Application of empirical predictive modeling using conventional and alternative fecal indicator bacteria in eastern North Carolina waters

A critical review of model construction and performance for nowcast systems for faecal contamination in recreational beaches

Faecal contamination is a widespread environmental and public health problem on recreational beaches around the world. The implementation of predictive models has been recommended by the World Health Organization as a complement to traditional monitoring to assist decision-makers and reduce health risks. Despite several advances that have been made in the modeling of faecal coliforms, tools and algorithms from machine learning are still scarcely used in the field and their implementation in nowcast systems is delayed. Here, we perform a literature review on modeling strategies to predict faecal contamination in recreational beaches in the last two decades and the implementation of models in nowcast systems to aid management. Models constructed for surface waters of continental (lakes, rivers and streams), estuarine and marine coastal ecosystems were analyzed and compared based on performance metrics for continuous (i.e. regression; R2, Root Mean Square Error: RMSE) and categorical (i.e. classification; accuracy, sensitivity, specificity) responses. We found 67 articles matching the search criteria and 40 with information allowing to evaluate and compare predictive ability. In early 2000, Multiple Linear Regressions were common, followed by a peak of Artificial Neural Networks (ANNs) from 2010 to 2015, and the rise of Machine learning techniques, such as decision trees (CART and Random Forest) since 2015. ANNs and decision trees presented better accuracy than the remaining models. Rainfall and its lags were important predictor variables followed by water temperature. Specificity was much higher than sensitivity in all modeling strategies, which is typical in data sets where one category (e.g. closed beach) is far less common than the normal state (i.e. unbalanced data sets). We registered the implementation of statistical models in early warning systems in 6 countries, mainly by public beach quality management institutions, followed by NGOs in conjunction with universities. We identified critical steps towards improving model construction, evaluation and usage: i) the need to balance the data set previous to model training, ii) the need to separate data set in training, validation and test to perform an honest evaluation of model performance and iii) the transduction of model outputs to plain language to relevant stakeholders. Integrating into a single framework in situ monitoring, model construction and nowcasting systems could help to improve decision making systems to protect users from bathing in contaminated waters. Still the reduction of arrival of faecal coliforms to aquatic ecosystems (e.g. by improving sewage treatment systems) will be the ultimate factor in reducing health risk.

Using a combination of quantitative culture, molecular, and infrastructure data to rank potential sources of fecal contamination in Town Creek Estuary, North Carolina

Estuarine water quality is declining worldwide due to increased tourism, coastal development, and a changing climate. Although well-established methods are in place to monitor water quality, municipalities struggle to use the data to prioritize infrastructure for monitoring and repair and to determine sources of contamination when they occur. The objective of this study was to assess water quality and prioritize sources of contamination within Town Creek Estuary (TCE), Beaufort, North Carolina, by combining culture, molecular, and geographic information systems (GIS) data into a novel contamination source ranking system. Water samples were collected from TCE at ten locations on eight sampling dates in Fall 2021 (n = 80). Microbiological water quality was assessed using US Environmental Protection Agency (U.S. EPA) approved culture-based methods for fecal indicator bacteria (FIB), including analysis of total coliforms (TC), Escherichia coli (EC), and Enterococcus spp. (ENT). The quantitative microbial source tracking (qMST) human-associated fecal marker, HF183, was quantified using droplet digital PCR (ddPCR). This information was combined with environmental data and GIS information detailing proximal sewer, septic, and stormwater infrastructure to determine potential sources of fecal contamination in the estuary. Results indicated FIB concentrations were significantly and positively correlated with precipitation and increased throughout the estuary following rainfall events (p < 0.01). Sampling sites with FIB concentrations above the U.S. EPA threshold also had the highest percentages of aged, less durable piping materials. Using a novel ranking system combining concentrations of FIB, HF183, and sewer infrastructure data at each site, we found that the two sites nearest the most aged sewage infrastructure and stormwater outflows were found to have the highest levels of measurable fecal contamination. This case study supports the inclusion of both traditional water quality measurements and local infrastructure data to support the current need for municipalities to identify, prioritize, and remediate failing infrastructure.

Fecal Bacteria Contamination of Floodwaters and a Coastal Waterway From Tidally-Driven Stormwater Network Inundation

Inundation of coastal stormwater networks by tides is widespread due to sea-level rise (SLR). The water quality risks posed by tidal water rising up through stormwater infrastructure (pipes and catch basins), out onto roadways, and back out to receiving water bodies is poorly understood but may be substantial given that stormwater networks are a known source of fecal contamination. In this study, we (a) documented temporal variation in concentrations of Enterococcus spp. (ENT), the fecal indicator bacteria standard for marine waters, in a coastal waterway over a 2-month period and more intensively during two perigean spring tide periods, (b) measured ENT concentrations in roadway floodwaters during tidal floods, and (c) explained variation in ENT concentrations as a function of tidal inundation, antecedent rainfall, and stormwater infrastructure using a pipe network inundation model and robust linear mixed effect models. We find that ENT concentrations in the receiving waterway vary as a function of tidal stage and antecedent rainfall, but also site-specific characteristics of the stormwater network that drains to the waterway. Tidal variables significantly explain measured ENT variance in the waterway, however, runoff drove higher ENT concentrations in the receiving waterway. Samples of floodwaters on roadways during both perigean spring tide events were limited, but all samples exceeded the threshold for safe public use of recreational waters. These results indicate that inundation of stormwater networks by tides could pose public health hazards in receiving water bodies and on roadways, which will likely be exacerbated in the future due to continued SLR.

