PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of aemPermissionsJournals.ASM.orgJournalAEM ArticleJournal InfoAuthorsReviewers
 
Appl Environ Microbiol. 2017 February 15; 83(4): e02881-16.
Published online 2017 February 1. Prepublished online 2016 December 9. doi:  10.1128/AEM.02881-16
PMCID: PMC5288815

Distribution and Differential Survival of Traditional and Alternative Indicators of Fecal Pollution at Freshwater Beaches

Shuang-Jiang Liu, Editor
Shuang-Jiang Liu, Chinese Academy of Sciences;

ABSTRACT

Alternative indicators have been developed that can be used to identify host sources of fecal pollution, yet little is known about how their distribution and fate compare to traditional indicators. Escherichia coli and enterococci were widely distributed at the six beaches studied and were detected in almost 95% of water samples (n = 422) and 100% of sand samples (n = 400). Berm sand contained the largest amount of E. coli (P < 0.01), whereas levels of enterococci were highest in the backshore (P < 0.01). E. coli and enterococci were the lowest in water, using a weight-to-volume comparison. The gull-associated Catellicoccus marimammalium (Gull2) marker was found in over 80% of water samples, regardless of E. coli levels, and in 25% of sand samples. Human-associated Bacteroides (HB) and Lachnospiraceae (Lachno2) were detected in only 2.4% of water samples collected under baseflow and post-rain conditions but produced a robust signal after a combined sewage overflow, despite low E. coli concentrations. Burdens of E. coli and enterococci in water and sand were disproportionately high in relation to alternative indicators when comparing environmental samples to source material. In microcosm studies, Gull2, HB, and Lachno2 quantitative PCR (qPCR) signals were reduced twice as quickly as those from E. coli and enterococci and approximately 20% faster than signals from culturable E. coli. High concentrations of alternative indicators in source material illustrated their high sensitivity for the identification of fecal sources; however, differential survival and the potential for long-term persistence of traditional fecal indicators complicate the use of alternative indicator data to account for the levels of E. coli and enterococci in environmental samples.

IMPORTANCE E. coli and enterococci are general indicators of fecal pollution and may persist in beach sand, making their use problematic for many applications. This study demonstrates that gull fecal pollution is widespread at Great Lakes beaches, whereas human and ruminant contamination is evident only after major rain events. An exploration of sand as a reservoir for indicators found that E. coli was ubiquitous, while gull host markers were detected in only 25% of samples. In situ sand beach microcosms provided decay rate constants for E. coli and enterococci relative to alternative indicators, which establish comparative benchmarks that would be helpful to distinguish recent from past pollution. Overall, alternative indicators are useful for identifying sources and assessing potentially high health risk contamination events; however, beach managers should be cautious in attempting to directly link their detection to the levels of E. coli or enterococci.

KEYWORDS: beaches, water quality, human-associated indicators, gull-associated indicators, qPCR

INTRODUCTION

Fecal contamination of recreational waters can be a serious threat to public health. Due to the vast diversity of fecal-borne human pathogens, the USEPA has recommended the use of fecal indicator bacteria (FIB), commonly Escherichia coli and enterococci, to determine if fecal pollution is present. Historically, single-day recreational water advisory thresholds were 235 CFU/100 ml for E. coli and 61 CFU/100 ml for enterococci in freshwater (1). The recent Recreational Water Quality Criteria, published by the USEPA in 2012, established a similar range for beach action values of 235 CFU/100 ml for E. coli and 70 CFU/100 ml enterococci for freshwater, which relates to an unacceptable health risk to beachgoers of 36 illnesses per 1,000 people (2). The advisory and closure FIB threshold values reflect multiple epidemiological studies that assessed the predictive nature of FIB based on the rate of illness reported by beachgoers (3).

Despite the wide use of FIB in marine and freshwater systems, it has been shown that enterococci and E. coli are poorly associated with the presence of human pathogens in environmental waters (4,6). Due to their ubiquitous nature in warm-blooded animals (7,9), fecal indicators can only indicate that fecal contamination may be present; however, certain host sources are more likely than others to carry human pathogens (10). Additionally, E. coli, the most commonly used indicator in freshwater systems, has been found to survive in the environment (11,14). Environmental persistence of fecal indicator bacteria undermines the utility for recreational water quality monitoring because the presence of these organisms would not necessarily indicate a recent contamination event, and in some cases, it can lead to an overestimation of the associated public health risk.

To address some of the pitfalls associated with traditional FIB monitoring, alternative indicators have been identified that are host associated and are a major focus of current water quality research. The use of alternative indicators shows promise for the detection of fecal pollution sources and can lead to increased accuracy in identifying health risks to beachgoers as well as aid in the mitigation of pollution sources. Alternative indicator assays commonly target anaerobic fecal bacteria due to their high abundance in animal and human feces and their limited survival in the external environment (15,17). Bacteria within the order Bacteroidales are common targets for quantitative PCR (qPCR) assays and have been used for the detection of sewage (18,21), ruminant (19, 22), canine (19), and avian (23, 24) sources of fecal pollution. Certain members within the order Clostridiales have also demonstrated host-associated patterns (25). Our lab has developed a qPCR assay that targets the second most abundant human-associated Lachnospiraceae (Lachno2) in sewage (26) and has been used to track sewage contamination in environmental waters (27). Gull-associated qPCR assays targeting Catellicoccus marimammalium have also been developed (28,30), and field studies have demonstrated that gulls are a common source of degraded water quality at marine and freshwater recreational beaches (31).

Although recreational waters are the primary monitoring focus for beach managers, numerous studies have documented high concentrations of fecal indicators within beach sand (12, 14, 32). Recent research has reported the recovery of bacterial and viral human pathogens from beach sand, providing evidence that sand contact may play an important role in beach-associated gastrointestinal (GI) illness (33,35). Based on a potential bacterial pathogen reservoir in beach sand, the scientific community has called for implementation of a sand monitoring program (35).

An understanding of the concentrations of alternative host-associated indicators compared with traditional indicators in source fecal samples is needed to establish the prevalence and sensitivity of these indicators and to interpret environmental monitoring results. Benchtop microcosm studies have assessed the survival of E. coli and enterococci in beach sand; however, only a few studies have examined the decay of alternative indicators (33, 36,38). The dynamic conditions present in the beach environment cannot be readily replicated in the laboratory; thus, results from previous laboratory survival studies are difficult to directly apply to what might be expected in the environment.

This study employs a comprehensive survey of beaches in urban and rural areas to evaluate the efficacy of molecular methods to assess fecal contamination in sand and water samples. With the goal of providing valuable information to beach managers regarding the usefulness of alternative indicators, the present study combined field surveys and microcosm experiments to explore the utility of traditional and alternative fecal indicators within beach sand and water. The present study sought to test the following hypotheses: (i) alternative indicators are more sensitive than E. coli or enterococci for detecting fecal pollution; (ii) differential survival can affect the relative concentrations of traditional and alternative fecal indicators in the beach environment; and (iii) sand acts as a long-term reservoir for E. coli, and E. coli may be detected without evidence of host sources. Overall, this work demonstrates that alternative indicators are useful for detecting recent pollution events from specific sources. However, beach managers are challenged to respond to elevated levels of E. coli and enterococci, and the differences in persistence between alternative indicators and traditional indicators, coupled with the probability of multiple ongoing pollution inputs, precludes making inferences about the causes of elevated FIB levels from alternative indicator measurements.

RESULTS

Comparison of levels of E. coli and enterococci in the beach environment.

We compared the levels of culturable E. coli and enterococci across three distinct zones (backshore sand, berm sand, and water) to determine if FIB reservoirs are present in sand and if differences exist between zones. Overall, we found that levels of E. coli at the six beaches were significantly higher in berm sand (P < 0.01) than in backshore sand, as well as water in a comparison of equal weight to volume (Fig. 1). Densities of enterococci were significantly higher in the backshore sand (P < 0.01) than in berm sand and water samples. Overall, some beaches had higher densities of E. coli and enterococci in berm and backshore sand than others; however, in water, the mean densities of each indicator were very similar at all beaches. Although berm sand had higher levels of E. coli than in water, the concentrations of E. coli in paired sand and water samples collected along the same transect were correlated (Spearman's rho = 0.65, P < 0.01). The geometric mean concentrations of E. coli at the different beaches ranged from 35 to 88 CFU/100 ml for water, and 1.7 × 102 to 3.2 × 103 CFU/100 g of berm sand and 2 to 87 CFU/100 g of backshore sand. Levels of enterococci ranged from 13 to 30 CFU/100 ml for water, 60 to 1.4 × 103 CFU/100 g for berm sand, and 47 to 1.3 × 103 CFU/100 g for backshore sand. Of the 422 water samples collected, 84 (20%) samples exceeded the USEPA E. coli beach action value of 235 CFU/100 ml. We used the average E. coli levels across all water sites at a beach to determine advisory days. Similar percentages of advisory days at Bayview Beach (BV; 28%), Bradford Beach (BB; 17%), Atwater Park Beach (ATW; 27%), Doctor's Park (DP; 25%), Kohler-Andrae State Park (KA; 27%), and Point Beach State Park (PB; 18%) were observed at the six beaches.

