawater, and inversely to surf height classification. Only the correlation to water temperature held for individually analyzed beaches and for S. aureus concentrations in both seawater and sand. To provide context for these results, the prevalence of S. aureus in sand was compared to published fomite studies, and results suggested that beach prevalence was similar to that in homes.     

Determination of the optimal parameters in regression models for the prediction of chlorophyll-a: A case study of the Yeongsan Reservoir, Korea

Statistical regression models involve linear equations, which often lead to significant prediction errors due to poor statistical stability and accuracy. This concern arises from multicollinearity in the models, which may drastically affect model performance in terms of a trade-off scenario for effective water resource management logistics. In this paper, we propose a new methodology for improving the statistical stability and accuracy of regression models, and then show how to cope with pitfalls in the models and determine optimal parameters with a decreased number of predictive variables. Here, a comparison of the predictive performance was made using four types of multiple linear regression (MLR) and principal component regression (PCR) models in the prediction of chlorophyll-a (chl-a) concentration in the Yeongsan (YS) Reservoir, Korea, an estuarine reservoir that historically suffers from high levels of nutrient input. During a 3-year water quality monitoring period, results showed that PCRs could be a compact solution for improving the accuracy of the models, as in each case MLR could not accurately produce reliable predictions due to a persistent collinearity problem. Furthermore, based on R² (goodness of fit) and F-overall number (confidence of regression), and the number of explanatory variables (R-F-N) curve, it was revealed that PCR-F(7) was the best model among the four regression models in predicting chl-a, having the fewest explanatory variables (seven) and the lowest uncertainty. Seven PCs were identified as significant variables, related to eight water quality parameters: pH, 5-day biochemical oxygen demand, total coliform, fecal indicator bacteria, chemical oxygen demand, ammonia-nitrogen, total nitrogen, and dissolved oxygen. Overall, the results not only demonstrated that the models employed successfully simulated chl-a in a reservoir in both the test and validation periods, but also suggested that the optimal parameters should cautiously be considered in the design of regression models.     

Evaluation of statistical models for predicting Escherichia coli particle attachment in fluvial systems

Modeling surface water Escherichia coli fate and transport requires partitioning E. coli into particle-attached and unattached fractions. Attachment is often assumed to be a constant fraction or is estimated using simple linear models. The objectives of this study were to: (i) develop statistical models for predicting E. coli attachment and virulence marker presence in fluvial systems, and (ii) relate E. coli attachment to a variety of environmental parameters. Stream water samples (n = 60) were collected at four locations in a rural, mixed-use watershed between June and October 2012, with four storm events (>20 mm rainfall) being captured. The percentage of E. coli attached to particles (>5 μm) and the occurrences of virulence markers were modeled using water quality, particle concentration, particle size distribution, hydrology and land use factors as explanatory variables. Three types of statistical models appropriate for highly collinear, multidimensional data were compared: least angle shrinkage and selection operator (LASSO), classification and regression trees using the general, unbiased, interaction detection and estimation (GUIDE) algorithm, and multivariate adaptive regression splines (MARS). All models showed that E. coli particle attachment and the presence of E. coli virulence markers in the attached and unattached states were influenced by a combination of water quality, hydrology, land-use and particle properties. Model performance statistics indicate that MARS models outperform LASSO and GUIDE models for predicting E. coli particle attachment and virulence marker occurrence. Validating the MARS modeling approach in multiple watersheds may allow for the development of a parameterizing model to be included in watershed simulation models.     

Method for rapid and sensitive detection of Enterococcus sp. and Enterococcus faecalis/faecium cells in potable water samples

We have developed a rapid and robust technological solution including a membrane filtration and dissolution method followed by a molecular enrichment and a real-time PCR assay, for detecting the presence of Enterococcus sp. or Enterococcus faecalis/faecium per 100 mL of water in less than 5 h and we compared it to Method 1600 on mEI agar in terms of specificity, sensitivity, and limit of detection. The mEI and the Enterococcus sp.-specific assay detected respectively 73 (64.0%) and 114 (100%) of the 114 enterococcal strains tested. None of the 150 non-enterococcal strains tested was detected by both methods with the exception of Tetragenococcus solitarius for the Enterococcus sp. assay. The multiplexed E. faecalis/faecium assay efficiently amplified DNA from 47 of 47 (100%) E. faecalis and 27 of 27 (100%) E. faecium strains tested respectively, whereas none of the 191 non-E. faecalis/faecium strains tested was detected. By simultaneously detecting the predominant fecal enterococcal species, the E. faecalis/faecium-specific assay allows a better distinction between enterococcal strains of fecal origin and those provided by the environment than Method 1600. Our procedure allows the detection of 4.5 enterococcal colony forming units (CFU) per 100 mL in less than 5 h, whereas the mEI method detected 2.3 CFU/100 mL in 24 h (95% confidence). Thus, our innovative and highly effective method provides a rapid and easy approach to concentrate very low numbers of enterococcal cells present in a 100 mL water sample and allows a better distinction between fecal and environmental enterococcal cells than Method 1600.     