Relationship between coliphage and Enterococcus at southern California beaches and implications for beach water quality management

Coliphage have been suggested as an alternative to fecal indicator bacteria for assessing recreational beach water quality, but it is unclear how frequently and at what types of beaches coliphage produces a different management outcome. Here we conducted side-by-side sampling of male-specific and somatic coliphage by the new EPA dead-end hollow fiber ultrafiltration (D-HFUF-SAL) method and Enterococcus at southern California beaches over two years. When samples were combined for all beach sites, somatic and male-specific coliphage both correlated with Enterococcus. When examined categorically, Enterococcus would have resulted in approximately two times the number of health advisories as somatic coliphage and four times that of male-specific coliphage,using recently proposed thresholds of 60 PFU/100 mL for somatic and 30 PFU/100 mL for male-specific coliphage. Overall, only 12% of total exceedances would have been for coliphage alone. Somatic coliphage exceedances that occurred in the absence of an Enterococcus exceedance were limited to a single site during south swell events, when this beach is known to be affected by nearby minimally treated sewage. Thus, somatic coliphage provided additional valuable health protection information, but may be more appropriate as a supplement to FIB measurements rather than as replacement because: (a) EPA-approved PCR methods for Enterococcus allow a more rapid response, (b) coliphage is more challenging owing to its greater sampling volume and laboratory time requirements, and (c) Enterococcus' long data history has yielded predictive management models that would need to be recreated for coliphage.

Integrating culture and molecular quantification of microbial contaminants into a predictive modeling framework in a low-lying, tidally-influenced coastal watershed

Examinations of stormwater delivery in the context of tidal inundation are lacking. Along the coastal plains of the southeastern United States, tidal inundation is increasing in frequency and severity, often with dramatic adverse impacts on timely stormwater discharge, coastal flooding hazards, and even "sunny day flooding". Therefore, a comprehensive study was conducted to examine tidally-influenced stormwater outfalls discharging to Taylor's Creek, an estuary proximal to Beaufort, NC used regularly for recreation and tourism. Over a wide range of meteorological conditions, water samples were collected and analyzed for fecal indicator bacteria (FIB, used for water quality management) and previously published quantitative microbial source tracking (qMST) markers. Nineteen sampling events were conducted from July 2017-June 2018 with samples classified according to tidal state and defined as either inundated, receding, or transition. A first-of-its-kind multiple linear regression model was developed to predict concentrations of Enterococcus sp. by tidal cycle, salinity and antecedent rainfall. We demonstrated that the majority of variability associated with the concentration of Enterococcus sp. could be predicted by E. coli concentration and tidal phase. FIB concentrations were significantly (<0.05) influenced by tide with higher concentrations observed in samples collected during receding (low) tides (EC: log 3.12 MPN/100 mL; ENT: 2.67 MPN/100 mL) compared to those collected during inundated (high) (EC: log 2.62 MPN/100 mL; ENT: 2.11 MPN/100 mL) or transition (EC: log 2.74 MPN/100 mL; ENT: 2.53 MPN/100 mL) tidal periods. Salinity, was also found to significantly (<0.05) correlate with Enterococcus sp. concentrations during inundated (high) tidal conditions (sal: 17 ppt; ENT: 2.04 MPN/100 mL). Tide, not precipitation, was shown to be a significant driver in explaining the variability in Enterococcus sp. concentrations. Precipitation has previously been shown to be a driver of Enterococcus sp. concentrations, but our project demonstrates the need for tidal parameters to be included in the future development of water quality monitoring programs.

Examining coastal dynamics and recreational water quality by quantifying multiple sewage specific markers in a North Carolina estuary

Fecal contamination is observed downstream of municipal separate storm sewer systems in coastal North Carolina. While it is well accepted that wet weather contributes to this phenomenon, less is understood about the contribution of the complex hydrology in this low-lying coastal plain. A quantitative microbial assessment was conducted in Beaufort, North Carolina to identify trends and potential sources of fecal contamination in stormwater receiving waters. Fecal indicator concentrations were significantly higher in receiving water downstream of a tidally submerged outfall compared to an outfall that was permanently submerged (p < 0.001), though tidal height was not predictive of human-specific microbial source tracking (MST) marker concentrations at the tidally submerged site. Short-term rainfall (i.e. <12 h) was predictive of E. coli, Enterococcus spp., and human-specific MST marker concentrations (Fecal Bacteroides, BacHum, and HF183) in receiving waters. The strong correlation between 12-hr antecedent rainfall and Enterococcus spp. (r = 0.57, p < 0.001, n = 92) suggests a predictive model could be developed based on rainfall to communicate risk for bathers. Additional molecular marker data indicates that the delivery of fecal sources is complex and highly variable, likely due to the influence of tidal influx (saltwater intrusion from the estuary) into the low-lying stormwater pipes. In particular, elevated MST marker concentrations (up to 2.56 × 10⁴ gene copies HF183/mL) were observed in standing water near surcharging street storm drain. These data are being used to establish a baseline for stormwater dynamics prior to dramatic rainfall in 2018 and to characterize the interaction between complex stormwater dynamics and water quality impairment in coastal NC.