FIG 1
Heat map illustrating the geometric mean concentrations of E. coli and enterococci, as measured by membrane filtration, in sand and water samples collected at beaches during 2012 and 2013. Sand samples were compared to water samples on a weight-to-volume ...

Alternative indicator qPCR analysis.

We examined water samples with E. coli concentrations above the advisory threshold (n = 84) and below the advisory threshold (n = 60) to determine the presence of host-associated indicators, particularly for human or cattle sources, which are known to pose a human health risk, and gulls, which can also pose a health risk and are a common source of fecal pollution at beaches. In addition to the selected water samples, “paired” sand samples (i.e., sand collected on the same date and at the same site as water samples) were also analyzed for the presence of host-associated markers. Of beach samples with high E. coli levels, the majority of paired sand samples (n = 64/69) also exceeded the advisory criteria on weight-to-volume comparison (i.e., ≥235 CFU E. coli/100 g). Of the beach water with low levels of E. coli, approximately half of the paired samples (n = 20/41) also had <235 CFU/100 ml on a volume-mass basis. Water samples that were collected under combined sewer overflow (CSO) conditions were considered independently (n = 20).

The detection frequency of host-associated markers (Gull2, Lachno2, HB, and ruminant-associated Bacteroidetes [BacR]) in beach water with high E. coli levels and paired sand is shown in Fig. 2. The Gull2 marker was detected more frequently than any other host-associated marker in both water and sand and occurred in 83% of the water samples that had high E. coli levels and 28% of the paired sand samples. The Gull2 marker concentrations in water samples with high E. coli levels ranged from 1.0 × 102 to 6.6 × 105 copy number (CN)/100 ml and 1.0 × 104 to 4.3 × 105 CN/100 g in the paired sand samples. The human-associated HB marker was detected in 2.4% of the water samples above the E. coli advisory threshold and occurred only at the BV and ATW beaches (which are located in the greater Milwaukee area); HB was absent from the BB, DP, KA, and PB beaches. The HB marker concentrations, when detected, were relatively low compared to those in untreated sewage, ranging from 3.3 × 102 to 9.2 × 102 CN/100 ml in water, and they were absent in all paired sand samples. The second human-associated marker, Lachno2, was detected in 15% of the water samples above the advisory threshold and was found at four of the six beach sites, all at relatively low levels. In paired sand samples, the HB marker was absent, and the Lachno2 marker was detected in 7% of samples; however, a large number of samples (n = 49/69) were detected but not quantifiable (DNQ) for Lachno2.

FIG 2
Alternative indicator detection frequencies measured during advisory conditions for water samples with ≥235 E. coli/100 ml (n = 84) (A) and sand samples paired by date and transect (n = 69) (B) collected during 2012 and 2013. Water samples collected ...

The BacR assay was analyzed only for the two northernmost beaches, PB and KA, due to the close proximity to agricultural operations. For water samples with high E. coli levels, the BacR marker was detected in 16% of KA samples but was undetected in paired sand samples. In PB water samples with high E. coli levels, the BacR marker was detected in 53% of water samples and in 56% of paired sand samples. For both PB and KA, when BacR was detected, the average concentration was 1.8 × 103 CFU/100 ml in water and 1.5 × 104 CFU/100 g in sand.

We also examined samples with low levels of E. coli (<235 CFU/100 ml or g) to assess the occurrence of host markers in cases where water samples would be considered to have acceptable water quality. We found similar distributions and levels of gull contamination in these samples compared to samples with high E. coli levels. The Gull2 marker was detected in 82% (n = 49/60) of water samples with low E. coli levels and 15% (n = 6/41) of the paired sand samples. There was not a significant difference in the gull marker levels in water samples with high or low E. coli levels (P = 0.08). The two human-specific markers were absent in all water samples with <235 CFU of E. coli/100 ml; however, there was a small number of samples that had DNQ results (n = 1 for HB and n = 6 for Lachno2). In contrast, for the sand samples paired with these water samples, the Lachno2 marker was detected in 2/41 of sand samples and DNQ in 20/41 of samples. The HB marker was not detected in any paired sand samples with <235 CFU/100 ml. At the northern beaches potentially impacted by agricultural runoff, the ruminant marker was not detected in any samples with low E. coli levels.

Water samples collected at the ATW, BB, and BV beaches 1 to 3 days following a CSO had very low culturable E. coli concentrations, ranging between 0 and 23 CFU/100 ml. The human-associated markers were detected in 45% of the post-CSO samples. When detected, concentrations were relatively high and ranged from 2.7 × 103 to 1.4 × 104 CN/100 ml for HB and 1.5 × 103 to 1.2 × 104 CN/100 ml for Lachno2, which is one or more orders of magnitude higher than in water samples collected under non-CSO conditions. The two human-associated markers were highly correlated within samples collected post-CSO (Spearman's rho = 0.99, P < 0.01).

Concentrations of host-associated markers and ratios to FIB.

We assessed the concentrations and variability of markers in gull, untreated sewage, and ruminant fecal sources (Fig. S1). We found that gulls consistently had a high abundance of the Gull2 marker per gram of gull feces. Concentrations of qPCR markers for E. coli and enterococci, however, were highly varied in gull fecal samples and ranged from 6.9 × 104 to 1.1 × 1010 CN/g for E. coli and 7.0 × 105 to 2.1 × 108 CN/g for enterococci. Gull2 was, on average, three to four orders of magnitude higher than either E. coli or enterococci as measured by qPCR. In sewage, concentrations of Lachno2 were approximately 1.5-fold higher than those of HB. These two human-associated markers were at levels similar to those of enterococci but were two orders of magnitude higher than E. coli. The ruminant marker BacR was found at concentrations approximately 4-fold higher than the concentrations of E. coli and enterococci.

We compared the concentrations of alternative indicators, E. coli, and enterococci in fecal sources to their concentrations in environmental samples (Table 1). The ratios were highly varied, particularly for Gull2, as indicated by the high standard deviation (SD) of the mean, which was not unexpected given the high variability in source material. The range of ratios for each alternative indicator is shown in Fig. S2 to S5. The ratio of Gull2 to E. coli concentrations was significantly higher in gull fecal samples than in either sand or water samples (P < 0.01). The same results were found for ratios of Gull2 to enterococci. There was no significant difference in the ratio of Gull2 to E. coli or the ratio of Gull2 to enterococci in a comparison of sand and water beach samples, which could suggest that decay dynamics in these two matrices are similar or that gull fecal droppings constantly deposit these fecal organisms in consistent proportions. Ratios of the BacR marker to E. coli or enterococci in cow feces compared to environmental samples were significantly higher (P < 0.01). Far fewer environmental samples were positive for the human markers; therefore, trends were difficult to assess. Ratios of human markers to enterococci actually increased in environmental samples compared with the fecal source material.

TABLE 1
Average concentrations and standard deviations of host-associated marker sequences detected in source material and environmental samples, and comparison to fecal indicator abundance

FIB, gull, and sewage marker decay.

In addition to extensive field sampling, we assessed the decay of traditional and alternative indicators in the beach environment using in situ microcosm experiments. For gull microcosms, the initial average concentrations (t = 0) of culturable E. coli and enterococci were 5.0 × 105 and 8.9 × 104 CFU/g, respectively. The initial average concentrations of E. coli, enterococci, and Gull2 markers detected by qPCR were 2.4 × 104, 1.0 × 106, and 4.1 × 106 CN/g, respectively. The concentrations of Gull2, E. coli, and enterococci detected in the microcosms over time are shown in Fig. 3 and S6. At 35 days following inoculation, qPCR markers and cultured FIB were reduced by four to five orders of magnitude, and further loss past this time point was minimal. Linear regression analysis was carried out using the first-order model of decay for marker concentrations within the linear range of detection (Table 2). The Gull2 marker decay constant was largest (k = −0.337 day−1) compared to other qPCR targets and culturable indicators measured in the gull microcosm experiment. The Gull2 decay constant was significantly larger than E. coli and enterococci measured by qPCR and enterococci measured by culture (P < 0.01). There was no statistical difference in the decay constants for the Gull2 marker and culture-based E. coli (P = 0.029).

