https://orcid.org/0000-0002-3946-1918
Figures
Abstract
This study investigated the sources of fecal pollution in surface and groundwaters collected from three urban watersheds in Texas during dry and wet weather and identified the impact of precipitation on water quality. Water samples (n = 316 and 107 for dry and wet events, respectively) were collected biweekly from eight sampling sites (6 sites from creeks and ponds, and 2 well sites) during two-year monitoring and analyzed for six physico-chemical parameters and ten qPCR-based assays targeting general (E. coli, Enterococcus, and universal Bacteroidales), human (BacHum and HF183), animal (Rum2Bac, BacCow, BacCan), and avian (Chicken/Duck-Bac and GFD) fecal markers. Elevated concentrations of NO3-N and NO2-N were observed in ponds and creeks sites during wet weather. Fecal markers analysis indicated higher concentrations of Rum2Bac, BacCow, and BacCan markers in most of pond and creek sites under wet weather, suggesting stormwater runoff contributed to non-point sources of fecal contamination by animal sources. Furthermore, sporadically higher concentrations of these markers were detected at groundwater sampling sites, demonstrating the significant human health risk. Multivariate statistical analysis such as cluster analysis (CA) and principal coordinate analysis (PCoA) was performed to identify relationship between sampling sites; while CA majorly classified ponds, creeks, and well sites separately, PCoA identified similarities in water quality characteristics between waters of wells with ponds and creeks. Overall, results indicate ruminant and dog fecal contamination is a major concern during storm events, consequently impacting surface and groundwater quality of the study.
Citation: Vadde KK, Moghadam SV, Jafarzadeh A, Matta A, Phan DC, Johnson D, et al. (2024) Precipitation impacts the physicochemical water quality and abundance of microbial source tracking markers in urban Texas watersheds. PLOS Water 3(2): e0000209. https://doi.org/10.1371/journal.pwat.0000209
Editor: Mohan Amarasiri, Kitasato University - Sagamihara Campus: Kitasato Daigaku - Sagamihara Campus, JAPAN
Received: September 19, 2023; Accepted: January 3, 2024; Published: February 1, 2024
Copyright: © 2024 Vadde et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data is included in the Supporting Information file.
Funding: This study was funded in part by the City of San Antonio (Proposition 1 Edwards Aquifer Protection Projects) and the National Science Foundation (Award number: 1759963) grant to VK. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Surface and groundwater bodies in urban environments are vital resources for drinking, agricultural, recreational, and industrial needs [1]. Contamination of these water sources with sewage/human and animal feces is a major concern for human health as it can lead to potential gastrointestinal (GI) infections and waterborne outbreaks [2, 3]. The major sources of fecal pollution in urban areas are failing sewage infrastructures, discharges from wastewater treatment plants (WWTPs) and sewer overflows, and feces of different animal sources originating from domestic, livestock, and wildlife that can transport to water resources through surface runoff [1]. Precipitation (wet weather) and resulting flow of stormwater have been identified as one of the most common non-point sources of fecal pollution responsible for deterioration of water quality and posing threat to human and ecosystem health [4, 5]. Stormwater runoff can carry microbial contaminants from animal and human waste and have been directly linked to disease outbreaks due to presence of waterborne pathogens [6, 7]. Therefore, understanding the magnitude and sources of fecal contamination in water bodies under dry and wet weather conditions is critical for accurate assessment of water quality and to take appropriate mitigation efforts by management authorities to prevent human health risks and economic consequences [8, 9].
Screening of general fecal indicator bacteria (FIB) such as Escherichia coli, enterococci, and fecal coliforms is performed traditionally to indicate potential presence of pathogens [10]. However, several previous studies reported the survival and growth of FIB in different environments such as soils, sands, sediments, and aquatic vegetation under favorable settings, indicating their inability to discriminate between recent fecal contamination and those adapted to secondary habitats [11]. Furthermore, FIB monitoring showed a poor correlation with pathogen presence in drinking, recreational and groundwaters [12, 13]. Importantly, FIB does not discriminate sources of fecal contamination as they are commonly present in feces of human and animals [14, 15]. Identification of fecal contamination sources is vital for implementing best management practices (BMPs). To overcome limitations of FIB monitoring, microbial source tracking (MST) techniques have been developed to identify potential sources of fecal contamination by targeting DNA sequences or molecular marker genes of fecal bacteria, enteric viral and mitochondrial DNA, which are associated with human or animal hosts [16–18]. Currently, MST methodologies are employed in various aquatic environments for detection and quantification of fecal contamination sources from a wide range of hosts including humans, ruminants, horses, pigs, cattle, dogs, and birds [17, 19–22].
In general, US residents have access to safe water for drinking, domestic, and recreational purposes due to efficient water treatment facilities and stringent environmental policies [23]. However, despite these safety measures, fecal waste containing waterborne pathogens is often recognized as leading cause for deterioration of water bodies in U.S., and around 300,000 kilometres of rivers and streams are designated as impaired [24]. Furthermore, precipitation and resulting runoff is considered one of the major sources of fecal matter transported to surface and groundwater in urban areas [25]. Several previous studies indicated a strong correlation of FIB with precipitation and associated runoff carrying fecal sources from upland to receiving water bodies [8]. Additionally, intense rainfall is anticipated to enhance leaching of contaminated surface waters to sub-surfaces, affecting groundwater quality [24]. As karst aquifers (current watersheds of interest are in karst terrain) have rapid interactions between groundwater and surface waters due to large voids and conduits, contaminated surface waters can greatly impact groundwater quality during periods of heavy precipitation and urban storm water runoff [26]. Therefore, as extreme weather including flooding is predicted to increase with climate change, BMPs aimed at reducing fecal sources, restoring deteriorated waterbodies, and managing risk of future fecal contamination impacts at spatial scales necessitates application of MST and also thorough understanding of relationship between MST markers and precipitation [6, 27].
The current urban watersheds of interest, Cibolo Creek (CC), Salado Creek (SC), and Upper San Antonio River (USAR) watersheds, are three important watersheds of San Antonio River Basin located in Edwards Aquifer region of Texas [28]. Edwards Aquifer is one of the most permeable and productive aquifers in the United States and groundwaters of this aquifer serve as an important source for drinking, agricultural, industrial, and recreational needs to more than two million people in south central Texas [29]. However, a recent integrated report on San Antonio River Basin water quality by Texas Commission on Environmental Quality (TCEQ) indicated that these urban watersheds are facing fecal contamination issues and also anticipated that stormwater runoff and poorly maintained septic systems are potential sources of pollution [30]. Furthermore, previous studies demonstrated the presence of nutrient and fecal contamination in surface waters of other watersheds in the Edwards aquifer and raised concerns regarding the groundwater quality of the region [31–33]. However, there have been limited studies to identify sources of fecal contamination in surface and groundwater resources and establishing a direct relationship between surface water quality of creeks and ponds to groundwater quality in this region.
In this regard, objectives of this study were to 1) determine the extent of nutrient and fecal pollution in surface and groundwaters collected from eight different sites including ponds, creeks, and wells of three watersheds during wet and dry weather, 2) understand the relationship between MST markers concentrations and wet weather, and 3) identify relationship between surface water quality of creeks/ponds and groundwater quality of wells in the monitored watersheds by applying multivariate statistical methods. To ascertain the extent of fecal contamination, all water samples were analyzed for FIB and MST-based qPCR assays targeting general (E. coli, enterococci, and universal Bacteroidales), human (BacHum and HF183), ruminant (Rum2Bac), cattle (BacCow), canine (BacCan), and avian (Chicken/Duck-Bac and GFD) fecal markers. Multivariate statistical analyses methods such hierarchical cluster analysis (CA) and principal coordinate analysis (PCoA) were conducted to identify the relationship between surface water quality of creeks/ponds and groundwater quality of wells in the monitored watersheds. Information from this study could be valuable in developing BMPs to reduce fecal contamination levels and also for advancement of surface and groundwater water quality management at other watersheds/aquifers.