Predicting Escherichia coli loads in cascading dams with machine learning: An integration of hydrometeorology, animal density and grazing pattern

Accurate prediction of Escherichia coli contamination in surface waters is challenging due to considerable uncertainty in the physical, chemical and biological variables that control E. coli occurrence and sources in surface waters. This study proposes a novel approach by integrating hydro-climatic variables as well as animal density and grazing pattern in the feature selection modeling phase to increase E. coli prediction accuracy for two cascading dams at the US Meat Animal Research Center (USMARC), Nebraska. Predictive models were developed using regression techniques and an artificial neural network (ANN). Two adaptive neuro-fuzzy inference system (ANFIS) structures including subtractive clustering and fuzzy c-means (FCM) clustering were also used to develop models for predicting E. coli. The performances of the predictive models were evaluated and compared using root mean squared log error (RMSLE). Cross-validation and model performance results indicated that although the majority of models predicted E. coli accurately, ANFIS models resulted in fewer errors compared to the other models. The ANFIS models have the potential to be used to predict E. coli concentration for intervention plans and monitoring programs for cascading dams, and to implement effective best management practices for grazing and irrigation during the growing season.

Integration of weather conditions for predicting microbial water quality using Bayesian Belief Networks

Levels of fecal indicator bacteria (FIB) provide a surrogate measure of the microbial quality of water used for a wide range of applications. Despite the common use of these measures, a significant limitation is a delay in results due to the time required for cultivation and enumeration of FIB. Testing requires at least 18-24 h, and therefore, FIB cannot be used to identify current or real-time microbial water quality. An approach of nowcasting or empirical modelling approaches that incorporate water quality, environmental, and weather variables to predict FIB levels in real-time has been developed with some success. However, FIB levels are dependent on a complex interaction of numerous variables, which can be challenging to model with ordinary linear regression or classification methods most commonly applied. In this study, novel use of Bayesian Belief Networks (BBNs) that allow for a probabilistic representation of complex variable interactions is investigated for real-time modelling of FIB levels surface waters. In particular, the integration of both water quality measures and current/historical weather for prediction of fecal coliforms and Escherichia coli levels is achieved using BBNs. For 4-bin classification of fecal coliform levels, BBNs increased prediction accuracy by 25%-54% compared to other previously used techniques including logistic regression, Naïve Bayes, and random forests. Binary prediction of E. coli levels exceeding a threshold of 20 CFU/100 mL was also significantly improved using BBNs with prediction accuracies >90% for all monitoring sites. Advantages of the BBN approach are also demonstrated identifying the ability to make predictions from incomplete monitoring data as well as probabilistic inference of variable importance in FIB levels. In particular, the results indicate that water quality surrogates such as conductivity are essential to real-time prediction of FIB. The results and models described in this work can be readily utilized to provide accurate and real-time assessments of FIB levels in surface waters utilizing commonly monitored parameters.

Storm runoff differentially influences the nutrient concentrations and microbial contamination at two distinct beaches in northern China

With the escalating coastal development and loss of vegetated landscape, the volume of storm runoff increases significantly in Chinese coastal cities. To protect human health and valuable recreational resources, it is necessary to develop a quantitative understanding of coastal pollution. Here we studied the influence of storm runoff on the nutrients and microbial pathogens at two popular bathing beaches in northern China. Dongshan Beach, located near the mouth of an urban river, is influenced by non-point source pollution while Tiger-Rock Beach, a coastal beach, is primarily influenced by a point source from a storm drain outfall. Storm runoff significantly (P < 0.001) decreased the salinity and Chl a post-storm at both the beaches, but only reduced the concentration of dissolved inorganic N at Tiger-Rock Beach. Escherichia coli decreased by 68.7% at Dongshan Beach, possibly due to the dilution effect of the stormflow, contradicting the notion of elevated fecal contamination in coastal beaches from storm runoff. Vibrio parahaemolyticus increased at both beaches post-storm, by 155.7% at Dongshan Beach and 136.7% at Tiger-Rock Beach. Regardless of storm impact, both E. coli and V. parahaemolyticus were much higher at Dongshan Beach than that at Tiger-Rock, suggesting the influence of different surrounding topographies. Lastly, the statistical models developed based on the environmental and microbial parameters regression showed predictive power (adjusted R² > 0.5) to estimate the concentration of E. coli at Dongshan Beach and V. parahaemolyticus at Tiger-Rock Beach. Overall, the results suggest the unique role of the individual beaches in attenuating the effect of rainfall on the concentration of microbial pathogens in bathing water quality and provide unique predictive models for recreational water management and public health protection.

Statistical models of fecal coliform levels in Pacific Northwest estuaries for improved shellfish harvest area closure decision making

There is a substantial need for tools that effectively predict spatial and temporal fecal pollution patterns in estuarine waters. In this study, statistical models of exceedances of shellfish fecal coliform (FC) water quality criteria were developed using a 10-year dataset of FC levels and environmental data. Performance (sensitivity, specificity, and predictive capacity) of five different types of models was tested (MLR regression, Tobit (censored) regression, Firth's binary logistic regression (BLR), classification trees, and mixed-effects regression) for each of three conditionally managed shellfish-harvesting areas in Tillamook Bay, Oregon (USA). The most influential variables were related to precipitation and river stage height in the wet season and wind and tidal-stage in the dry season. Classification tree and Firth's BLR approaches better explained exceedances of shellfish water quality standards than the current closure thresholds. Findings demonstrate the utility of statistical modeling approaches for improved management of shellfish harvesting waters.