FIG 3
Concentrations of markers and fecal indicators measured over time during the gull microcosm experiment. Circles represent average concentrations for triplicate microcosms. Error bars indicate the standard deviations about the mean concentrations.
TABLE 2
First-order decay coefficients with standard deviations calculated for qPCR markers and fecal indicator bacteria in gull and sewage microcosm experiments

For sewage microcosms, the initial average concentrations of E. coli and enterococci were 2.5 × 103 and 2.1 × 103 CFU/g, respectively, as measured by culture methods. The initial average concentrations of E. coli, enterococci, HB, and Lachno2 were 5.3 × 103 CN/g, 1.5 × 105 CN/g, 4.7 × 104 CN/g, and 6.4 × 104 CN/g, respectively, as measured by qPCR. The concentrations of indicators measured over time are shown in Fig. 4 and S7. All indicators by both qPCR and culture were detected throughout the duration of the experiment, with the exception of E. coli measured by qPCR, which had concentrations below the limit of quantification after 33 days. The HB marker had a decay constant of −0.175 day−1, which was similar to culture-based enterococci and significantly larger than culture-based E. coli and enterococci detected by qPCR (P < 0.01). The Lachno2 marker first-order rate constant was significantly larger than that for enterococci detected by qPCR (P < 0.01) but similar to the value for culture-based E. coli. Most notably, the Lachno2 marker and HB were found to have statistically different decay rate constants, with the HB marker lost at a higher rate than Lachno2; however, concentrations generally remained at an order of magnitude above culturable indicator concentrations throughout the experiment.

FIG 4
Concentrations of markers and fecal indicators measured over time during the sewage microcosm experiment. Circles represent average concentrations for triplicate microcosms. Error bars indicate the standard deviations about the mean concentrations. E. ...

DISCUSSION

Fecal contamination of recreational water poses a threat to beachgoer health, and the resulting beach advisories and/or closures can have serious economic consequences (39). Beach water quality monitoring practices, which typically rely on culture-based enumeration of E. coli and/or enterococci, fall short in their ability to provide beach managers with timely and detailed information concerning sources of fecal pollution that could pose a health risk for beachgoers. This study examined the use of alternative indicators for identifying sources in instances where elevated levels of fecal indicators are detected in water and explored how alternative indicators persist in sand compared to E. coli and enterococci.

Evidence of gull contamination in water was widespread at all beaches, consistent with previous reports in marine regions (31, 40), with the Gull2 marker detected in 83% of samples with high E. coli levels and 82% of samples with low E. coli levels. Traditional FIB concentrations in gull feces have been reported to be highly varied, with concentrations of E. coli and enterococci reported to range between 102 and 108 CFU/g (41) and between 102 and 1010 CFU/g (42,44), respectively. We found similar variability in the gull fecal samples we analyzed, with concentrations of FIB ranging over six orders of magnitude for E. coli and three orders of magnitude for enterococci (Fig. S1). The large range of E. coli and enterococci in gull feces has not been shown to vary by season, geography, or age of gull (42, 45); the omnivorous diet and scavenging tendencies of gull populations may explain much of this variation. The Gull2 marker was much more consistent and, on average, was four orders of magnitude higher than FIB. Next-generation sequencing has revealed that Catellicoccus is the most abundant genus in gull feces, representing, on average, 55% of the total community, which illustrates the utility of Catellicoccus marker assays to detect gull waste (43). The same study showed that Enterococcus spp. and Escherichia spp. represent a smaller fraction of the population, at ~10% (43). Gull feces can also contain some human pathogens, such as Campylobacter and Salmonella (46). Compared to human sewage, the presence of gull fecal pollution poses a comparatively lessened risk of illness, yet can lead to an excessive number of beach closings (10).

In contrast, human sources were rarely detected in beach water but, when detected, were at only two of the urban beaches that have nearby stormwater outfalls. Stormwater outfalls have been reported to be frequently contaminated with sanitary sewage (47, 48), and these discharges can act as local sewage sources in the absence of known contamination events, such as sewage overflows. When human indicators were detected, the Lachno2 marker was detected more frequently. This result could reflect differences in marker survival and/or differences in initial marker concentrations. Newton et al. found that the concentration of HB was significantly correlated with Lachno2 (R2 = 0.86) in water samples collected in the Milwaukee inner harbor (26). In that study and the present study, Lachno2 was found at 1.5-fold-higher levels in sewage than HB. The two human markers, compared to traditional FIB alone, provide greater reliability in the detection of human fecal pollution, a finding that was exemplified in the assessment of post-CSO samples. CSOs occur 1 to 3 times per year in Milwaukee and are a regional rather than local source of fecal pollution to beaches. Previous studies in this system reported that levels of E. coli and enterococci are generally low during and after CSO/sanitary sewer overflow (SSO) events, but human markers have been detected at >105 CN/100 ml in the open waters of Lake Michigan adjacent to the BB and BV beach sites (27, 49). In this study, 1 day following release of a CSO, the HB and Lachno2 markers produced a robust signal in the swimming area, while E. coli was well below the limit for water quality advisories and in some samples, was absent. These data demonstrate that alternative indicators, such as Lachno2 and HB, are very useful to assess serious water quality concerns where dilute pollution could present a serious health risk to beachgoers.

The ruminant marker was only tested for at rural beaches and was detected in 16% and 53% of water samples with elevated E. coli at KA and PB, respectively. Both beaches are near river discharge points (Fig. 5), with PB closer to its river. All of the samples in which the ruminant marker was detected had elevated E. coli levels and were collected on the same date; considering these sites span almost 2 km of shoreline, this demonstrates that there was widespread contamination on this day. When the BacR marker was detected in water, E. coli levels averaged 1,500 CFU/100 ml, and levels of enterococci averaged 1,900 CFU/100 ml, with ratios of alternative indicators to traditional FIB lower than what was found in source material (Table 1), suggesting the pollution could be attenuated in the environment. Alternatively, these ratios could have also been affected by the presence of other sources of additional traditional FIB. Compared to either E. coli or enterococci, the high abundance and low variability of BacR in ruminant fecal samples underscore the reliability of BacR for the detection of fecal pollution from agricultural runoff.

FIG 5
Beach sites: Manitowoc County: Point Beach State Park; Sheboygan County: Kohler-Andrea State Park Beach; Milwaukee County: Doctor's Park Beach, Atwater Beach, Bradford Beach, and Bayview Beach. (Map created using QGIS version 2.10.1-Pisa.)

Sand has been widely considered an intermittent source of fecal indicator bacteria to water (12, 50,52). At all beaches examined, the highest E. coli densities (on a per-weight basis) were found within berm sand samples, while densities of enterococci were found to be higher in backshore sand. Previous work has noted high levels of E. coli in wash-zone beach sand (12, 53); however, few studies have compared multiple beaches concurrently for both indicators, allowing us to benchmark one against the other. These findings suggest that E. coli in beach sand is favored under high-moisture conditions, and enterococci are favored under low-moisture conditions, irrespective of source inputs. Water samples harbored lower concentrations of E. coli and enterococci per 100 ml, compared to 100-g berm or backshore sand samples, consistent with a recent study by Staley et al. (54). Although a determination of bacterial transfer dynamics between sand and water are not within the aims of this study, the high correlation of FIB between berm sand and water suggests that the sand FIB-carrying capacity is large and has the potential to seed FIB to the nearshore water.

The distribution and decay of alternative indicators in sand were examined to assess how alternative indicator persistence compared with traditional indicators that are commonly used in water quality monitoring programs. Despite >80% of all water samples showing evidence of gull waste, only 25% of sand samples from the six beaches were positive for the Gull2 marker, but all had E. coli and enterococci present. The results from water samples support the conclusion that the main external source of E. coli to the beach environment was likely gulls, with the exception of occasional widespread regional contamination.