2. Materials and methods
2.1. Study area and sampling sites selection
This study was carried out at CC, SC, and USAR watersheds of San Antonio River Basin in Texas, which have drainage area of 366 (upper CC watershed only), 270, and 898 km2, respectively. Among these, SC watershed has more developed lands (62.6%) compared to USAR (30.1%) and upper CC watersheds (20.4%) [30]. A total of eight sampling sites were chosen from three watersheds based on presence and proximity to pollution sources in contributing, recharge, and transition zones of Edwards Aquifer within Bexar County region (S1 Table). Among eight sampling sites (Fig 1), four sampling sites consisting of two wells (W-1 and W-2, which are used for domestic purposes) and two ponds (P-1 and P-2, which are less than 1 km2 in surface area) are located in USAR watershed and all these sites are situated in the recharge zone. For CC watershed, one sampling site (C-1) from Cibolo creek was selected which flows in the recharge zone of north-eastern Bexar County. Cibolo creek is a stream in South Central Texas that runs about 154 km across six counties and confluences with the San Antonio River [34]. For SC watershed, one recreational pond sampling site (P-3, which is less than 1km2 in surface area) was selected that is located in the contributing zone; one site from each of Elm Waterhole creek (C-2) and Panther Springs creek (C-3) that flows in recharge and transition zones were also selected. Elm Waterhole and Panther Springs creeks are small streams that run approximately 22 and 20 km respectively, before converging into Salado Creek [35, 36]. Among sampling sites, two groundwater sites (W-1 and W-2) are anticipated to be influenced by three surface water sites (P-1, P-2, and C-3) due to their proximity (around 2-3km) and are located in recharge zone (Fig 1). The geographical coordinates and land-use information of these sites are given in S1, S2 Tables respectively.
Basemap was downloaded from the United States Geological Survey (USGS) website (https://usgs.maps.arcgis.com/home/item.html?id=37a727110755454d8fce746d6fcbbe07).
2.2 Water sample collection, processing, and weather classification
Water samples were collected as described in our previous studies [32, 33]. Briefly, 1 litre of water sample was collected using sterile 1-liter Nalgene (Rochester, NY) bottles from each sampling site on a bi-weekly basis over two years from January 2018 to February 2020, which accounted for 54 sampling events. For well sites, water samples were collected before softening or other treatments [37]. Sampling was conducted after getting permissions from the San Antonio River Authority and the Shavano Park Water Utility Department. All water samples were transported on ice to the laboratory at the University of Texas at San Antonio (UTSA, San Antonio, TX), where samples were processed within 6hrs of collection. 300mL of water samples were filtered in duplicate on a vacuum manifold through 0.45-μm-pore-size, 47 mm diameter polycarbonate membranes (Pall Corporation, MI) and immediately stored at -80°C until DNA extraction. Sterile deionized water was filtered during each sampling event to check cross-contamination during sample processing. Precipitation data for sampling sites were obtained from USGS National Water Information System and are presented in S3 Table. Based on total precipitation received within 7 days before sample collection, weather was classified as “wet” if total rainfall was 10 mm or more and “dry” if no or less than 10 mm of rainfall was received [9]. Among 54 sampling events, 14 are classified as “wet” and 40 as “dry weather”.
2.3 Physico-chemical analysis of water samples
Six physico-chemical parameters including water temperature (W.T), dissolved oxygen (DO), nitrate-N (NO3-N), nitrite-N (NO2-N), and ammonium-N (NH4-N), which are known to impact microbial activity and water pollution [38], were analyzed in ground and surface waters collected during dry and wet weather. Physico-chemical parameters were measured as mentioned in our previous studies [32, 33]. DO, W.T and pH were measured on-site using an Intellical LDO101 Field Luminescent/Optical probe and HQ40d portable multi-meter (HACH, CO). NH4-N, NO2-N and NO3-N concentrations were measured in the UTSA laboratory using USEPA Salicylate Method 10205 (HACH TNTplus 830 ultra-low range kit), USEPA Diazotization Method 10207 (HACH TNTplus 839 low range kit) and Dimethylphenol Method 10206 (HACH TNTplus 835 low range kit), respectively, using a HACH DR 2800 spectrophotometer. The physico-chemical analysis results were compared to the acceptable maximum contaminant levels (MCLs) set by the United States Environmental Protection Agency (USEPA) and TCEQ [39]. For statistical analysis, concentrations with below detection limits (BDLs) were given a value of zero [33].
2.4 Genomic DNA extraction and qPCR analyses
Genomic DNA extraction from membrane filters was carried out using DNeasy PowerSoil Kit (Qiagen, Germany). Briefly, after thawing to the room temperature, the membrane filter was transferred into the PowerBead tube and DNA was extracted following the manufacturer’s protocol. Extraction blanks were processed in all batches of DNA extractions to assess cross-contamination. Nanodrop OneC Spectrophotometer (Thermo Scientific, Wilmington, DE) was used to determine DNA purity and concentration (ng/μL) and all extracted DNA were stored at −80°C until molecular analysis.
To analyze fecal contamination in ground and surface waters collected from three watersheds during wet and dry conditions, the following previously published ten qPCR assays were carried out: E. coli (EC23S857), enterococcus (Entero1), universal Bacteroidales (BacUni), human-associated Bacteroidales (HF183, BacHum), canine-associated Bacteroidales (BacCan), ruminant-associated Bacteroidales (Rum2Bac), cattle-associated Bacteroidales (BacCow), and avian-associated fecal markers (Chicken/Duck-Bac and GFD). The primers and probes utilized for the qPCR assays are presented in Table 1. E. coli and enterococci fecal markers employed in this study were reported to correlate with public health risks [40]. Additionally, HF183 and Entero1 assays have been incorporated as Recreational Water Quality Criteria (RWQC) standards in the United States by USEPA [41, 42]. The fecal markers applied in this study were validated and tested in several previous studies for their applicability in environmental waters [43–45]. For all qPCR assays, plasmid containing target sequence for respective assay was purchased from Integrated DNA Technologies (IDT, Skokie, IL) and were used as standards. All qPCR amplifications were performed in 25 μL reaction mixtures using CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA). Among ten qPCR assays, GFD qPCR amplifications were performed in the reaction mixture (25 μL) containing 12.5 μl of SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, CA), 200 nM each of forward and reverse primers and 2 μL of template DNA. The remaining qPCR assays were performed in the reaction mixture (25 μL) containing 12.5 μL of iTaq Universal Probes Supermix (Bio-Rad, CA), 100 nM of respective probe, 800 nM each of respective forward and reverse primers, and 2 μL of template DNA. All qPCR amplifications were performed with an initial denaturation for 2 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C (57°C and 54°C for GFD and Entero1 assays, respectively). To separate non-specific products for GFD assay, melting curve analysis (temperature rise from 60°C to 95°C at 0.4°C per minute) was performed after amplification and amplified products which showed a distinct peak at a temperature of 84 ± 0.2°C were considered positive. All qPCR reactions for samples, standards, and negative controls were performed in duplicate, and gene copies of fecal markers in samples were determined based on standard curve produced using 106 to 101 copies/ reaction of plasmid standards. The qPCR data analysis including along limit of detection (LOD), limit of quantification (LOQ) and detected but not quantifiable (DNQ) and quality control for absence of PCR inhibitors and cross-contaminations during sample processing, DNA extraction, and qPCR amplifications was assessed as described in our previous study [32].
2.5 Statistical analysis
All statistical analyses were conducted using the GraphPad Prism version 9.3.1 (LaJolla, CA), SPSS version 25.0 (Chicago, IL), and R program [46]. In total, 16 water quality parameters consisting of physico-chemical and fecal marker data of all samples were analyzed for statistical analysis and normal distribution of data was assessed using the Shapiro–Wilk test [47]. The concentrations of FIB and MST markers were log-transformed to attain assumptions of equal variances and normal distribution. The non-detects were assigned as 0 and the DNQs were assigned the value of LOQ/√2 [48]. The significant differences between dry and wet weather and spatial variations were tested using student’s t-test and one-way analysis of variance (ANOVA), following Kruskal–Wallis test with Dunn’s multiple comparisons. To study relationship between surface and ground waters, multivariate statistical analyses including CA and PCoA were applied as described in our previous studies [6, 49].