A cross comparison of QPCR to agar-based or defined substrate test methods for the determination of Escherichia coli and enterococci in municipal water quality monitoring programs

Molecular methods such as quantitative, real-time polymerase chain reaction (QPCR) are intended to shorten the period between sampling and publicly available results. Cross comparison studies in Racine, WI, USA evaluated QPCR against agar-based (US EPA Method 1600) and defined substrate (IDEXX Colilert-18^®) methods for the detection and quantification of Escherichia coli and enterococci in a variety of aqueous environments (wastewater, stormwater, and surface water). Regulatory outcomes were also compared based on choice of indicator and method. Positive correlation was seen between QPCR cell equivalents and viable cells through the wastewater treatment process and in all surface water samples (river or freshwater bathing beach) but not in direct stormwater discharge. For surface water samples, correlation improved with the application of a site-specific corrective factor, with regulatory action correctly predicted 98% of the time at bathing beaches. This study suggests the potential utility of QPCR for certain water quality monitoring applications.     

Efficacy of monitoring and empirical predictive modeling at improving public health protection at Chicago beaches

Efforts to improve public health protection in recreational swimming waters have focused on obtaining real-time estimates of water quality. Current monitoring techniques rely on the time-intensive culturing of fecal indicator bacteria (FIB) from water samples, but rapidly changing FIB concentrations result in management errors that lead to the public being exposed to high FIB concentrations (type II error) or beaches being closed despite acceptable water quality (type I error). Empirical predictive models may provide a rapid solution, but their effectiveness at improving health protection has not been adequately assessed. We sought to determine if emerging monitoring approaches could effectively reduce risk of illness exposure by minimizing management errors. We examined four monitoring approaches (inactive, current protocol, a single predictive model for all beaches, and individual models for each beach) with increasing refinement at 14 Chicago beaches using historical monitoring and hydrometeorological data and compared management outcomes using different standards for decision-making. Predictability (R²) of FIB concentration improved with model refinement at all beaches but one. Predictive models did not always reduce the number of management errors and therefore the overall illness burden. Use of a Chicago-specific single-sample standard—rather than the default 235 E. coli CFU/100 ml widely used—together with predictive modeling resulted in the greatest number of open beach days without any increase in public health risk. These results emphasize that emerging monitoring approaches such as empirical models are not equally applicable at all beaches, and combining monitoring approaches may expand beach access.     

Relationships between human adenoviruses and faecal indicator organisms in European recreational waters

Human adenoviruses (HAdV) may be implicated in some disease outbreaks associated with recreational water exposures, typically in swimming pools. Modern molecular methods can be used to detect HAdV in environmental water samples. During the EU FP6 Project VIROBATHE a database of over 290 HAdV analyses with corresponding faecal indicator organism (FIO) determinations was gathered and used to explore statistical associations between HAdV and FIO results. The FIOs measured were Escherichia coli, intestinal enterococci and somatic coliphage. Statistically significant trends of increasing proportions of HAdV-positive results in categories of increasing FIO concentration were found in freshwater but not seawater samples. The analysis of these trends in freshwater samples was refined, the trends remaining statistically significant when using categories of 0.5 log₁₀ intervals of FIO concentration. Logistic regression models were then developed to predict the probability of a HAdV-positive outcome from FIO concentration. Potential applications of these models to predict the probability of HAdV-positive outcomes from routine FIO determinations used to describe recreational water quality exposures and to classify recreational water quality are discussed.     

Evaluation of low-copy genetic targets for waterborne bacterial pathogen detection via qPCR

Recent developments in water quality research have highlighted difficulties in accurately predicting the incidence of pathogens within freshwater based on the viability, culturability and metabolic activity of indicator organisms. QPCR-driven assays are candidates to replace standard culture-based methods, however, protocols suitable for routine use have yet to be sufficiently validated. The objective of this study was to evaluate five oligonucleotide primers sets (ETIR, SINV, exoT, VS1 and ipaH2) for their potential applicability in qPCR assays to detect contamination from five waterborne bacterial pathogens (Escherichia coli O157:H7, Salmonella Typhimurium, Campylobacter jejuni, Pseudomonas aeruginosa, and Shigella flexneri). An enrichment-free qPCR protocol was also tested using S. Typhimurium-seeded source water, combining membrane filtration and mechanical, chemical and enzymatic lysis techniques to recover the bacterial cells. All five primer sets were found to have high specificity and sensitivity for the tested organisms. Four of the primers were able to detect pathogen loads as low as 10 cells/mL while 200 cells/mL of C. jejuni were detectable in pure culture. Although sensitivity decreased in an artificially contaminated environmental matrix, it was still possible to detect as few as 10 S. Typhimurium cells without enrichment. The primers and protocols evaluated in this study have demonstrated potential for further validation for possible application alongside traditional indicator techniques.     

Examining spatially varying relationships between land use and water quality using geographically weighted regression I: model design and evaluation

Traditional regression techniques such as ordinary least squares (OLS) can hide important local variations in the model parameters, and are not able to deal with spatial autocorrelations existing in the variables. A recently developed technique, geographically weighted regression (GWR), is used to examine the relationships between land use and water quality in eastern Massachusetts, USA. GWR models make great improvements of model performance over OLS models, which is proved by F-test and comparisons of model R² and corrected Akaike Information Criterion (AIC_c) from both GWR and OLS. GWR models also improve the reliabilities of the relationships by reducing spatial autocorrelations. The application of GWR models finds that the relationships between land use and water quality are not constant over space but show great spatial non-stationarity. GWR models are able to reveal the information previously ignored by OLS models on the local causes of water pollution, and so improve the model ability to explain local situation of water quality. The results of this study suggest that GWR technique has the potential to serve as a useful tool for environmental research and management at watershed, regional, national and even global scales.