Implementation of an automated beach water quality nowcast system at ten California oceanic beaches

Fecal indicator bacteria like Escherichia coli and entercococci are monitored at beaches around the world to reduce incidence of recreational waterborne illness. Measurements are usually made weekly, but FIB concentrations can exhibit extreme variability, fluctuating at shorter periods. The result is that water quality has likely changed by the time data are provided to beachgoers. Here, we present an automated water quality prediction system (called the nowcast system) that is capable of providing daily predictions of water quality for numerous beaches. We created nowcast models for 10 California beaches using weather, oceanographic, and other environmental variables as input to tuned regression models to predict if FIB concentrations were above single sample water quality standards. Rainfall was used as a variable in nearly every model. The models were calibrated and validated using historical data. Subsequently, models were implemented during the 2017 swim season in collaboration with local beach managers. During the 2017 swim season, the median sensitivity of the nowcast models was 0.5 compared to 0 for the current method of using day-to-week old measurements to make beach posting decisions. Model specificity was also high (median of 0.87). During the implementation phase, nowcast models provided an average of 140 additional days per beach of updated water quality information to managers when water quality measurements were not made. The work presented herein emphasizes that a one-size-fits all approach to nowcast modeling, even when beaches are in close proximity, is infeasible. Flexibility in modeling approaches and adaptive responses to modeling and data challenges are required when implementing nowcast models for beach management.

Maxent estimation of aquatic Escherichia coli stream impairment

BACKGROUND

The leading cause of surface water impairment in United States' rivers and streams is pathogen contamination. Although use of fecal indicators has reduced human health risk, current approaches to identify and reduce exposure can be improved. One important knowledge gap within exposure assessment is characterization of complex fate and transport processes of fecal pollution. Novel modeling processes can inform watershed decision-making to improve exposure assessment.

METHODS

We used the ecological model, Maxent, and the fecal indicator bacterium Escherichia coli to identify environmental factors associated with surface water impairment. Samples were collected August, November, February, and May for 8 years on Sinking Creek in Northeast Tennessee and analyzed for 10 water quality parameters and E. coli concentrations. Univariate and multivariate models estimated probability of impairment given the water quality parameters. Model performance was assessed using area under the receiving operating characteristic (AUC) and prediction accuracy, defined as the model's ability to predict both true positives (impairment) and true negatives (compliance). Univariate models generated action values, or environmental thresholds, to indicate potential E. coli impairment based on a single parameter. Multivariate models predicted probability of impairment given a suite of environmental variables, and jack-knife sensitivity analysis removed unresponsive variables to elicit a set of the most responsive parameters.

RESULTS

Water temperature univariate models performed best as indicated by AUC, but alkalinity models were the most accurate at correctly classifying impairment. Sensitivity analysis revealed that models were most sensitive to removal of specific conductance. Other sensitive variables included water temperature, dissolved oxygen, discharge, and NO3. The removal of dissolved oxygen improved model performance based on testing AUC, justifying development of two optimized multivariate models; a 5-variable model including all sensitive parameters, and a 4-variable model that excluded dissolved oxygen.

DISCUSSION

Results suggest that E. coli impairment in Sinking Creek is influenced by seasonality and agricultural run-off, stressing the need for multi-month sampling along a stream continuum. Although discharge was not predictive of E. coli impairment alone, its interactive effect stresses the importance of both flow dependent and independent processes associated with E. coli impairment. This research also highlights the interactions between nutrient and fecal pollution, a key consideration for watersheds with multiple synergistic impairments. Although one indicator cannot mimic theplethora of existing pathogens in water, incorporating modeling can fine tune an indicator's utility, providing information concerning fate, transport, and source of fecal pollution while prioritizing resources and increasing confidence in decision making.

Development of a Nowcasting System Using Machine Learning Approaches to Predict Fecal Contamination Levels at Recreational Beaches in Korea

Microbial contamination in beach water poses a public health threat due to waterborne diseases. To reduce the risk of exposure to fecal contamination, informing beachgoers in advance about the microbial water quality is important. Currently, determining the level of fecal contamination takes 24 h. The objective of this study is to predict the current level of fecal contamination (enterococcus [ENT] and ) using readily available environmental variables. Artificial neural network (ANN) and support vector regression (SVR) models were constructed using data from the Haeundae and Gwangalli Beaches in Busan City. The input variables included the tidal level, air and water temperature, solar radiation, wind direction and velocity, precipitation, discharge from the wastewater treatment plant, and suspended solid concentration in beach water. The dependence of fecal contamination on the input variables was statistically evaluated; precipitation, discharge from the wastewater treatment plant, and wind direction at the two beaches were positively correlated to the changes in the two bacterial concentrations ( < 0.01), whereas solar radiation was negatively correlated ( < 0.01). The performance of the ANN model for predicting ENT and at Gwangalli Beach was significantly higher than that of the SVR model with the training dataset ( < 0.05). Based on the comparison of residual values between the predicted and observed fecal indicator bacteria concentrations in two models, the ANN demonstrated better performance than SVR. This study suggests an effective prediction method to determine whether a beach is safe for recreational use.