3. Results and discussion
3.1 Physico-chemical analysis of ground and surface waters under dry and wet weather
A total of 316 and 107 water samples were collected from eight sampling sites during dry (n = 40) and wet (n = 14) events, respectively. The range of measured physico-chemical parameters at each site under dry and wet weather is given in Table 2 and average concentration (mg/L) of nutrients/nitrogen species (NO3-N, NO2-N, and NH4-N) detected at different sites during dry and wet conditions is presented in Fig 2. W.T of groundwaters (W-1 and W-2) were not significantly varied and ranged from 19.2 to 29.2°C, with an average of 22.6 and 22.1°C in dry and wet weather samples, respectively. However, W.T of surface waters (P-1, P-2, P-3, C-1, C-2, and C-3) varied significantly (ranging from 8.3 to 34.8°C) with an average of 22.1 and 20°C in dry and wet weather, respectively. W.T showed significant statistical variation between dry and wet weather samples (p < 0.05), but no spatial variation was observed (p > 0.05). DO levels in groundwaters were in the range of 5.1 to 9.1 mg/L, and their average concentrations in dry and wet weather were 7.6 and 7.7 mg/L, respectively. Similarly, DO levels in surface waters ranged from 0.8 to 17.5 mg/L with an average of 9.4 and 7.8 mg/L in dry and wet weather, respectively. DO levels exhibited significant spatial variation (p < 0.05) among monitored sites but no seasonal (dry vs wet weather) variation was observed. Overall, average DO levels were relatively low in groundwaters than in surface waters, which is consistent due to higher average temperatures in groundwaters (DO decreases as temperature increases). However, low average DO levels in surface waters of wet weather (which had low average W.T) could be due to entry of higher concentrations of nutrients or waste from agriculture, industrial and municipal sewage after rainfall events [50, 51]. Furthermore, DO levels were below 3 mg/L in surface waters collected from P-1, P-3, and C-3 sites during several wet weather events and their levels were below 1 mg/L (hypoxic condition) at P-3 site during a few wet events (Table 2), demonstrating deprived water quality and significant nutrient pollution at these sites. According to TCEQ’s 2014 Integrated Report (IR), 21.43% of water bodies in USAR, SC, and CC watersheds showed depressed levels of DO and concern for aquatic life [30]. The pH levels in ground and surface waters showed significant variation during dry and wet weather conditions. The pH of ground waters was in range of 5.2 to 8.3, while their levels in surface waters were ranged from 6.3 to 10. The average pH levels of groundwater samples in dry and wet weather were 6.9 and 7.0, respectively. Similarly, the average pH levels in surface waters were 7.9 and 7.7 in dry and wet weather, respectively. The pH of monitored sites showed significant spatial and seasonal (dry vs wet weather) variations (p < 0.05). Overall, groundwaters of current study, particularly from W-2 site, did not meet the U.S. National Secondary Drinking Water Regulations (6.5–8.5) on several occasions and showed low pH levels (<6). As per USEPA and TCEQ regulations, pH of 6.5 to 9 is acceptable for surface waters, although a pH range of 6.5 to 8 is ideal for maintaining aquatic life in natural waters and higher pH levels suggest potential nutrient pollution [39, 52]. In the current study, surface waters collected from ponds and creeks exceeded these pH (>9) values on multiple events, indicating potential nutrient pollution at these sites.
Average concentration (mg/L) of (A) Nitrate (NO3-N), (B) Nitrite (NO2-N), and (C) Ammonia (NH4-N) detected in the water samples collected from different sites during dry and wet events.
In the United States, NO3-N is considered a widespread pollutant in ground and surface waters that have adverse effects on human and ecological health [53]. In current study, NO3-N levels in groundwaters are in the range of 1.54 to 3.96 mg/L, with an average of 2.62 and 2.44 mg/L in dry and wet weather, respectively. For surface waters, NO3-N concentrations ranged from 0 to 4.57 mg/L and their average concentrations in dry and wet weather are 0.41 and 0.60 mg/L, respectively (Table 2 and Fig 2A). NO3-N concentrations showed significant statistical variation (p < 0.05) among sampling sites (spatial) and seasons (dry vs wet). Overall, both ground and surface waters met the U.S. Federal regulations (10 mg/L), however, all creek samples (particularly C-1 site) showed higher concentrations than estimated national background concentrations (0.24 mg/L) for streams [51, 54]. Furthermore, NO3-N concentrations at surface water sites were significantly higher during wet weather (Fig 2A), indicating stormwater runoff is contributing to these increased concentrations. The potential contributors to NO3-N include agricultural runoff, failing on-site septic systems, fertilized lawns, and wastewater treatment plants [55]. Agricultural activities are limited in the monitored watersheds [30] and elevated concentrations of NO3-N suggests discharge from fertilized lawns or fecal sources could be the primary contributors. Previous studies reported similar higher levels of NO3-N concentrations in surface waters of USAR, SC, and CC watersheds and suggested that septic systems and land application of treated wastewater might be the main source for increased concentrations [29, 36]. Among nutrients monitored, NO2-N was less detected and their concentrations were below MCLs limit (1 mg/L) set by USEPA. Among groundwater sites, NO2-N were detected only at W-1 site during wet events, with a range of 0 to 0.02 mg/L and an average of 0.015 mg/L. NO2-N concentrations in surface waters were in the range of 0 to 0.13 mg/L and their average concentrations in dry and wet weather were 0.02 and 0.05 mg/L, respectively (Fig 2B). Similar to NO3-N, high concentrations of NO2-N were detected at C-1 site and their concentrations were relatively high during wet weather, suggesting again that stormwater runoff is a major contributor to nutrient pollution. NH4-N concentrations in groundwaters were in the range of 0 to 0.05 mg/L and an average concentration of 0.03mg/L was observed in both dry and wet weather (Fig 2C). NH4-N concentrations in surface waters ranged from 0 to 1.61 mg/L and their average concentrations in dry and wet events were 0.08 and 0.06 mg/L, respectively. No significant spatial or seasonal variation in NH4-N concentrations was observed (p > 0.05). In contrast to NO3-N and NO2-N, NH4-N concentrations were higher in dry conditions than wet events (excluding C-1 and C-3 sites). Although NH4-N concentrations were within acceptable limits, their levels were relatively high at P-2 and P-3 sites. However, detection of NH4-N in aquatic environments designates increased anthropogenic activities and also entry of raw sewage, livestock manure, and agricultural runoff [49]. Overall, physico-chemical analysis reveals potential fecal pollution of ground and surface waters collected from three watershed areas, and accurate identification of fecal contamination source is required to implement BMPs.
3.2 qPCR performance characteristics and PCR inhibition
All the qPCR parameters (amplification efficiency, R2 and slope values) were within optimal recommended range (S4 Table) as mentioned in MIQE guidelines [56]. The LOD and LOQ values of the qPCR assays were range from 3 to 20 gene copies/reaction (S4 Table). PCR inhibition tests were carried out using BacUni assay on at least 12% of samples from each site and results indicated the matching concentrations of BacUni markers for undiluted and 10-fold diluted DNA samples, demonstrating the absence of PCR inhibitors. In these tests, a CT value proportional to a 10-fold dilution relative to the undiluted DNA templates resulted, suggesting that PCR inhibition did not interfere with the amplification efficiency. DNA extraction controls and no template controls (three per qPCR plate) assessment revealed the absence of cross-contamination during processing stages and qPCR experiments.
3.3 Trends in general and host-associated fecal markers during dry and wet weather events
The detection frequency of general and host-associated fecal markers in surface and groundwaters collected from three watersheds is presented in Table 3 and their median concentrations (log10 copies/100 mL) during dry and wet weather are shown in Figs 3 and 4. The absolute concentrations (log10 copies/100 mL) of host-associated fecal markers in ground and surface waters is presented in S5 to S11 Tables. The general fecal markers comprising E. coli, Entero1, and BacUni were more frequently detected (>82%) in the surface and groundwaters than host-associated fecal markers and are discussed in detail in following subsections.