We examined the time frame that it would take to reduce the Gull2 marker below the levels of E. coli and enterococci in sand. This study is one of the first to deploy in situ microcosms in the beach environment to mimic environmental conditions. The vast majority of previous studies that reported the survival of indicators in beach sand have been performed using controlled benchtop experiments (33, 36,38, 55, 56), which, by design, cannot reproduce the range of interrelated conditions in the natural environment, including daily temperature fluctuations, UV radiation, and humidity cycles. Our microcosm experiments were designed to act as chambers that were subjected to natural temperature variations, allowing for the passage of water and nutrients, while ensuring the microbial integrity of the inoculated sand contained inside. Due to the large variation of FIB in gull fecal samples, microcosm inocula derived from pooled gull droppings were used to mimic average initial concentrations of all the indicators.

Although the Gull2 marker decay constant was greater than culturable E. coli or enterococci, this marker was consistently detected at higher concentrations for approximately 30 days before concentrations dropped below cultured traditional indicator levels. Because the microcosms mimicked beach conditions, this time frame could be a useful benchmark for beach managers when assessing gull sources at beaches. Interestingly, the decay constant for the Gull2 marker was similar to rates obtained for lab benchtop microcosms containing sand and water from Santa Cruz, CA, that utilized seawater (36). The Gull2 assay used in this study is reported to have a high level of sensitivity and specificity during a multilaboratory study of gull-associated assays (29). The only host species shown thus far to cross-react with Gull2 is pigeons. Genetic markers that target Catellicoccus, such as Gull2, appear to be very robust for source detection and could be useful for evaluating the success of gull deterrent programs that are aimed at reducing this nuisance bird loafing on beaches, which can result in accumulation of FIB in the sand.

Sand microcosms with sewage as a fecal source demonstrated that the human-associated markers HB and Lachno2 also had a higher decay rate constant than culturable E. coli. Similar to the gull microcosms, we used untreated sewage as the inoculum; thus, the relative proportions of each indicator at the start of the experiment were similar to an actual contamination event. The human-associated markers were detected at levels above culturable E. coli and enterococci in sand for ~50 days, suggesting that these markers could give indications of sewage impacts to beaches over this time frame. Since humans are reservoirs for many human pathogens, human sources create a serious health risk to beachgoers, and an assessment of residual contamination in sand might be useful since inputs may be sporadic and rain driven, making it difficult to detect these sources in water.

Lachno2 was detected but not quantified in a high number (69/110) of sand samples. The HB marker decay constant was larger than that of the Lachno2 marker. Differences in the human marker decay patterns suggest that old pollution may result in the sole detection of Lachno2, given similar initial concentrations of the two markers; however, there are several other alternative explanations for this result. Samples with trace amounts of DNA template, due to environmental dilution and/or attenuation, can lead to DNQ results. Specifically for microbial source tracking studies, DNQ results may be the result of the presence of old fecal pollution. Additionally, amplification below the limit of quantification may indicate low-level cross-reactivity with nontarget organisms, such as those indigenous to the natural microbial community. We observed a greater-than-expected number of DNQ samples in sand, but very few were observed in water samples for Lachno2. Environmental interference has been reported with alternative indicator assays, targeting Bacteroides (57, 58), which highlights that previously uncharacterized organisms could interfere with assays targeting fecal bacteria. Low levels of amplification, such as with the DNQ samples, can occur when there is a large amount of similar but nontarget sequence (59). Large numbers of DNQ results were reported in a recent study of 41 microbial source tracking markers, where the authors suggested that detection thresholds were very important in determining if a source is present or absent (8). Cross-reactivity with canine feces could be another possible explanation, as Lachno2 has been detected in some canine fecal samples (60); however, testing with an established canine marker (61) produced negative results (data not shown). Further validation of this marker in the sand matrix is needed.

Understanding the dynamics of both traditional and alternative fecal indicators in the beach environment is essential for effectively identifying fecal pollution sources and evaluating potential health risks. This is especially important as beach managers move toward implementing molecular testing methodologies. From this study and others, it is clear that it is virtually impossible to interpret single-day or even extensive multiday survey data to identify sources of E. coli observed based on the presence of alternative indicators. Contamination scenarios are complex and involve repeated deposition, differential survival of indicators, and interchange between sand, water, and other matrices, such as wrack. Despite this, the high abundance and consistency of alternative indicators in source fecal material demonstrate that beach managers could reliably employ alternative indicators to detect specific suspected sources from recent pollution events, such as human sources from stormwater or sewage overflows, or impacts from nearby agricultural watersheds. The lack of correlation between elevated E. coli levels and identification of sources with alternative indicators in both water and sand suggests that elevated E. coli levels should not be the only criterion for choosing samples for testing by qPCR methods.

Growing evidence, including the results put forth in this study, substantiates that persistence of fecal indicators in beach sand is a major confounder of monitoring programs. The absence of alternative indicators of the most probable sources for a beach may be a result of differential decay and could be considered evidence that pollution is from a past rather than recent pollution event. Our microcosm studies suggest that source-associated indicators will be at higher concentrations than the culturable FIB associated with that source when inputs occurred >30 days prior for the Gull2 marker and >50 days prior for sewage markers. Alternatively, E. coli and enterococci that occur with no other evidence of fecal pollution could represent strains that are naturalized, a phenomenon worth further exploration (62). Overall, with proper interpretation of monitoring results, the use of alternative indicators can improve the breadth of beach pollution assessments and aid in source identification at recreational beaches.

MATERIALS AND METHODS

Study area and sample collection.

This study was conducted during the summer months of 2012 to 2013 along the western coastline of Lake Michigan at six Wisconsin beaches. Beaches included Point Beach State Park (PB), Kohler-Andrae State Park (KA), Doctor's Park (DP), Atwater Park Beach (ATW), Bradford Beach (BB), and Bayview Beach (BV) (Fig. 5). The beach site map was created using QGIS version 2.10.1 (63). At each of the six beaches, in addition to water samples, sand was collected at the berm (wash zone wetted by wave action) and backshore beach zones (dry sand). Backshore sand was only collected during 2013 sampling dates. Sampling was conducted along three sites, spaced 50 to 100 m apart, parallel to the shoreline for PB, KA, BB, and BV, while ATW and DP were sampled at four sites. At each beach, the berm zone was considered the lakeward portion of the beach within the range of wet sand subjected to wave action. The backshore was defined as the generally dry portion of the beach between the vegetation line and the berm, which is wetted from intermittent wave action with dry periods over a span of days to weeks. Nearshore water was collected at a depth of 0.3 m approximately 3 to 5 m from shore. Water was collected in sterile 1-liter Nalgene bottles, and all sand samples were collected in sterile Whirl-Pak bags. Sand and water samples were transported to the lab on ice and processed within 24 h of collection. Sand moisture content was determined based on the mass difference before and after a 24-h drying period at 45°C.

Culture-based fecal indicator enumeration.

E. coli and enterococci were isolated from sand samples using techniques adapted from those developed by Boehm et al. (50). To isolate cells from sand, 45 g of either backshore or berm sand was shaken in 450 ml of sterile water for 2 min by hand. Water and sand extracts were filtered through a 0.45-μm-pore-size nitrocellulose filter (Millipore, Billerica, MA) and transferred to modified membrane-thermotolerant E. coli (mTEC) and membrane Enterococcus indoxyl-d-glucoside (mEI) agar plates (Becton Dickinson, Franklin Lakes, NJ). Incubation and enumeration were performed according to USEPA methods 1603 and 1600 for E. coli and enterococci, respectively (64, 65). FIB concentrations were reported as CFU/100 ml or CFU/100 g (dry weight). Various filter volumes or sample dilutions were used to attain colony counts within a target countable range of 10 to 300 CFU. For samples filtered at 100 ml having colony counts fewer than 10, whole numbers were reported; otherwise, concentrations were reported to two significant figures, per Myers et al. (66).

Sample processing and selection.