Precipitation thresholds for fecal bacterial indicators in the Chesapeake Bay

Many coastal states of the United States restrict harvest of shellfish from select areas based on some environmental trigger. Such areas are classified as being conditionally approved. In Maryland, the trigger is an inch or more of rainfall that has fallen in the last 24 h. This study used 11 years of monitoring data to test the relationship between daily rainfall totals and densities of fecal indicators in Maryland shellfish harvest waters. Precipitation and fecal coliform (FC) water monitoring data from 2004 to 2014 were matched by date and watershed. The influence of antecedent rainfall conditions (i.e. rainfall in the preceding days or weeks) and the distance of each monitoring station to land were compared to the percent of samples exceeding the FDA criterion for managing shellfish harvest areas. Sample stations beyond 1000m from land had FC densities consistently below the FDA criterion and were excluded from further analysis. Rainfall events greater than an inch tended to result in significantly elevated FC for the following two days, followed by lower levels thereafter. The total amount of rain in the last three weeks was positively related to the proportion of samples with FC greater than the FDA criterion. Bay-wide, the percent of samples exceeding the FDA criterion rose from seven percent for rainfall less than an inch to 37% following one or more inches of rain. Watersheds were classified based on the percent of FC densities over the criterion when rainfall was an inch or more, with 41 of 81 watersheds showing FC responses indicative of potential conditionally approved areas, those shellfish growing areas where the one inch precipitation trigger may be applied. These areas largely overlapped the current conditionally approved areas defined by Maryland. The percent of open water, wetlands, and poorly drained soils explained a significant amount of the variability (R² = 0.72) in the difference in percent of samples exceeding the FDA criterion when rainfall was greater than an inch and when it was less than an inch. Logistic regression analysis showed that the current trigger of one inch of rain in 24 h is predictive of FC densities over the FDA criterion, though the appropriate threshold will most likely depend on how far the particular shellfish growing area is from land and antecedent rain conditions. In watersheds with relatively high percentages of open water to total watershed size, higher rainfall thresholds might be appropriate. The approach taken in this study could be applied to individual stations and sub-watersheds, potentially allowing the reclassification of some shellfish harvest areas.

Geostatistical Prediction of Microbial Water Quality Throughout a Stream Network Using Meteorology, Land Cover, and Spatiotemporal Autocorrelation

Predictive modeling is promising as an inexpensive tool to assess water quality. We developed geostatistical predictive models of microbial water quality that empirically modeled spatiotemporal autocorrelation in measured fecal coliform (FC) bacteria concentrations to improve prediction. We compared five geostatistical models featuring different autocorrelation structures, fit to 676 observations from 19 locations in North Carolina's Jordan Lake watershed using meteorological and land cover predictor variables. Though stream distance metrics (with and without flow-weighting) failed to improve prediction over the Euclidean distance metric, incorporating temporal autocorrelation substantially improved prediction over the space-only models. We predicted FC throughout the stream network daily for one year, designating locations "impaired", "unimpaired", or "unassessed" if the probability of exceeding the state standard was ≥90%, ≤10%, or >10% but <90%, respectively. We could assign impairment status to more of the stream network on days any FC were measured, suggesting frequent sample-based monitoring remains necessary, though implementing spatiotemporal predictive models may reduce the number of concurrent sampling locations required to adequately assess water quality. Together, these results suggest that prioritizing sampling at different times and conditions using geographically sparse monitoring networks is adequate to build robust and informative geostatistical models of water quality impairment.

Microbial quality of agricultural water in Central Florida

The microbial quality of water that comes into the edible portion of produce is believed to directly relate to the safety of produce, and metrics describing indicator organisms are commonly used to ensure safety. The US FDA Produce Safety Rule (PSR) sets very specific microbiological water quality metrics for agricultural water that contacts the harvestable portion of produce. Validation of these metrics for agricultural water is essential for produce safety. Water samples (500 mL) from six agricultural ponds were collected during the 2012/2013 and 2013/2014 growing seasons (46 and 44 samples respectively, 540 from all ponds). Microbial indicator populations (total coliforms, generic Escherichia coli, and enterococci) were enumerated, environmental variables (temperature, pH, conductivity, redox potential, and turbidity) measured, and pathogen presence evaluated by PCR. Salmonella isolates were serotyped and analyzed by pulsed-field gel electrophoresis. Following rain events, coliforms increased up to 4.2 log MPN/100 mL. Populations of coliforms and enterococci ranged from 2 to 8 and 1 to 5 log MPN/100 mL, respectively. Microbial indicators did not correlate with environmental variables, except pH (P<0.0001). The invA gene (Salmonella) was detected in 26/540 (4.8%) samples, in all ponds and growing seasons, and 14 serotypes detected. Six STEC genes were detected in samples: hly (83.3%), fliC (51.8%), eaeA (17.4%), rfbE (17.4%), stx-I (32.6%), stx-II (9.4%). While all ponds met the PSR requirements, at least one virulence gene from Salmonella (invA-4.8%) or STEC (stx-I-32.6%, stx-II-9.4%) was detected in each pond. Water quality for tested agricultural ponds, below recommended standards, did not guarantee the absence of pathogens. Investigating the relationships among physicochemical attributes, environmental factors, indicator microorganisms, and pathogen presence allows researchers to have a greater understanding of contamination risks from agricultural surface waters in the field.