Markers quantified at each sampling site and event (x-axis) are shown as log10 copies/100 mL (y-axis). White and grey color box plots represent dry and wet sampling events. The boundaries of the box indicated the 25th and 75th percentile and the middle line within the box represented the 50th percentile (median). Whiskers above and below the box are the 10th and 90th percentiles.
Markers quantified at each sampling site and event (x-axis) are shown as log10 gene copies/100 mL (y-axis). White and grey color box plots represent dry and wet sampling events. The boundaries of the box indicated the 25th and 75th percentile and the middle line within the box represented the 50th percentile (median). Whiskers above and below the box are the 10th and 90th percentiles.
3.3.1 Occurrence and concentrations of general fecal markers.
The general fecal markers analyzed include two FIB (E. coli and Entero1) and one universal Bacteroidales MST (BacUni) marker. Among these, Entero1 was detected in most of the ground and surface waters collected during dry (97%) and wet (98%) events (Table 3). The concentration of Entero1 in groundwaters (W-1 and W-2 sites) was relatively low, ranging from 2.24 to 4.62 log10 copies/100 mL and their mean concentrations were 2.70±0.28 and 2.38±0.19 log10 copies/100 mL in dry and wet weather, respectively. In surface waters, Entero1 concentrations ranged from 2.37 to 6.86 log10 copies/100 mL and their mean concentrations during dry and wet weather were 4.23± 0.94 and 4.44±1.62 log10 copies/100 mL, respectively. Entero1 marker was frequently detected with high concentrations at P-3 and C-3 sites of SC watershed (Table 3 and Fig 3). Entero1 concentrations in ground and surface waters showed significant statistical variation between dry and wet weather events (p<0.05). E. coli marker was the second most frequently detected general fecal marker in ground and surface waters and were detected in 94% and 96% of samples collected during dry and wet weather, respectively (Table 3). Similar to Entero1, concentration of E. coli marker in groundwaters was low and ranged from 2.22 to 3.50 log10 copies/100 mL, with a mean concentration of 2.53±0.30 and 2.37±0.20 log10 copies/100 mL in dry and wet weather, respectively. The concentrations of E. coli marker in surface waters was in the range of 2.22 to 6.73 log10 copies/100 mL and mean concentrations in dry and wet weather were 3.06±0.22 and 3.70±0.29 log10 copies/100 mL, respectively. The highest E. coli marker concentrations were frequently detected at C-3 site of SC watershed and concentrations of E. coli markers in ground and surface waters did not differ significantly between dry and wet weather (p>0.05). BacUni markers were detected in 93% and 96% of samples collected during dry and wet weather, respectively (Table 3). The concentration of BacUni marker in groundwaters were in the range of 2.21 to 3.53 log10 copies/100 mL, with a mean concentration of 2.49 and 2.34 log10 copies/100 mL in dry and wet weather, respectively. In surface waters, BacUni marker concentrations varied significantly and ranged from 2.26 to 7.67 log10 copies/100 mL. The mean concentration of BacUni markers in surface waters collected during dry and wet weather were 4.71±0.30 and 5.06±0.34 log10 copies/100 mL, respectively. BacUni marker concentrations in ground and surface waters differed significantly between dry and wet weather (p<0.05) and were frequently detected with high concentrations at P-3 and C-3 sites that are located in SC watershed.
Overall, sampling sites from SC watershed (mainly P-3 and C-3 sites) showed higher levels of all three general fecal markers indicating this watershed is facing higher fecal contamination issues compared to other watersheds monitored. Furthermore, analysis of general fecal markers indicated the presence of higher concentrations (about 0.5 to 3 orders of magnitude) of these markers in surface waters collected during wet weather compared to dry weather, although this trend was not observed for groundwaters. The high concentrations of general fecal markers in surface waters during wet weather indicate high levels of fecal contamination occurring at these sampling sites after storm events [57]. In this regard, detection, and quantification of host-associated MST markers are necessary for identifying source of fecal pollution and sites of major concern for possible public health risks and to apply necessary BMPs. Furthermore, as the Entero1 marker was detected with higher frequency, this study recommends the application of the Entero1 marker for assessing the general fecal contamination of environmental samples when limited resources are available. Similar to our findings, USEPA also recently suggested the quantification of Enterococcus spp. for reliable detection of fecal indicator bacteria at freshwater and marine beaches [58].
3.3.2 Occurrence and concentrations of human-associated MST markers.
Two human-associated MST markers viz HF183 and BacHum were applied to monitor human fecal contamination and the rationale for using two markers is that they could detect minute amounts of human feces or diluted sewage entering environmental waters and avoid false positive identification [8]. The occurrence of BacHum and HF183 markers in ground and surface waters was much lower than all other monitored host-associated fecal markers (Table 3). Overall, BacHum was detected only in 8.8% and 11.2% of samples (ground and surface waters) collected during dry and wet weather. Among surface waters, BacHum markers were quantifiable in few samples (Quantifiable samples (QS), n = 11/240 for dry and n = 11/84 for wet events) collected from P-3, C-1, C-2, and C-3 sites, and their concentrations were in the range of 2.25 to 4.02 log10 copies/100 mL (S5 Table). The mean concentrations of BacHum markers in quantifiable samples of surface waters collected during dry and wet weather were 3.27±1.65 and 2.52±0.29 log10 copies/100 mL and their highest concentrations were observed more frequently at C-1 and C-3 sites. However, BacHum marker concentrations were not in quantifiable range in any of groundwaters. HF183 markers were detected only in 7.5% and 12.5% of all water samples collected during dry and wet weather, respectively (Table 3). Similar to BacHum, HF183 markers were not quantifiable in groundwaters but were in quantifiable range in a few surface waters (QS, n = 9/240 for dry and n = 5/84 for wet events) collected from P-1, C-1, C-2, and C-3 sites. HF183 markers concentrations ranged from 2.21 to 3.85 log10 copies/100 mL (S6 Table) and their mean concentrations in dry and wet weather were 2.90±1.50 and 2.75±0.33 log10 copies/100 mL. Similar to BacHum, higher levels of HF183 markers were detected at C-1 and C-3 sites. Statistical analysis to identify significant differences in concentrations of BacHum and HF183 makers during dry and wet weather could not be performed due to the presence of less quantifiable samples. Overall, frequent detection of human fecal markers at C-1 site (which has less population and no septic tanks within 1km radius) could be associated with transfer of human fecal markers from upstream sites of Cibolo Creek [59]. Our previous study carried out at Cibolo Creek that flows in Kendell County (upstream to C-1 site) indicated the presence of human fecal contamination due to entry of effluents from WWTP [59]. The C-3 site of SC watershed is the most populated location of this study and higher levels of human fecal markers could be associated with leaks in septic system infrastructures [31]. Musgrove et al., 2016 [31] reported increased NO3-N levels in streams that flow through San Antonio segment of Edwards aquifers could be due to septic leakage and land application of treated waste from septic systems. Additionally, simultaneous detection of BacHum and HF183 markers was mainly observed at the C-3 site during several sampling events (14 out of 54 events, S5, S6 Tables), strongly supporting the presence of human fecal contamination at this site. Similarly, simultaneous detection of human fecal markers was also observed at the C-1, C-2, and P-3 sites on a few occasions.