Water samples (400 ml) and sand eluent (100 ml) were also filtered through 0.22-μm-pore-size nitrocellulose filters (Millipore, Billerica, MA) for DNA extraction procedures. Filters were folded using ethanol-sterilized forceps and transferred to 2-ml sterile screw-cap tubes. Both water and sand DNA filters were stored at −80°C until DNA extraction was performed. Water samples were chosen for qPCR analysis if the culture counts exceeded 235 CFU/100 ml. A total of 84/422 samples were found to exceed the E. coli water criteria. We also selected a subset of water (n = 60) samples that had low levels of E. coli, which were collected at the same beach on the same day or within 1 day of the exceedance water samples. Paired sand samples, collected at the same transect and day as the selected water samples, were also analyzed by qPCR. Of the 84 water samples with ≥235 CFU E. coli/100 ml, 69 paired berm sand samples were analyzed by qPCR. Of the 60 water samples with ≤235 CFU E. coli /100 ml, 41 paired berm sand samples were analyzed. Sand was not collected on some of the sample dates. Additional water samples from Bradford (n = 6), Atwater (n = 8), and Bayview Beaches (n = 6), collected 1 day and 3 days after a combined sewer overflow (CSO) event in the Milwaukee estuary, were also analyzed using the same qPCR assays. We did not collect sand samples under CSO conditions. A summary of beach locations, number of days sampled, total samples collected, and samples analyzed using qPCR by beach can found in Table S1.

DNA extraction and qPCR analysis.

To extract DNA from the selected samples, frozen nitrocellulose filters were removed from −80°C storage and placed on dry ice. Filters were then manually crushed into small pieces using sterile steel spatulas. DNA was extracted from the crushed filter pieces using the Fast DNA Spin kit for soil (MP Biomedicals, Solon, OH), according to the manufacturer's instructions. The concentration and purity of DNA were determined using a NanoDrop ND-1000 (Thermo Fisher Scientific, Inc., Pittsburgh, PA). Extraction efficiencies have consistently been ~20%, as previously reported (67).

The qPCR assays were carried out using an ABI StepOne Plus real-time PCR system with TaqMan (Applied Biosystems, Foster City, CA) hydrolysis probe chemistry. All qPCR assays included in this study were previously published, and the primer-probe sequences can be found in Table S2. The assays employed in this study targeted E. coli (55), Enterococcus spp. (Entero) (68), human-associated Lachnospiraceae (Lachno2) (26), human-associated Bacteroides (HB) (19, 67, 69), gull-associated Catellicoccus marimammalium (Gull2) (28, 70), and ruminant-associated Bacteroidetes (BacR) (22). Standard curves were created using six serial 1:10 dilutions from 1.5 × 106 to 15 copies per reaction of a linearized plasmid containing the target sequence. Standard curves were run in triplicate and were included on each run. The slope, y-intercept, and assay efficiencies can be found in Table S3. All samples were run in duplicate 25-μl reaction mixtures containing 1× TaqMan gene expression master mix (Applied Biosystems), primers and probes at final concentrations of 1 μM and 80 nM, respectively, and 25 to 100 ng of extracted DNA. The PCR cycling conditions were performed as follows: 2 min at 50°C to activate the uracil-N-glycosylase (UNG), 10 min at 95°C to inactivate UNG and activate the Taq polymerase, 40 cycles of 95°C for 15 s, and 60°C for 1 min.

The copy number (CN) was converted to CN/100 ml of water or CN/100 g of sand (original sample) using the DNA elution volume (150 μl), sample filtration volume, total sample mass (for sand), and wet/dry mass (for sand). The lower limit of quantification (LLQ) was determined for each assay using a cycle threshold (CT) corresponding to the standard curve dilution that was within the linear range. The limit of reliable quantification or LLQ was 15 copies per reaction, which is equivalent to 112 CN/100 ml for water samples filtered at 400 ml. The LLQ occurred at a CT of 35, with the exception of the Gull2 assay, which had a CT of 38. Any amplification that occurred below the LLQ but above background was recorded as detected but not quantifiable (DNQ). All beach water samples were analyzed using the E. coli, Entero, Gull2, Lachno2, and HB assays. Beach sand samples were analyzed using the Gull2, Lachno2, and HB assays. Selected water and sand samples that were collected from either Point Beach State Park or Kohler-Andrae State Park were also analyzed using the ruminant-specific BacR assay due to the proximal agricultural land use practices (14).

Gull, ruminant, and untreated sewage sample analysis.

To examine the variation of host-associated genetic markers (Gull2, Lachno2, HB, and BacR) and their relationship to traditional indicators in source material, we performed qPCR on gull fecal pellets (n = 22), sewage influent samples (n = 43), and bovine fecal samples (n = 6). Fecal and sewage influent samples were processed using the Fast DNA Spin kit for soil (MP Biomedicals, Solon, OH), according to the manufacturer's protocol. Additional information about the collection of these samples can be found in the supplemental material. We also analyzed these samples using the E. coli and Entero qPCR assays and compared the relative concentrations of host markers to those of E. coli and Entero.

In situ beach microcosms.

Microcosms were constructed using polyvinyl chloride (PVC) pipe material cut into 5 by 9-cm pieces. End caps for the microcosms were prepared by drilling 20 to 30 1-mm-diameter holes into PVC knockout test caps (Oatey, Cleveland, OH) and affixing a sterile 0.22-μm-pore filter to the interior surface using standard silicone sealant. The microcosm design was adapted from that of Alm et al. (71). Prior to microcosm use, the interior surfaces of the PVC pipe were sterilized using a 70% ethanol wash.

For gull fecal microcosm experiments, 13 fecal pellets were collected from metropolitan Milwaukee locations. A 1-ml aliquot of sterile phosphate-buffered saline (PBS) was added to each pellet and vortexed for 30 s to create a pooled fecal slurry. All fecal slurries were combined for a single pooled sample and used within 24 h of collection. Initial concentrations of E. coli and enterococci were determined for the pooled sample via membrane filtration and incubation on mod-mTEC and mEI media, as described previously. Beach sand was collected at the berm from Bradford Beach and tested for levels of E. coli and enterococci. A 5-ml subsample of the fecal slurry was inoculated into 5,000 g of sand and homogenized manually. Inoculated sand was divided into 21 prepared microcosms, which were used for seven triplicate time points. The remaining inoculated sand was used for the triplicate measurement at time 0. With permission from Milwaukee County Parks, inoculated microcosms were transferred to Bayview Beach on ice and buried within the berm zone in a single layer 7 to 10 cm below surface level, spaced 5 cm apart. After burial, a random selection of three microcosms were removed every 7 to 10 days and transferred to the lab for analysis. All samples were analyzed using previously described culture-based methods. Sand extracts were also filtered for DNA, extracted, and analyzed using the E. coli, Entero, and Gull2 qPCR assays. The experiment was run for 57 days.

Sewage microcosms used the same setup protocol as outlined above. The inoculum for the sewage microcosm consisted of untreated sewage influent collected at the South Shore Water Reclamation facility supplied by Milwaukee Metropolitan Sewerage District (MMSD). Sewage influent was transferred to the laboratory at 4°C and used within 4 h of collection. A 100-ml aliquot of sewage was inoculated into 3,000 g of sand, homogenized manually, and divided into 15 microcosms. Time points were 7 to 10 days apart, with the final triplicate microcosms sacrificed at 51 days after inoculation, for a total of six time points. Sand extracts were processed as described in the previous experiment and analyzed using the E. coli, Entero, HB, and Lachno2 qPCR assays.

Data analysis.

Data analysis and statistical procedures were performed in R (version 3.1.1) using R Core packages. Data visualization and figure generation were carried out with the lattice and ggplot2 packages and package dependencies (72). Culture counts were log transformed (those with no detectable CFU were given a value of 1). Counts of E. coli and enterococci were compared using the 2-tailed Student's t test, assuming equal variance and significance at a P value of ≤0.01. The geometric mean was determined independently for each beach in backshore sand, berm sand, and water samples. Spearman's rank order coefficient was used to assess the correlation between E. coli levels in water and berm sand samples along the same transect, and to assess the correlation between human-associated marker concentrations in samples collected under CSO conditions. The geometric mean values were used as the input for the heatmap generation. Ratios of alternative indicators to E. coli and enterococci were calculated for each sample individually. The average and standard deviation of sample ratios were reported.

The arithmetic mean and standard deviation were calculated for each replicate time point in both microcosm experiments. Linear regressions were propagated using the first-order exponential decay equation ln(C/C0) = kt, where C0 is the initial concentration, k is the decay rate constant in days−1, and t is equal to elapsed time in days. Analysis of variance (ANOVA) was used to evaluate if significant differences exist between the regression coefficients for indicator decay in microcosm experiments. Differences in decay rate coefficients were deemed significant at a P value of ≤0.01. In discussing differences in decay constants among fecal indicators and qPCR markers, the absolute value of the decay constant was used.