Citizen science-based water quality monitoring: Constructing a large database to characterize the impacts of combined sewer overflow in New York City

To protect recreational water users from waterborne pathogen exposure, it is crucial that waterways are monitored for the presence of harmful bacteria. In NYC, a citizen science campaign is monitoring waterways impacted by inputs of storm water and untreated sewage during periods of rainfall. However, the spatial and temporal scales over which the monitoring program can sample are constrained by cost and time, thus hindering the construction of databases that benefit both scientists and citizens. In this study, we first illustrate the scientific value of a citizen scientist monitoring campaign by using the data collected through the campaign to characterize the seasonal variability of sampled bacterial concentration as well as its response to antecedent rainfall. Second, we examine the efficacy of the HyServe Compact Dry ETC method, a lower cost and time-efficient alternative to the EPA-approved IDEXX Enterolert method for fecal indicator monitoring, through a paired sample comparison of IDEXX and HyServe (total of 424 paired samples). The HyServe and IDEXX methods return the same result for over 80% of the samples with regard to whether a water sample is above or below the EPA's recreational water quality criteria for a single sample of 110 enterococci per 100mL. The HyServe method classified as unsafe 90% of the 119 water samples that were classified as having unsafe enterococci concentrations by the more established IDEXX method. This study seeks to encourage other scientists to engage with citizen scientist communities and to also pursue the development of cost- and time-efficient methodologies to sample environmental variables that are not easily collected or analyzed in an automated manner.

Spatial and temporal microbial pollution patterns in a tropical estuary during high and low river flow conditions

Spatial and temporal patterns of coastal microbial pollution are not well documented. Our study examined these patterns through measurements of fecal indicator bacteria (FIB), nutrients, and physiochemical parameters in Hilo Bay, Hawai'i, during high and low river flow. >40% of samples tested positive for the human-associated Bacteroides marker, with highest percentages near rivers. Other FIB were also higher near rivers, but only Clostridium perfringens concentrations were related to discharge. During storms, FIB concentrations were three times to an order of magnitude higher, and increased with decreasing salinity and water temperature, and increasing turbidity. These relationships and high spatial resolution data for these parameters were used to create Enterococcus spp. and C. perfringens maps that predicted exceedances with 64% and 95% accuracy, respectively. Mapping microbial pollution patterns and predicting exceedances is a valuable tool that can improve water quality monitoring and aid in visualizing FIB hotspots for management actions.

Isolating the impact of septic systems on fecal pollution in streams of suburban watersheds in Georgia, United States

The presence of multiple sources of fecal pollution at the watershed level presents challenges to efforts aimed at identifying the influence of septic systems. In this study multiple approaches including targeted sampling and monitoring of host-specific Bacteroidales markers were used to identify the impact of septic systems on microbial water quality. Twenty four watersheds with septic density ranging from 8 to 373 septic units/km² were monitored for water quality under baseflow conditions over a 3-year period. The levels of the human-associated HF183 marker, as well as total and ruminant Bacteroidales, were quantified using quantitative polymerase chain reaction. Human-associated Bacteroidales yield was significantly higher in high density watersheds compared to low density areas and was negatively correlated (r = -0.64) with the average distance of septic systems to streams in the spring season. The human marker was also positively correlated with the total Bacteroidales marker, suggesting that the human source input was a significant contributor to total fecal pollution in the study area. Multivariable regression analysis indicates that septic systems, along with forest cover, impervious area and specific conductance could explain up to 74% of the variation in human fecal pollution in the spring season. The results suggest septic system impact through contributions to groundwater recharge during baseflow or failing septic system input, especially in areas with >87 septic units/km². This study supports the use of microbial source tracking approaches along with traditional fecal indicator bacteria monitoring and land use characterization in a tiered approach to isolate the influence of septic systems on water quality in mixed-use watersheds.

Monitoring and predicting the fecal indicator bacteria concentrations from agricultural, mixed land use and urban stormwater runoff

While the urban runoff are increasingly being studied as a source of fecal indicator bacteria (FIB), less is known about the occurrence of FIB in watershed with mixed land use and ongoing land use and land cover (LULC) change. In this study, Escherichia coli (EC) and fecal streptococcus (FS) were monitored from 2012 to 2013 in agricultural, mixed and urban LULC and analyzed according to the most probable number (MPN). Pearson correlation was used to determine the relationship between FIB and environmental parameters (physicochemical and hydrometeorological). Multiple linear regressions (MLR) were used to identify the significant parameters that affect the FIB concentrations and to predict the response of FIB in LULC change. Overall, the FIB concentrations were higher in urban LULC (EC=3.33-7.39; FS=3.30-7.36log10MPN/100mL) possibly because of runoff from commercial market and 100% impervious cover (IC). Also, during early-summer season; this reflects a greater persistence and growth rate of FIB in a warmer environment. During intra-event, however, the FIB concentrations varied according to site condition. Anthropogenic activities and IC influenced the correlation between the FIB concentrations and environmental parameters. Stormwater temperature (TEMP), turbidity, and TSS positively correlated with the FIB concentrations (p>0.01), since IC increased, implying an accumulation of bacterial sources in urban activities. TEMP, BOD5, turbidity, TSS, and antecedent dry days (ADD) were the most significant explanatory variables for FIB as determined in MLR, possibly because they promoted the FIB growth and survival. The model confirmed the FIB concentrations: EC (R(2)=0.71-0.85; NSE=0.72-0.86) and FS (R(2)=0.65-0.83; NSE=0.66-0.84) are predicted to increase due to urbanization. Therefore, these findings will help in stormwater monitoring strategies, designing the best management practice for FIB removal and as input data for stormwater models.