3.3.3 Occurrence and concentrations of canine-associated MST markers.
The detection frequency of BacCan markers in all samples (ground and surface waters) collected from three watersheds during dry and wet weather were 38% and 53%, respectively (Table 3). Among groundwaters, although BacCan marker was detected at both W-1 and W-2 sites, their concentrations were quantifiable in only two samples (QS, n = 1 for each of dry and wet weather) collected from W-2 site with 2.22 (dry) and 2.35 (wet weather) log10 copies/100 mL (S7 Table). For surface waters, BacCan markers were frequently detected at quantifiable range in all six sampling sites (QS, n = 54/240 for dry and n = 37/84 for wet weathers) and high concentrations of this marker were detected frequently at P-1, P-3, and C-3 sites (Fig 3). The concentrations of BacCan markers in all surface waters were in the range of 2.23 to 5.76 log10 copies/100 mL and their mean concentrations in dry and wet weather were 2.75±1.14 and 3.11±0.40 log10 copies/`100 mL respectively. BacCan marker concentrations showed significant statistical variation (p<0.05) among dry and wet weather samples. The higher levels of BacCan markers observed at P-1, P-3, and C-3 sites could be attributed to large number of unrestrained dog (around 34,363 in San Antonio) population and dogs-owning households (57% in Bexar County) in nearby areas [32, 60]. Therefore, presence of significant dog fecal contamination during wet weather could be related to the entry of non-point source of dog feces into surface waters during storm events. However, presence of elevated BacCan markers in groundwaters of W-2 site could be associated with leaching of dog fecal material from P-1 and C-3 sites due to closer proximity (Fig 1). A recent tracer tests carried at Panther Springs Creek (C-3 site) reported a groundwater velocity of 1,134 to 5,300 meters per day [61], suggesting the possibility of seepage from C-3 site to W-2 site due to proximity (<2500 meters).
3.3.4 Occurrence and concentrations of ruminant and cattle-associated MST markers.
The ruminant-associated MST marker (Rum2Bac) and cattle-associated MST marker (BacCow) were more prevalent in monitored ground and surface waters (Table 3). The occurrence of Rum2Bac markers in water samples collected from eight sampling sites during dry and wet weather was 61% and 74% respectively. The Rum2Bac marker was the most frequently detected host-associated marker in the two groundwater sites tested. Although Rum2Bac was detected at both groundwater sites (W-1 and W-2), only W-2 site has quantifiable range concentrations on a few occasions (QS, n = 1 for each of dry and wet weathers) and concentrations of these two samples were 2.22 (dry) and 2.45 (wet weather) log10 copies/100 mL (S8 Table). All six surface water sites showed quantifiable levels of Rum2Bac markers on multiple sampling events (QS, n = 23/240 for dry and n = 17/84 for wet weathers), and their concentrations were ranged from 2.22 to 3.41 log10 copies/100 mL (S8 Table). The mean concentrations of Rum2Bac markers in all surface waters collected during dry and wet weather were 2.35±0.11 and 2.48±0.9 log10 copies/100 mL. Significant statistical variation (p<0.05) in the concentrations of Rum2Bac markers was observed for dry and wet weather and their highest concentrations were detected at P-1, P-3, and C-3 sites (Fig 4).
The detection frequency of BacCow markers in all ground and surface waters collected during dry and wet weather were 49% and 59% respectively (Table 3). BacCow markers were sporadically detected at both groundwater sites (W-1 and W-2) but their concentrations were quantifiable at W-1 site during a single dry sampling event with 2.23 log10 copies/100 mL (S9 Table). Similar to Rum2Bac, BacCow marker was frequently quantified at all six surface water sites (QS, n = 75/240 for dry and n = 28/84 for wet weather). BacCow marker concentrations in surface waters ranged from 2.21 to 4.48 log10 copies/100 mL and their mean concentrations during dry and wet weather were 2.79±0.29 and 2.96±0.32 respectively. The concentrations of BacCow markers showed significant statistical variation (p<0.05) among dry and wet weather samples. BacCow markers were frequently detected with high concentrations at all three pond sites (Fig 4).
The three monitored watersheds of Bexar County are highly inhabited by numerous ruminant wildlife animals such as deer and elk [32], and fecal waste from these animals can significantly impact microbial water quality. Therefore, higher levels of ruminant fecal markers detected during wet weather could be associated with non-point sources of fecal entry into surface water bodies during storm events. However, presence of higher levels of ruminant fecal markers in groundwaters could be associated with seepage from P-1, P-2, and C-3 sites as mentioned earlier [61]. As cattle population is low in the three urban watersheds monitored, detection of BacCow markers could be due to cross-reactivity of markers to the ruminant feces [17]. Several previous reports also indicated the cross-reactivity of BacCow markers with numerous non-target hosts including pig, dog, and deer feces [62, 63].
3.3.5 Occurrence and concentrations of avian-associated MST markers.
Among avian-associated MST makers, GFD markers were more prevalent and were detected in 66% and 79% of samples (both ground and surface waters) collected during dry and wet weather respectively (Table 3). However, considering only groundwater sites (W-1 and W-2), GFD marker was less detected and was quantifiable at W-1 site in a single sample (QS, n = 1) collected during dry weather with a concentration of 3.33 log10 copies/100 mL (S10 Table). In contrast, GFD marker was the most prevalent host-associated MST marker in surface waters, suggesting avian fecal pollution has a strong influence on surface water quality. Furthermore, GFD markers were quantifiable (with a range of 2.22 to 4.46 log10 copies/100 mL) at all surface water sites collected during several sampling events (QS, n = 140/240 for dry and n = 57/84 for wet weathers) and their mean concentrations during dry and wet weather were 2.89±0.11 and 2.93±0.28 log10 copies/100 mL, respectively. The concentrations of GFD markers across sampling sites collected during dry weather were relatively similar to wet weather (Fig 4) and no significant statistical variation (p>0.05) was observed between dry and wet weather samples.
Compared to GFD, CD-Bac markers were relatively low in ground and surface waters and were detected in 32% and 36% of samples collected during dry and wet weather (Table 3). Similar to GFD marker, detection frequency of CD-Bac markers in groundwaters is very low and was quantified in only one sample collected from W-1 site during a dry sampling event and concentration was 2.24 log10 copies/100 mL (S11 Table). CD-Bac markers were detected more frequently in surface waters and were quantifiable at all pond sites during dry and wet weather (QS, n = 67/240 for dry and n = 22/84 for wet weather) than creek sites, where concentrations at C-2 and C-3 sites were not quantifiable during wet weathers. The concentrations of CD-Bac markers in surface waters were in the range of 2.21 to 5.62 log10 copies/100 mL and their mean concentrations during dry and wet weather were 3.09±0.39 and 3.81±0.58 log10 copies/100 mL. The mean concentrations of CD-Bac markers in water samples were significantly (p < 0.05) different in dry and wet weather and highest concentrations were observed at P-1 and P-3 sites. The low incidence of CD-Bac markers was expected as this marker targets members of Bacteroidales in avian feces and studies indicated inconsistency in occurrence of Bacteroidales members in avian gut or excreta, with few reporting their complete absence [64, 65]. However, more prevalence of GFD marker is anticipated as they target Helicobacter sp. that lives in the gut of a wide range of avian species including seagulls, chicken, duck, and waterfowls [21]. Also, higher concentrations of GFD markers at all surface water sites are consistent with higher avian population in Texas compared to other states of USA and suggests that bird fecal pollution is significantly impacting water quality at these sites. A recent study reported that Texas has recorded over 615 bird species with more than 50% are migratory birds that has crucial stop-over points across Texas [66]. Furthermore, higher occurrence of GFD markers during spring and fall/winter seasons is consistent with higher migratory birds stop-over in Texas during spring (mid-April to mid-May) and fall migration/passage (late August to mid-November) [67].