Supplementary Material

Supplemental material:

ACKNOWLEDGMENTS

We thank Katherine Halmo, Morgan Schroeder, Melinda Bootsma, Deb Dila, and Hayley Templar for their assistance with sample collection and processing. We also thank Patrick Siwula and Natalie Rumball for their assistance with manuscript editing. The Wisconsin Department of Natural Resources gave permission for our sampling at Point Beach State Park and Kohler-Andrae State Park. We also thank the Milwaukee County Parks Department for allowing our experimentation.

This study was funded by the University of Wisconsin Sea Grant Institute, under a grant from the National Sea Grant College Program, National Oceanic and Atmospheric Administration, U.S. Department of Commerce, and the State of Wisconsin, federal grant NA10OAR4170070, project R/HCE-10.

Footnotes

Supplemental material for this article may be found at https://doi.org/10.1128/AEM.02881-16.

REFERENCES

1. USEPA. 1986. Ambient water quality criteria for bacteria–1986. Office of Water Regulation and Standards, United States Environmental Protection Agency, Washington, DC: http://www.ugra.org/pdfs/EPA1986.pdf.
2. USEPA. 2012. Recreational water quality criteria. Office of Water, United States Environmental Protection Agency, Washington, DC.
3. Dufour AP. 1984. Health effects criteria for fresh recreational waters. EPA-600/1-84-004. Office of Research and Development, United States Environmental Protection Agency, Washington, DC.
4. Field KG, Samadpour M 2007. Fecal source tracking, the indicator paradigm, and managing water quality. Water Res 41:3517–3538. doi:.10.1016/j.watres.2007.06.056 [PubMed] [Cross Ref]
5. Wilkes G, Edge T, Gannon V, Jokinen C, Lyautey E, Medeiros D, Neumann N, Ruecker N, Topp E, Lapen DR 2009. Seasonal relationships among indicator bacteria, pathogenic bacteria, Cryptosporidium oocysts, Giardia cysts, and hydrological indices for surface waters within an agricultural landscape. Water Res 43:2209–2223. doi:.10.1016/j.watres.2009.01.033 [PubMed] [Cross Ref]
6. Harwood VJ, Levine AD, Scott TM, Chivukula V, Lukasik J, Farrah SR, Rose JB 2005. Validity of the indicator organism paradigm for pathogen reduction in reclaimed water and public health protection. Appl Environ Microbiol 71:3163–3170. doi:.10.1128/AEM.71.6.3163-3170.2005 [PMC free article] [PubMed] [Cross Ref]
7. Ervin JS, Russell TL, Layton BA, Yamahara KM, Wang D, Sassoubre LM, Cao Y, Kelty CA, Sivaganesan M, Boehm AB, Holden PA, Weisberg SB, Shanks OC 2013. Characterization of fecal concentrations in human and other animal sources by physical, culture-based, and quantitative real-time PCR methods. Water Res 47:6873–6882. doi:.10.1016/j.watres.2013.02.060 [PubMed] [Cross Ref]
8. Boehm AB, Van De Werfhorst LC, Griffith JF, Holden PA, Jay JA, Shanks OC, Wang D, Weisberg SB 2013. Performance of forty-one microbial source tracking methods: a twenty-seven lab evaluation study. Water Res 47:6812–6828. doi:.10.1016/j.watres.2012.12.046 [PubMed] [Cross Ref]
9. Kelty CA, Varma M, Sivaganesan M, Haugland RA, Shanks OC 2012. Distribution of genetic marker concentrations for fecal indicator bacteria in sewage and animal feces. Appl Environ Microbiol 78:4225–4232. doi:.10.1128/AEM.07819-11 [PMC free article] [PubMed] [Cross Ref]
10. Soller JA, Schoen ME, Varghese A, Ichida AM, Boehm AB, Eftim S, Ashbolt NJ, Ravenscroft JE 2014. Human health risk implications of multiple sources of faecal indicator bacteria in a recreational waterbody. Water Res 66:254–264. doi:.10.1016/j.watres.2014.08.026 [PubMed] [Cross Ref]
11. Yamahara KM, Walters SP, Boehm AB 2009. Growth of enterococci in unaltered, unseeded beach sands subjected to tidal wetting. Appl Environ Microbiol 75:1517–1524. doi:.10.1128/AEM.02278-08 [PMC free article] [PubMed] [Cross Ref]
12. Alm E, Burke J, Spain A 2003. Fecal indicator bacteria are abundant in wet sand at freshwater beaches. Water Res 37:3978–3982. doi:.10.1016/S0043-1354(03)00301-4 [PubMed] [Cross Ref]
13. Phillips MC, Solo-Gabriele HM, Reniers AJHM, Wang JD, Kiger RT, Abdel-Mottaleb N 2011. Pore water transport of enterococci out of beach sediments. Mar Pollut Bull 62:2293–2298. doi:.10.1016/j.marpolbul.2011.08.049 [PMC free article] [PubMed] [Cross Ref]
14. Cloutier DD, Alm EW, McLellan SL 2015. The influence of land-use, nutrients, and geography on microbial communities and fecal indicator abundance at Lake Michigan beaches. Appl Environ Microbiol 81:4904–4913. doi:.10.1128/AEM.00233-15 [PMC free article] [PubMed] [Cross Ref]
15. Savichtcheva O, Okabe S 2006. Alternative indicators of fecal pollution: relations with pathogens and conventional indicators, current methodologies for direct pathogen monitoring and future application perspectives. Water Res 40:2463–2476. doi:.10.1016/j.watres.2006.04.040 [PubMed] [Cross Ref]
16. Kreader CA. 1998. Persistence of PCR-detectable Bacteroides distasonis from human feces in river water. Appl Environ Microbiol 64:4103–4105. [PMC free article] [PubMed]
17. Fiksdal L, Maki JS, LaCroix SJ, Staley JT 1985. Survival and detection of Bacteroides spp., prospective indicator bacteria. Appl Environ Microbiol 49:148–150. [PMC free article] [PubMed]
18. Bernhard AE, Field KG 2000. Identification of nonpoint sources of fecal pollution in coastal waters by using host-specific 16S ribosomal DNA genetic markers from fecal anaerobes. Appl Environ Microbiol 66:1587–1594. doi:.10.1128/AEM.66.4.1587-1594.2000 [PMC free article] [PubMed] [Cross Ref]
19. Kildare BJ, Leutenegger CM, McSwain BS, Bambic DG, Rajal VB, Wuertz S 2007. 16S rRNA-based assays for quantitative detection of universal, human-, cow-, and dog-specific fecal Bacteroidales: a Bayesian approach. Water Res 41:3701–3715. doi:.10.1016/j.watres.2007.06.037 [PubMed] [Cross Ref]
20. Kreader CA. 1995. Design and evaluation of Bacteroides DNA probes for the specific detection of human fecal pollution. Appl Environ Microbiol 61:1171–1179. [PMC free article] [PubMed]
21. Shanks OC, Kelty CA, Sivaganesan M, Varma M, Haugland RA 2009. Quantitative PCR for genetic markers of human fecal pollution. Appl Environ Microbiol 75:5507–5513. doi:.10.1128/AEM.00305-09 [PMC free article] [PubMed] [Cross Ref]
22. Reischer GH, Kasper DC, Steinborn R, Mach RL, Farnleitner AH 2006. Quantitative PCR method for sensitive detection of ruminant fecal pollution in freshwater and evaluation of this method in alpine karstic regions. Appl Environ Microbiol 72:5610–5614. doi:.10.1128/AEM.00364-06 [PMC free article] [PubMed] [Cross Ref]
23. Fremaux B, Boa T, Yost CK 2010. Quantitative real-time PCR assays for sensitive detection of Canada goose-specific fecal pollution in water sources. Appl Environ Microbiol 76:4886–4889. doi:.10.1128/AEM.00110-10 [PMC free article] [PubMed] [Cross Ref]
24. Kobayashi A, Sano D, Hatori J, Ishii S, Okabe S 2013. Chicken- and duck-associated Bacteroides-Prevotella genetic markers for detecting fecal contamination in environmental water. Appl Microbiol Biotechnol 97:7427–7437. doi:.10.1007/s00253-012-4469-2 [PubMed] [Cross Ref]