Storm loads of culturable and molecular fecal indicators in an inland urban stream

Elevated concentrations of fecal indicator bacteria in receiving waters during wet-weather flows are a considerable public health concern that is likely to be exacerbated by future climate change and urbanization. Knowledge of factors driving the fate and transport of fecal indicator bacteria in stormwater is limited, and even less is known about molecular fecal indicators, which may eventually supplant traditional culturable indicators. In this study, concentrations and loading rates of both culturable and molecular fecal indicators were quantified throughout six storm events in an instrumented inland urban stream. While both concentrations and loading rates of each fecal indicator increased rapidly during the rising limb of the storm hydrographs, it is the loading rates rather than instantaneous concentrations that provide a better estimate of transport through the stream during the entire storm. Concentrations of general fecal indicators (both culturable and molecular) correlated most highly with each other during storm events but not with the human-associated HF183 Bacteroides marker. Event loads of general fecal indicators most strongly correlated with total runoff volume, maximum discharge, and maximum turbidity, while event loads of HF183 most strongly correlated with the time to peak flow in a hydrograph. These observations suggest that collection of multiple samples during a storm event is critical for accurate predictions of fecal indicator loading rates and total loads during wet-weather flows, which are required for effective watershed management. In addition, existing predictive models based on general fecal indicators may not be sufficient to predict source-specific genetic markers of fecal contamination.

Predictive statistical models linking antecedent meteorological conditions and waterway bacterial contamination in urban waterways

Although the relationships between meteorological conditions and waterway bacterial contamination are being better understood, statistical models capable of fully leveraging these links have not been developed for highly urbanized settings. We present a hierarchical Bayesian regression model for predicting transient fecal indicator bacteria contamination episodes in urban waterways. Canals, creeks, and rivers of the New York City harbor system are used to examine the model. The model configuration facilitates the hierarchical structure of the underlying system with weekly observations nested within sampling sites, which in turn were nested inside of the harbor network. Models are compared using cross-validation and a variety of Bayesian and classical model fit statistics. The uncertainty of predicted enterococci concentration values is reflected by sampling from the posterior predictive distribution. Issuing predictions with the uncertainty reasonably reflected allows a water manager or a monitoring agency to issue warnings that better reflect the underlying risk of exposure. A model using only antecedent meteorological conditions is shown to correctly classify safe and unsafe levels of enterococci with good accuracy. The hierarchical Bayesian regression approach is most valuable where transient fecal indicator bacteria contamination is problematic and drainage network data are scarce.

Sunny with a chance of gastroenteritis: predicting swimmer risk at California beaches

Traditional beach management that uses concentrations of cultivatable fecal indicator bacteria (FIB) may lead to delayed notification of unsafe swimming conditions. Predictive, nowcast models of beach water quality may help reduce beach management errors and enhance protection of public health. This study compares performances of five different types of statistical, data-driven predictive models: multiple linear regression model, binary logistic regression model, partial least-squares regression model, artificial neural network, and classification tree, in predicting advisories due to FIB contamination at 25 beaches along the California coastline. Classification tree and the binary logistic regression model with threshold tuning are consistently the best performing model types for California beaches. Beaches with good performing models usually have a rainfall/flow related dominating factor affecting beach water quality, while beaches having a deteriorating water quality trend or low FIB exceedance rates are less likely to have a good performing model. This study identifies circumstances when predictive models are the most effective, and suggests that using predictive models for public notification of unsafe swimming conditions may improve public health protection at California beaches relative to current practices.

Predicting water quality at Santa Monica Beach: evaluation of five different models for public notification of unsafe swimming conditions

Bathing beaches are monitored for fecal indicator bacteria (FIB) to protect swimmers from unsafe conditions. However, FIB assays take ∼24 h and water quality conditions can change dramatically in that time, so unsafe conditions cannot presently be identified in a timely manner. Statistical, data-driven predictive models use information on environmental conditions (i.e., rainfall, turbidity) to provide nowcasts of FIB concentrations. Their ability to predict real time FIB concentrations can make them more accurate at identifying unsafe conditions than the current method of using day or older FIB measurements. Predictive models are used in the Great Lakes, Hong Kong, and Scotland for beach management, but they are presently not used in California - the location of some of the world's most popular beaches. California beaches are unique as point source pollution has generally been mitigated, the summer bathing season receives little to no rainfall, and in situ measurements of turbidity and salinity are not readily available. These characteristics may make modeling FIB difficult, as many current FIB models rely heavily on rainfall or salinity. The current study investigates the potential for FIB models to predict water quality at a quintessential California Beach: Santa Monica Beach. This study compares the performance of five predictive models, multiple linear regression model, binary logistic regression model, partial least square regression model, artificial neural network, and classification tree, to predict concentrations of summertime fecal coliform and enterococci concentrations. Past measurements of bacterial concentration, storm drain condition, and tide level are found to be critical factors in the predictive models. The models perform better than the current beach management method. The classification tree models perform the best; for example they correctly predict 42% of beach postings due to fecal coliform exceedances during model validation, as compared to 28% by the current method. Artificial neural network is the second best model which minimizes the number of incorrect beach postings. The binary logistic regression model also gives promising results, comparable to classification tree, by adjusting the posting decision thresholds to maximize correct beach postings. This study indicates that predictive models hold promise as a beach management tool at Santa Monica Beach. However, there are opportunities to further refine predictive models.