In summary, groundwaters showed higher concentrations of general and host-associated fecal markers (except BacCan and Rum2Bac on a few occasions at W-2 site) in dry weather compared to wet weather, which is consistent with nutrient concentrations. The lower concentrations of general and host-associated fecal markers in wet weather could be associated with their dilution during leaching process [68, 69]. However, in case of surface waters, detection of higher concentrations of general and host-associated fecal markers (particularly Rum2Bac, BacCow, and BacCan) during wet weather showed consistency with higher nutrient pollution, indicating precipitation-mediated runoff carrying ruminant and dog feces contributed to non-point sources of fecal contamination at three watersheds and suggests they could be responsible for deterioration of water quality [8, 25]. Consistent with our study, a recent MST study carried out at five mixed-use watersheds in Michigan’s Lower Peninsula indicated that precipitation mediated runoff carrying bovine and porcine feces applied to agricultural lands has led to non-point source of fecal contamination and deterioration of water quality in monitored creeks [70]. In the current study, findings of host-associated fecal markers analysis also indicated that Rum2Bac, BacCow, BacCan, and GFD markers were more frequently detected in surface waters during dry and wet weather than human fecal markers (BacHum and HF183), suggesting that animal fecal contamination is the major concern at three monitored watersheds. Furthermore, frequent detection of higher levels of animal fecal markers at P-1, P-3, and C-3 sites indicates these sampling sites are of significant concern, and among three watersheds, SC watershed is facing serious fecal contamination issues. In addition, karst topography of study area increases surface water and groundwater interaction and can contaminate groundwaters with nutrient and microbial contaminants [26]. The W-2 groundwater site (which is closer to P-1 and C-3 sites) showed sporadically higher concentrations of Rum2Bac, BacCow, and BacCan fecal markers, demonstrating significant human health risk due to zoonotic pathogens at this site. Although results indicate the presence of host-associated fecal pollution at several monitored sites, we believe certain shortcomings of the current study need to be addressed in future work to improve the conclusion. For instance, Rum2Bac, BacCan, and BacCow markers reported elsewhere indicated a high cross-reactivity with other non-target fecal sources, resulting in poor conclusions [62, 63]. Therefore, future studies at these watersheds necessitate site-specific validation of MST markers for more accurate identification of host-associated fecal pollution. Furthermore, although the general fecal marker “Entero1” was detected in most of the samples tested (412 out of 423 samples), host-associated fecal markers were not detected in the 39 samples (around 10%) that were positive for the Entero1 marker. The results highlight the potential presence of fecal contamination from other sources of the study area, such as swine or raccoons, that were not assessed in the current study or could be due to the low concentration of host-associated fecal markers tested in this study [71].
3.4 Relationship between ground and surface water in the study area
Multivariate statistical analysis using CA and PCoA was applied to explore relationship between monitored ground and surface water sites and group them based on similarities in water quality characteristics (both physico-chemical and fecal markers parameters). The dendrogram of sampling sites generated by CA is presented in Fig 5A. All eight sampling sites were grouped into two main clusters at (Dlink/Dmax) <60. The results of CA are convincing, as clustered sampling sites poses similar water quality characteristics and within clusters, majority of sampling sites share similar land use patterns. For instance, cluster 1 includes only groundwater sites viz W-1 and W-2, and cluster 2 was formed by surface waters sampling sites consisting of P-1, P-2, P-3, C-1, C-2, and C-3. Furthermore, sampling sites in cluster 2 are divided into 2 groups where C-1 site is separated from remaining surface water sites (Fig 5A). As mentioned earlier, C-1 site is located in the CC watershed, which is less developed and less populated area (S2 Table). Among water quality characteristics, C-1 site showed highest NO3-N and NO2-N levels compared to all other surface water sites (Fig 2A and 2B). Also, C-1 site showed quantifiable concentrations of human fecal markers on more sampling events (excluding C-3 site) compared to all other surface water sites (Table 3 and Fig 3). Among the remaining five surface water sites, pond sites clustered separately and C-2 and C-3 sites from SC watershed clustered separately, indicating similarities in water quality among these samples. Overall, CA results support previous studies that demonstrated correlation between watersheds or catchment areas and water chemistry, including nutrient and microbial pollutants [72, 73].
(A) Hierarchical clustering analysis: Dendrogram showing clustering of sampling sites based on water quality characteristics. (B) Principal coordinate analysis (PCoA): Bray-Curtis-based dissimilarity matrix comparing sampling sites of wells, ponds, and creeks.
PCoA was performed to identify similarities between each of water samples collected from ground and surface water sites during different sampling events. PCoA results obtained using Bray-Curtis dissimilarity matrix are more convincing regarding similarities between sampling sites and influence of surface waters on ground waters. The results showed that nutrient and microbial composition in water samples was significantly different between ground and surface water sites (Fig 5B). The ground and surface waters primarily formed three distinct clusters (with few exceptions). However, water samples collected from ponds and creeks formed two close clusters suggesting significant similarities among these surface waters. The groundwaters from W-1 and W-2 sites formed a distinct cluster with few surface water samples primarily from P-1, P-2, and C-3 sites, indicating similarities in water quality characteristics of these sites (S1 Fig). These results also highlight interaction between surface water from P-1, P-2, and C-3 with groundwaters of W-1 and W-2 sites. As mentioned earlier, fecal markers quantification results for groundwaters (primarily from W-2 site) corresponded with P-1 and C-3 sites (which are close to W-2 site), demonstrating influence of surface waters on groundwater quality. Several previous studies demonstrated that PCoA is an effective multivariate statistical method to visualize distance between samples or analyse proximity matrix between samples [74]. It is widely applied statistical tool to identify spatiotemporal similarities between samples in an extensive range of studies, including water quality, ecology, and epidemiology [6, 75, 76]. For instance, Wei et al., 2021 [75] applied PCoA and spatially categorized 30 sites of Guangzhou urban waterway system into two groups based on organophosphate pesticides concentrations in water and surficial sediments collected during dry and wet weathers; Moghadam et al., 2022 [6] efficiently classified water samples collected from 8 sampling sites of two Texas rivers into flooded and non-flooded categories based on bacterial communities.
3.5 Fecal pollution implications on human and ecosystem health of the study area
In general, human exposure to microbial and chemical pollutants in waters is considered to be responsible for about 70% of human illnesses [77]. Human contact with microbial pollutants primarily occur when untreated sewage or animal fecal-contaminated ground or surface waters are used for recreation or drinking [25]. Ingestion of harmful pathogens such as Campylobacter jejuni, Salmonella spp, and norovirus can lead to GI illnesses [78]. Therefore, proper protection and monitoring are required for water bodies used for drinking and recreational purpose.
Among surface water sites, P-3 site, which is located in Classen memorial lake, is used for recreational purposes. To understand fecal marker gene abundance data with regard to human health risk, RWQC standards established by USEPA for HF183 and gull feces-associated Catellicoccus marker gene were applied [42]. Also, USEPA has made RWQC recommendations for enterococci in recreational waters determined by a qPCR method using calibrator cell equivalent (CCE) densities model and a threshold value of 1280 or 3.1 log10 CCE/100 mL or above is anticipated to cause 32 illnesses per 1000 primary contact recreators [40]. Although Entero1 marker concentrations at P-3 site are above 3.1 log10 copies/100 mL and indicate potential human health risk, it should be noted that our results were not based on CCE densities making it difficult to conclude risk of exposure. For HF183 markers, a median concentration of 3.6 log10 copies/100 mL is hypothesized to cause GI illness in 30 out of 1,000 primary contact recreators [42, 79]. HF183 markers at P-3 site (S6 Table) were below RWQC benchmark, suggesting primary contact with these waters may not pose risks. GI illness risk associated with GFD markers detected in this study is evaluated based on gull feces-associated Catellicoccus marker standards established by Brown and colleagues [80]. According to Brown et al., 2015 [80], a median concentration exceeding 6.60 log10 copies/100 mL of gull feces-associated Catellicoccus markers are hypothesized to cause GI illness in 30 out of 1000 primary contact recreators. GFD marker concentrations at P-3 site were well below RWQC benchmark, suggesting low risk. As there are no RWQC benchmarks for remaining animal markers (Rum2Bac, BacCan, and BacCow) detected at P-3 site, it is difficult to interpret human health risk associated with these markers.
In case of groundwater sites, waters of W-1 and W-2 sites are used for drinking or domestic purposes. The sporadic detection of HF183, BacHum, Rum2Bac, and BacCan markers at these sites, which are in karst-dominated terrain, indicates risk of GI infections due to potential ingestion of zoonotic pathogens [81, 82]. In largest documented case of waterborne disease outbreak in Bexar County, TX, approximately 2,000 people developed illness after a well in Braun Station, San Antonio was contaminated with sewage [83]. Therefore, fecal marker analysis for groundwaters demonstrate necessity for regular monitoring and proper protection. Furthermore, future studies at these watersheds necessitate application of updated MST markers recommended by USEPA for more accurate identification of human health risks.