25. McLellan SL, Newton RJ, Vandewalle JL, Shanks OC, Huse SM, Eren MA, Sogin ML 2013. Sewage reflects the distribution of human faecal Lachnospiraceae. Environ Microbiol 15:2213–2227. doi:.10.1111/1462-2920.12092 [PMC free article] [PubMed] [Cross Ref]
26. Newton RJ, Vandewalle JL, Borchardt MA, Gorelick MH, McLellan SL 2011. Lachnospiraceae and Bacteroidales alternative fecal indicators reveal chronic human sewage contamination in an urban harbor. Appl Environ Microbiol 77:6972–6981. doi:.10.1128/AEM.05480-11 [PMC free article] [PubMed] [Cross Ref]
27. Newton RJ, Bootsma MJ, Morrison HG, Sogin ML, McLellan SL 2013. A microbial signature approach to identify fecal pollution in the waters off an urbanized coast of Lake Michigan. Microb Ecol 65:1011–1023. doi:.10.1007/s00248-013-0200-9 [PMC free article] [PubMed] [Cross Ref]
28. Lu J, Santo Domingo JW, Lamendella R, Edge T, Hill S 2008. Phylogenetic diversity and molecular detection of bacteria in gull feces. Appl Environ Microbiol 74:3969–3976. doi:.10.1128/AEM.00019-08 [PMC free article] [PubMed] [Cross Ref]
29. Sinigalliano CD, Ervin JS, Van De Werfhorst LC, Badgley BD, Ballesté E, Bartkowiak J, Boehm AB, Byappanahalli M, Goodwin KD, Gourmelon M, Griffith J, Holden PA, Jay J, Layton B, Lee C, Lee J, Meijer WG, Noble R, Raith M, Ryu H, Sadowsky MJ, Schriewer A, Wang D, Wanless D, Whitman R, Wuertz S, Santo Domingo JW 2013. Multi-laboratory evaluations of the performance of Catellicoccus marimammalium PCR assays developed to target gull fecal sources. Water Res 47:6883–6896. doi:.10.1016/j.watres.2013.02.059 [PubMed] [Cross Ref]
30. Ryu H, Griffith JF, Khan IUH, Hill S, Edge TA, Toledo-Hernandez C, Gonzalez-Nieves J, Santo Domingo J 2012. Comparison of gull feces-specific assays targeting the 16S rRNA genes of Catellicoccus marimammalium and Streptococcus spp. Appl Environ Microbiol 78:1909–1916. doi:.10.1128/AEM.07192-11 [PMC free article] [PubMed] [Cross Ref]
31. Araújo S, Henriques IS, Leandro SM, Alves A, Pereira A, Correia A 2014. Gulls identified as major source of fecal pollution in coastal waters: A microbial source tracking study. Sci Total Environ 470–471:84–91. doi:.10.1016/j.scitotenv.2013.09.075 [PubMed] [Cross Ref]
32. Whitman R, Harwood VJ, Edge TA, Nevers M, Byappanahalli M, Vijayavel K, Brandão J, Sadowsky MJ, Alm EW, Crowe A, Ferguson D, Ge Z, Halliday E, Kinzelman J, Kleinheinz G, Przybyla-Kelly K, Staley C, Staley Z, Solo-Gabriele HM 2014. Microbes in beach sands: integrating environment, ecology and public health. Rev Environ Sci Biotechnol 13:329–368. doi:.10.1007/s11157-014-9340-8 [PMC free article] [PubMed] [Cross Ref]
33. Yamahara KM, Sassoubre LM, Goodwin KD, Boehm AB 2012. Occurrence and persistence of bacterial pathogens and indicator organisms in beach sand along the California coast. Appl Environ Microbiol 78:1733–1745. doi:.10.1128/AEM.06185-11 [PMC free article] [PubMed] [Cross Ref]
34. Heaney CD, Sams E, Wing S, Marshall S, Brenner K, Dufour AP, Wade TJ 2009. Contact with beach sand among beachgoers and risk of illness. Am J Epidemiol 170:164–172. doi:.10.1093/aje/kwp152 [PubMed] [Cross Ref]
35. Solo-Gabriele HM, Harwood VJ, Kay D, Fujioka RS, Sadowsky MJ, Whitman RL, Wither A, Caniça M, Carvalho da Fonseca R, Duarte A, Edge TA, Gargaté MJ, Gunde-Cimerman N, Hagen F, McLellan SL, Nogueira da Silva A, Novak Babič M, Prada S, Rodrigues R, Romão D, Sabino R, Samson RA, Segal E, Staley C, Taylor HD, Veríssimo C, Viegas C, Barroso H, Brandão JC 2015. Beach sand and the potential for infectious disease transmission: observations and recommendations. J Mar Biol Assoc U K 96:1–20.
36. Brown KI, Boehm AB 2015. Comparative decay of Catellicoccus marimmalium and enterococci in beach sand and seawater. Water Res 83:377–384. doi:.10.1016/j.watres.2015.06.055 [PubMed] [Cross Ref]
37. Eichmiller JJ, Borchert AJ, Sadowsky MJ, Hicks RE 2014. Decay of genetic markers for fecal bacterial indicators and pathogens in sand from Lake Superior. Water Res 59:99–111. doi:.10.1016/j.watres.2014.04.005 [PubMed] [Cross Ref]
38. Staley ZR, Robinson C, Edge TA 2016. Comparison of the occurrence and survival of fecal indicator bacteria in recreational sand between urban beach, playground and sandbox settings in Toronto, Ontario. Sci Total Environ 541:520–527. doi:.10.1016/j.scitotenv.2015.09.088 [PubMed] [Cross Ref]
39. Rabinovici SJM, Bernknopf RL, Wein AM, Coursey DL, Whitman RL 2004. Economic and health risk trade-offs of swim closures at a Lake Michigan beach. Environ Sci Technol 38:2737–2745. doi:.10.1021/es034905z [PubMed] [Cross Ref]
40. Converse RR, Kinzelman JL, Sams EA, Hudgens E, Dufour AP, Ryu H, Santo-Domingo JW, Kelty CA, Shanks OC, Siefring SD, Haugland RA, Wade TJ 2012. Dramatic improvements in beach water quality following gull removal. Environ Sci Technol 46:10206–10213. [PubMed]
41. Meerburg BG, Koene MGJ, Kleijn D 2011. Escherichia coli concentrations in feces of geese, coots, and gulls residing on recreational water in The Netherlands. Vector Borne Zoonotic Dis 11:601–603. doi:.10.1089/vbz.2010.0218 [PubMed] [Cross Ref]
42. Fogarty LR, Haack SK, Wolcott MJ, Whitman RL 2003. Abundance and characteristics of the recreational water quality indicator bacteria Escherichia coli and enterococci in gull faeces. J Appl Microbiol 94:865–878. doi:.10.1046/j.1365-2672.2003.01910.x [PubMed] [Cross Ref]
43. Koskey AM, Fisher JC, Traudt MF, Newton RJ, McLellan SL 2014. Analysis of the gull fecal microbial community reveals the dominance of Catellicoccus marimammalium in relation to culturable enterococci. Appl Environ Microbiol 80:757–765. doi:.10.1128/AEM.02414-13 [PMC free article] [PubMed] [Cross Ref]
44. Wright ME, Solo-Gabriele HM, Elmir S, Fleming LE 2009. Microbial load from animal feces at a recreational beach. Mar Pollut Bull 58:1649–1656. doi:.10.1016/j.marpolbul.2009.07.003 [PMC free article] [PubMed] [Cross Ref]
45. Lévesque B, Brousseau P, Bernier F, Dewailly É, Joly J 2000. Study of the bacterial content of ring-billed gull droppings in relation to recreational water quality. Water Res 34:1089–1096. doi:.10.1016/S0043-1354(99)00266-3 [Cross Ref]
46. Kinzelman J, McLellan SL, Amick A, Preedit J, Scopel CO, Olapade O, Gradus S, Singh A, Sedmak G 2008. Identification of human enteric pathogens in gull feces at Southwestern Lake Michigan bathing beaches. Can J Microbiol 54:1006–1015. doi:.10.1139/W08-096 [PubMed] [Cross Ref]
47. Sauer EP, Vandewalle JL, Bootsma MJ, McLellan SL 2011. Detection of the human specific Bacteroides genetic marker provides evidence of widespread sewage contamination of stormwater in the urban environment. Water Res 45:4081–4091. doi:.10.1016/j.watres.2011.04.049 [PubMed] [Cross Ref]