Comparisons of statistical models to predict fecal indicator bacteria concentrations enumerated by qPCR- and culture-based methods

Recently, the United States Environmental Protection Agency (USEPA) revised their recreational water quality criteria, in which adjustments were made by approving enterococci (ENT) quantitative PCR (qPCR) as an alternative, rapid method and advocating the use of predictive models for water quality management. The implementation of qPCR-based methods and prediction models are meant to decrease the time between sample collection and public advisories and notifications. To date, few studies have compared qPCR-based models to culture-based prediction models and none of these studies have been conducted in coastal estuarine systems. In this study, we created prediction models using qPCR-based fecal indicator bacteria (FIB) data in dual-use recreational and shellfish harvesting waters and compared them to published ENT and Escherichia coli (EC) culture-based prediction models in eastern North Carolina estuaries. Furthermore, an empirical statistical model was created to predict qPCR inhibition levels so that proper remediation techniques can be applied when it is a problem. Predictor variable selection in both qPCR- and culture-based ENT models was very similar; both models included 14-day rain total, dissolved oxygen, and salinity/conductivity, with 89 and 90% of qPCR and culture data described, respectively. Using ENT management action thresholds, qPCR- and culture-based models showed high accuracy in management decisions. The qPCR model had 92 and 96% accuracy using the 110 and 1000 cell equivalents (CE)/100 ml thresholds, respectively, and the culture model had 90% accuracy in management decisions with the 110 MPN/100 ml threshold. EC models for qPCR- and culture-based concentrations used similar independent variables (14-day humidity, salinity/conductivity, a rain/storm variable, and a measure of air temperature), with each model explaining 26 and 55% of the data variation, respectively. When using different thresholds that were logs apart for management decisions, the two EC models accurately predicted management decisions; qPCR models correctly predicted management decisions 96 and 77% of the time (using 31 and 320 CE/100 ml, respectively) while culture models correctly predicted management decisions 96 and 88% percent of the time (with 31 and 320 MPN/100 ml, respectively). Equivalency between models was shown in our non-point source impacted estuaries, with ENT models performing slightly better than EC models. In addition, inhibition of the qPCR was a major issue that had to be addressed. An inhibition model was created with easily obtained meteorological data and accounted for a high level of data variability (adjusted R(2) = 0.82).

Recommendations following a multi-laboratory comparison of microbial source tracking methods

Microbial source tracking (MST) methods were evaluated in the Source Identification Protocol Project (SIPP), in which 27 laboratories compared methods to identify host sources of fecal pollution from blinded water samples containing either one or two different fecal types collected from California. This paper details lessons learned from the SIPP study and makes recommendations to further advance the field of MST. Overall, results from the SIPP study demonstrated that methods are available that can correctly identify whether particular host sources including humans, cows and birds have contributed to contamination in a body of water. However, differences between laboratory protocols and data processing affected results and complicated interpretation of MST method performance in some cases. This was an issue particularly for samples that tested positive (non-zero Ct values) but below the limits of quantification or detection of a PCR assay. Although false positives were observed, such samples in the SIPP study often contained the fecal pollution source that was being targeted, i.e., the samples were true positives. Given these results, and the fact that MST often requires detection of targets present in low concentrations, we propose that such samples be reported and identified in a unique category to facilitate data analysis and method comparisons. Important data can be lost when such samples are simply reported as positive or negative. Actionable thresholds were not derived in the SIPP study due to limitations that included geographic scope, age of samples, and difficulties interpreting low concentrations of target in environmental samples. Nevertheless, the results of the study support the use of MST for water management, especially to prioritize impaired waters in need of remediation. Future integration of MST data into quantitative microbial risk assessments and other models could allow managers to more efficiently protect public health based on site conditions.

Evaluation of equivalence between different methods for enumeration of fecal indicator bacteria before and after adoption of the new Bathing Water Directive and risk assessment of pollution

The quality of bathing water is of considerable public importance due to the possibility of fecal contamination. In 2009, Croatia implemented the new European Bathing Water Directive (BWD, 2006/7/EC) establishing stricter microbiological standards for new parameters with new reference methods. This study aims to evaluate the equivalence of different methods according to the old and revised BWD and to provide the possibility of data comparison. Furthermore, the directive requires the establishment of the bathing water profile (BWP) for pollution risk assessment. The estimation of consistency of pollution risk assessment with obtained microbiological results was also performed. Six marine beaches of the Municipality of Rijeka (Croatia) were examined during the 2009 season. Statistical analysis showed equivalence between determination methods for fecal contamination indicators. Based on the current water classification results, the need for correction of estimated pollution risks and recommendations for inclusion of historical microbiological data during BWP enactment was noticed.

Predicting Salmonella populations from biological, chemical, and physical indicators in Florida surface waters

Coliforms, Escherichia coli, and various physicochemical water characteristics have been suggested as indicators of microbial water quality or index organisms for pathogen populations. The relationship between the presence and/or concentration of Salmonella and biological, physical, or chemical indicators in Central Florida surface water samples over 12 consecutive months was explored. Samples were taken monthly for 12 months from 18 locations throughout Central Florida (n = 202). Air and water temperature, pH, oxidation-reduction potential (ORP), turbidity, and conductivity were measured. Weather data were obtained from nearby weather stations. Aerobic plate counts and most probable numbers (MPN) for Salmonella, E. coli, and coliforms were performed. Weak linear relationships existed between biological indicators (E. coli/coliforms) and Salmonella levels (R(2) < 0.1) and between physicochemical indicators and Salmonella levels (R(2) < 0.1). The average rainfall (previous day, week, and month) before sampling did not correlate well with bacterial levels. Logistic regression analysis showed that E. coli concentration can predict the probability of enumerating selected Salmonella levels. The lack of good correlations between biological indicators and Salmonella levels and between physicochemical indicators and Salmonella levels shows that the relationship between pathogens and indicators is complex. However, Escherichia coli provides a reasonable way to predict Salmonella levels in Central Florida surface water through logistic regression.