4. Conclusion
In summary, results of nutrient/nitrogen species monitoring indicated that elevated concentrations of NO3-N and NO2-N were observed in surface waters during wet weather representing increase in nutrient pollution of these waters during storm events. Host-associated fecal markers analysis indicated a similar increase in concentrations of ruminant and dog fecal contamination in surface waters during wet weather, suggesting stormwater runoff is causing non-point sources of fecal pollution. Higher frequency in detection of animal fecal markers at P-1, P-3, and C-3 sites indicates these sampling sites are of significant concern, and among three watersheds, the SC watershed is facing serious fecal contamination issues. The sporadic detection of fecal markers at well sites indicates interaction and influence of surface waters on groundwaters. Multivariate statistical analysis such as CA and PCoA identified relationship between sampling sites; CA classified ponds, creeks, and well sites separately but PCoA showed a close relationship between waters of ponds and creeks with wells. Overall, results indicated that MST methods improved our understanding of fecal contamination sources at three watersheds, which can guide management authorities to implement BMPs necessary to improve water quality and reduce public health-related risks.
Supporting information
S1 Fig. The Bray-Curtis dissimilarity matrix-based comparing water samples collected from eight sampling sites during dry and wet sampling events.
https://doi.org/10.1371/journal.pwat.0000209.s001
(TIF)
S1 Table. Description and land-use information of sampling sites selected for the current study.
https://doi.org/10.1371/journal.pwat.0000209.s002
(DOCX)
S2 Table. Land-use variables within 1 km radius of study sites.
https://doi.org/10.1371/journal.pwat.0000209.s003
(DOCX)
S3 Table. Precipitation (in mm) recorded at different sites during wet sampling events (rainfall received within 7 days before sampling event).
https://doi.org/10.1371/journal.pwat.0000209.s004
(DOCX)
S4 Table. Performance characteristics (range) of qPCR assays were carried out in this study.
https://doi.org/10.1371/journal.pwat.0000209.s005
(DOCX)
S5 Table. Concentrations (log10 GC/100 mL) of BacHum markers in water samples collected from different sampling sites during dry and wet weather events.
https://doi.org/10.1371/journal.pwat.0000209.s006
(DOCX)
S6 Table. Concentrations (log10 GC/100 mL) of HF183 markers in water samples collected from different sampling sites during dry and wet weather events.
https://doi.org/10.1371/journal.pwat.0000209.s007
(DOCX)
S7 Table. Concentrations (log10 GC/100 mL) of BacCan markers in water samples collected from different sampling sites during dry and wet weather events.
https://doi.org/10.1371/journal.pwat.0000209.s008
(DOCX)
S8 Table. Concentrations (log10 GC/100 mL) of Rum2Bac markers in water samples collected from different sampling sites during dry and wet weather events.
https://doi.org/10.1371/journal.pwat.0000209.s009
(DOCX)
S9 Table. Concentrations (log10 GC/100 mL) of BacCow markers in water samples collected from different sampling sites during dry and wet weather events.
https://doi.org/10.1371/journal.pwat.0000209.s010
(DOCX)
S10 Table. Concentrations (log10 GC/100 mL) of GFD markers in water samples collected from different sampling sites during dry and wet weather events.
https://doi.org/10.1371/journal.pwat.0000209.s011
(DOCX)
S11 Table. Concentrations (log10 GC/100 mL) of CD-Bac markers in water samples collected from different sampling sites during dry and wet weather events.
https://doi.org/10.1371/journal.pwat.0000209.s012
(DOCX)
References
- 1. Derx J, Kılıç HS, Linke R, Cervero-Aragó S, Frick C, Schijven J, et al. Probabilistic fecal pollution source profiling and microbial source tracking for an urban river catchment. Science of The Total Environment. 2023;857:159533. pmid:36270368
- 2. Penakalapati G, Swarthout J, Delahoy MJ, McAliley L, Wodnik B, Levy K, et al. Exposure to Animal Feces and Human Health: A Systematic Review and Proposed Research Priorities. Environmental Science & Technology. 2017;51(20):11537–52. pmid:28926696
- 3. Fewtrell L, Kay D. Recreational Water and Infection: A Review of Recent Findings. Current Environmental Health Reports. 2015;2(1):85–94. pmid:25821715
- 4. Lee S, Suits M, Wituszynski D, Winston R, Martin J, Lee J. Residential urban stormwater runoff: A comprehensive profile of microbiome and antibiotic resistance. Sci Total Environ. 2020;723:138033. pmid:32392682
- 5. Ahmed W, Lobos A, Senkbeil J, Peraud J, Gallard J, Harwood VJ. Evaluation of the novel crAssphage marker for sewage pollution tracking in storm drain outfalls in Tampa, Florida. Water Research. 2018;131:142–50.
- 6. Moghadam SV, Vadde KK, Phan DC, Jafarzadeh A, Kapoor V. Assessing the impact of flooding on bacterial community structure and occurrence of potentially pathogenic bacteria in Texas Rivers after Hurricane Harvey. Journal of Hazardous Materials Letters. 2022;3:100058.
- 7. Price MT, Blackwood AD, Noble RT. Integrating culture and molecular quantification of microbial contaminants into a predictive modeling framework in a low-lying, tidally-influenced coastal watershed. Science of The Total Environment. 2021;792:148232. pmid:34147794
- 8. Ahmed W, Payyappat S, Cassidy M, Besley C. Enhanced insights from human and animal host-associated molecular marker genes in a freshwater lake receiving wet weather overflows. Scientific Reports. 2019;9(1):12503. pmid:31467317
- 9. Shrestha A, Kelty CA, Sivaganesan M, Shanks OC, Dorevitch S. Fecal pollution source characterization at non-point source impacted beaches under dry and wet weather conditions. Water Research. 2020;182:116014. pmid:32622131
- 10. Napier MD, Haugland R, Poole C, Dufour AP, Stewart JR, Weber DJ, et al. Exposure to human-associated fecal indicators and self-reported illness among swimmers at recreational beaches: a cohort study. Environ Health. 2017;16(1):103. pmid:28969670
- 11. Anderson KL, Whitlock JE, Harwood VJJA, microbiology e. Persistence and differential survival of fecal indicator bacteria in subtropical waters and sediments. 2005;71(6):3041–8.
- 12. Signor RS, Roser DJ, Ashbolt NJ, Ball JE. Quantifying the impact of runoff events on microbiological contaminant concentrations entering surface drinking source waters. J Water Health. 2005;3(4):453–68. pmid:16459849
- 13. Harwood VJ, Levine AD, Scott TM, Chivukula V, Lukasik J, Farrah SR, et al. Validity of the indicator organism paradigm for pathogen reduction in reclaimed water and public health protection. Appl Environ Microbiol. 2005;71(6):3163–70. pmid:15933017
- 14. Scott TM, Rose JB, Jenkins TM, Farrah SR, Lukasik J. Microbial source tracking: current methodology and future directions. Appl Environ Microbiol. 2002;68(12):5796–803. pmid:12450798
- 15. Simpson JM, Santo Domingo JW, Reasoner DJ. Microbial source tracking: state of the science. Environ Sci Technol. 2002;36(24):5279–88. pmid:12521151
- 16. Layton A, McKay L, Williams D, Garrett V, Gentry R, Sayler G. Development of Bacteroides 16S rRNA gene TaqMan-based real-time PCR assays for estimation of total, human, and bovine fecal pollution in water. Appl Environ Microbiol. 2006;72(6):4214–24. pmid:16751534
- 17. Kildare BJ, Leutenegger CM, McSwain BS, Bambic DG, Rajal VB, Wuertz S. 16S rRNA-based assays for quantitative detection of universal, human-, cow-, and dog-specific fecal Bacteroidales: a Bayesian approach. Water Res. 2007;41(16):3701–15. pmid:17644149
- 18. Malla B, Haramoto E. Host-specific mitochondrial DNA markers for tracking the sources of fecal pollution. Current Opinion in Environmental Science & Health. 2020;16:34–46.