48. Sercu B, Van De Werfhorst LC, Murray JLS, Holden PA 2011. Sewage exfiltration as a source of storm drain contamination during dry weather in urban watersheds. Environ Sci Technol 45:7151–7157. doi:.10.1021/es200981k [PubMed] [Cross Ref]
49. McLellan SL. 2004. Genetic diversity of Escherichia coli isolated from urban rivers and beach water. Appl Environ Microbiol 70:4658. doi:.10.1128/AEM.70.8.4658-4665.2004 [PMC free article] [PubMed] [Cross Ref]
50. Boehm AB, Griffith J, McGee C, Edge TA, Solo-Gabriele HM, Whitman R, Cao Y, Getrich M, Jay JA, Ferguson D, Goodwin KD, Lee CM, Madison M, Weisberg SB 2009. Faecal indicator bacteria enumeration in beach sand: a comparison study of extraction methods in medium to coarse sands. J Appl Microbiol 107:1740–1750. doi:.10.1111/j.1365-2672.2009.04440.x [PMC free article] [PubMed] [Cross Ref]
51. Phillips MC, Feng Z, Vogel LJ, Reniers AJHM, Haus BK, Enns AA, Zhang Y, Hernandez DB, Solo-Gabriele HM 2014. Microbial release from seeded beach sediments during wave conditions. Mar Pollut Bull 79:114–122. doi:.10.1016/j.marpolbul.2013.12.029 [PMC free article] [PubMed] [Cross Ref]
52. Bonilla TD, Nowosielski K, Cuvelier M, Hartz A, Green M, Esiobu N, McCorquodale DS, Fleisher JM, Rogerson A 2007. Prevalence and distribution of fecal indicator organisms in South Florida beach sand and preliminary assessment of health effects associated with beach sand exposure. Mar Pollut Bull 54:1472–1482. doi:.10.1016/j.marpolbul.2007.04.016 [PubMed] [Cross Ref]
53. Whitman RL, Nevers MB 2003. Foreshore sand as a source of Escherichia coli in nearshore water of a Lake Michigan beach. Appl Environ Microbiol 69:5555–5562. doi:.10.1128/AEM.69.9.5555-5562.2003 [PMC free article] [PubMed] [Cross Ref]
54. Staley ZR, Vogel L, Robinson C, Edge TA 2015. Differential occurrence of Escherichia coli and human Bacteroidales at two Great Lakes beaches. J Great Lakes Res 41:530–535. doi:.10.1016/j.jglr.2015.03.006 [Cross Ref]
55. Lee CM, Lin TY, Lin C-C, Kohbodi GA, Bhatt A, Lee R, Jay JA 2006. Persistence of fecal indicator bacteria in Santa Monica Bay beach sediments. Water Res 40:2593–2602. doi:.10.1016/j.watres.2006.04.032 [PubMed] [Cross Ref]
56. Feng F, Goto D, Yan T 2010. Effects of autochthonous microbial community on the die-off of fecal indicators in tropical beach sand. FEMS Microbiol Ecol 74:214–225. doi:.10.1111/j.1574-6941.2010.00921.x [PubMed] [Cross Ref]
57. Whitman RL, Byappanahalli MN, Spoljaric AM, Przybyla-Kelly K, Shively D, Nevers M 2014. Evidence for free-living Bacteroides in Cladophora along the shores of the Great Lakes. Aquatic Microb Ecol 72:117–126. doi:.10.3354/ame01688 [Cross Ref]
58. Olapade OA, Depas MM, Jensen ET, McLellan SL 2006. Microbial communities and fecal indicator bacteria associated with Cladophora mats on beach sites along Lake Michigan shores. Appl Environ Microbiol 72:1932–1938. doi:.10.1128/AEM.72.3.1932-1938.2006 [PMC free article] [PubMed] [Cross Ref]
59. Layton BA, Cao Y, Ebentier DL, Hanley K, Ballest E, Brandao J, Byappanahalli M, Converse R, Farnleitner AH, Gentry-Shields J, Gidley ML, Gourmelon M, Lee CS, Lee J, Lozach S, Madi T, Meijer WG, Noble R, Peed L, Reischer GH, Rodrigues R, Rose JB, Schriewer A, Sinigalliano C, Srinivasan S, Stewart J, Van De Werfhorst LC, Wang D, Whitman R, Wuertz S, Jay J, Holden PA, Boehm AB, Shanks O, Griffith JF 2013. Performance of human fecal anaerobe-associated PCR-based assays in a multi-laboratory method evaluation study. Water Res 47:6897–6908. doi:.10.1016/j.watres.2013.05.060 [PubMed] [Cross Ref]
60. Fisher JC, Murat Eren A, Green HC, Shanks OC, Morrison HG, Vineis JH, Sogin ML, McLellan SL 2015. Comparison of sewage and animal fecal microbiomes by using oligotyping reveals potential human fecal indicators in multiple taxonomic groups. Appl Environ Microbiol 81:7023–7033. doi:.10.1128/AEM.01524-15 [PMC free article] [PubMed] [Cross Ref]
61. Dick LK, Simonich MT, Field KG 2005. Microplate subtractive hybridization to enrich for Bacteroidales genetic markers for fecal source identification. Appl Environ Microbiol 71:3179–3183. doi:.10.1128/AEM.71.6.3179-3183.2005 [PMC free article] [PubMed] [Cross Ref]
62. Ishii S, Ksoll WB, Hicks RE, Sadowsky MJ 2006. Presence and growth of naturalized Escherichia coli in temperate soils from Lake Superior watersheds. Appl Environ Microbiol 72:612–621. doi:.10.1128/AEM.72.1.612-621.2006 [PMC free article] [PubMed] [Cross Ref]
63. QGIS Development Team. 2015. Quantum Geographic Information System. 2.10.1-Pisa Open Source Geospatial Foundation Project.
64. USEPA. 2002. Method 1603: Escherichia coli (E. coli) in water by membrane filtration using modified membrane-thermotolerant Escherichia coli agar (modified mTEC). EPA-821-R-02-023. United States Environmental Protection Agency, Washington, DC.
65. USEPA. 2006. Method 1600: enterococci in water by membrane filtration using membrane-Enterococcus indoxyl-B-d-glucoside agar (mEI). EPA-821-R-06-009. United States Environmental Protection Agency, Washington, DC.
66. Myers DM, Stoeckel DM, Bushon RN, Francy DS, Brady AM 1997. Fecal indicator bacteria, p 1–38. In National field manual for the collection of water-quality data: U.S. Geological Survey techniques of water-resources investigations, book 9. U.S. Geological Survey, Reston, VA.
67. Templar HA, Dila DK, Bootsma MJ, Corsi SR, McLellan SL 2016. Quantification of human-associated fecal indicators reveal sewage from urban watersheds as a source of pollution to Lake Michigan. Water Res 100:556–567. doi:.10.1016/j.watres.2016.05.056 [PubMed] [Cross Ref]
68. Ludwig W, Schleifer KH 2000. How quantitative is quantitative PCR with respect to cell counts? Syst Appl Microbiol 23:556–562. doi:.10.1016/S0723-2020(00)80030-2 [PubMed] [Cross Ref]
69. Bernhard AE, Field KG 2000. A PCR assay to discriminate human and ruminant feces on the basis of host differences in Bacteroides-Prevotella genes encoding 16S rRNA. Appl Environ Microbiol 66:4571–4574. doi:.10.1128/AEM.66.10.4571-4574.2000 [PMC free article] [PubMed] [Cross Ref]
70. Sinigalliano CD, Fleisher JM, Gidley ML, Solo-Gabriele HM, Shibata T, Plano LRW, Elmir SM, Wanless D, Bartkowiak J, Boiteau R, Withum K, Abdelzaher AM, He G, Ortega C, Zhu X, Wright ME, Kish J, Hollenbeck J, Scott T, Backer LC, Fleming LE 2010. Traditional and molecular analyses for fecal indicator bacteria in non-point source subtropical recreational marine waters. Water Res 44:3763–3772. doi:.10.1016/j.watres.2010.04.026 [PMC free article] [PubMed] [Cross Ref]
71. Alm E, Burke J, Hagan E 2006. Persistence and potential growth of the fecal indicator bacteria, Escherichia coli, in shoreline sand at Lake Huron. J Great Lakes Res 32:401–405. doi:.10.3394/0380-1330(2006)32[401:PAPGOT]2.0.CO;2 [Cross Ref]
72. R Core Team. 2014. R: a language and environment for statistical computing, version 3.1.1. R Foundation for Statistical Computing, Vienna, Austria.

Articles from Applied and Environmental Microbiology are provided here courtesy of American Society for Microbiology (ASM)