- 19. Green HC, Haugland RA, Varma M, Millen HT, Borchardt MA, Field KG, et al. Improved HF183 quantitative real-time PCR assay for characterization of human fecal pollution in ambient surface water samples. Appl Environ Microbiol. 2014;80(10):3086–94. pmid:24610857
- 20. Mieszkin S, Furet JP, Corthier G, Gourmelon M. Estimation of pig fecal contamination in a river catchment by real-time PCR using two pig-specific Bacteroidales 16S rRNA genetic markers. Appl Environ Microbiol. 2009;75(10):3045–54. pmid:19329663
- 21. Green HC, Dick LK, Gilpin B, Samadpour M, Field KG. Genetic markers for rapid PCR-based identification of gull, Canada goose, duck, and chicken fecal contamination in water. Appl Environ Microbiol. 2012;78(2):503–10. pmid:22081573
- 22. Kapoor V, Smith C, Santo Domingo JW, Lu T, Wendell D. Correlative Assessment of Fecal Indicators using Human Mitochondrial DNA as a Direct Marker. Environmental Science & Technology. 2013;47(18):10485–93. pmid:23919424
- 23. Gaffield SJ, Goo RL, Richards LA, Jackson RJ. Public health effects of inadequately managed stormwater runoff. Am J Public Health. 2003;93(9):1527–33. pmid:12948975
- 24. Coffey R, Paul MJ, Stamp J, Hamilton A, Johnson T. A review of water quality responses to air temperature and precipitation changes 2: nutrients, algal blooms, sediment, pathogens. JAWRA Journal of the American Water Resources Association. 2019;55(4):844–68.
- 25. McLellan SL, Sauer EP, Corsi SR, Bootsma MJ, Boehm AB, Spencer SK, et al. Sewage loading and microbial risk in urban waters of the Great Lakes. Elementa: Science of the Anthropocene. 2018;6. pmid:30393748
- 26. Diston D, Robbi R, Baumgartner A, Felleisen R. Microbial source tracking in highly vulnerable karst drinking water resources. J Water Health. 2018;16(1):138–49. pmid:29424727
- 27. Pfahl S, O’Gorman PA, Fischer EM. Understanding the regional pattern of projected future changes in extreme precipitation. Nature Climate Change. 2017;7(6):423–7.
- 28. Knebl MR, Yang ZL, Hutchison K, Maidment DR. Regional scale flood modeling using NEXRAD rainfall, GIS, and HEC-HMS/RAS: a case study for the San Antonio River Basin Summer 2002 storm event. Journal of Environmental Management. 2005;75(4):325–36. pmid:15854726
- 29. Abongwa PT, Den W, Teague A. Chemical and Carbon Isotopic Characterization of a Karst-Dominated Urbanized Watershed: Case of the Upper San Antonio River. Arch Environ Contam Toxicol. 2022;82(3):439–54. pmid:35347352
- 30.
SARA SARA. San Antonio River Basin 2018 Summary Report. 2018.
- 31. Musgrove M, Opsahl SP, Mahler BJ, Herrington C, Sample TL, Banta JR. Source, variability, and transformation of nitrate in a regional karst aquifer: Edwards aquifer, central Texas. Science of The Total Environment. 2016;568:457–69. pmid:27314899
- 32. Hinojosa J, Green J, Estrada F, Herrera J, Mata T, Phan D, et al. Determining the primary sources of fecal pollution using microbial source tracking assays combined with land-use information in the Edwards Aquifer. Water Research. 2020;184:116211. pmid:32721766
- 33. Vadde KK, Phan D, Moghadam SV, Jafarzadeh A, Matta A, Johnson D, et al. Fecal pollution source characterization in the surface waters of recharge and contributing zones of a karst aquifer using general and host-associated fecal genetic markers. Environmental Science: Processes & Impacts. 2022. pmid:36444711
- 34. Morrall D, McAvoy D, Schatowitz B, Inauen J, Jacob M, Hauk A, et al. A field study of triclosan loss rates in river water (Cibolo Creek, TX). Chemosphere. 2004;54(5):653–60. pmid:14599511
- 35. Johnson SS G; Veni G. Tracing Groundwater Flowpaths in the Edwards Aquifer Recharge Zone, Panther Springs Creek Basin, Northern Bexar County, TX. 2010. Contract No.: 5373.
- 36. Opsahl SP. Quality of surface-water runoff in selected streams in the San Antonio Segment of the Edwards Aquifer Recharge Zone, Bexar County, Texas, 1997− 2012. US Geological Survey Data Series. 2012;740:20.
- 37. Borchardt M, Stokdyk J, Kieke B, Muldoon M, Spencer S, Firnstahl A, et al. Sources and Risk Factors for Nitrate and Microbial Contamination of Private Household Wells in the Fractured Dolomite Aquifer of Northeastern Wisconsin. Environmental Health Perspectives. 2021;129:067004. pmid:34160249
- 38. Voytek MA, Ashen JB, Fogarty LR, Kirshtein JD, Landa ER. Detection of Helicobacter pylori and fecal indicator bacteria in five North American rivers. J Water Health. 2005;3(4):405–22. pmid:16459846
- 39.
Ground water and drinking water: National Primary Drinking Water Regulations [Internet]. U.S. Environmental Protection Agency. 2017. Available from: https://www.epa.gov/ground-water-and-drinking-water/national-primary-drinking-water-regulations.
- 40. Haugland RA, Siefring SD, Varma M, Dufour AP, Brenner KP, Wade TJ, et al. Standardization of enterococci density estimates by EPA qPCR methods and comparison of beach action value exceedances in river waters with culture methods. J Microbiol Methods. 2014;105:59–66. pmid:25038459
- 41.
USEPA. Method 1611: Enterococci in Water by TaqMan Quantitative Polymerase Chain Reaction (qPCR) Assay United States Environmental Protection Agency Washington, DC; 2012.
- 42. Boehm AB, Soller JA, Shanks OC. Human-Associated Fecal Quantitative Polymerase Chain Reaction Measurements and Simulated Risk of Gastrointestinal Illness in Recreational Waters Contaminated with Raw Sewage. Environmental Science & Technology Letters. 2015;2(10):270–5.
- 43. Kapoor V, Pitkanen T, Ryu H, Elk M, Wendell D, Santo Domingo JW. Distribution of human-specific bacteroidales and fecal indicator bacteria in an urban watershed impacted by sewage pollution, determined using RNA- and DNA-based quantitative PCR assays. Appl Environ Microbiol. 2015;81(1):91–9. pmid:25326295
- 44. Boehm AB, Van De Werfhorst LC, Griffith JF, Holden PA, Jay JA, Shanks OC, et al. Performance of forty-one microbial source tracking methods: a twenty-seven lab evaluation study. Water Res. 2013;47(18):6812–28. pmid:23880218
- 45. Pitkänen T, Ryu H, Elk M, Hokajärvi A-M, Siponen S, Vepsäläinen A, et al. Detection of Fecal Bacteria and Source Tracking Identifiers in Environmental Waters Using rRNA-Based RT-qPCR and rDNA-Based qPCR Assays. Environmental Science & Technology. 2013;47(23):13611–20. pmid:24187936
- 46.
R: A language and environment for statistical computing. R version 4.0. 2 [Internet]. 2020.
- 47. Kongprajug A, Denpetkul T, Chyerochana N, Mongkolsuk S, Sirikanchana K. Human Fecal Pollution Monitoring and Microbial Risk Assessment for Water Reuse Potential in a Coastal Industrial–Residential Mixed-Use Watershed. Frontiers in Microbiology. 2021;12.
- 48. Bradshaw JK, Snyder BJ, Oladeinde A, Spidle D, Berrang ME, Meinersmann RJ, et al. Characterizing relationships among fecal indicator bacteria, microbial source tracking markers, and associated waterborne pathogen occurrence in stream water and sediments in a mixed land use watershed. Water Res. 2016;101:498–509. pmid:27295624
- 49. Vadde K, Wang J, Cao L, Yuan T, McCarthy A, Sekar R. Assessment of Water Quality and Identification of Pollution Risk Locations in Tiaoxi River (Taihu Watershed), China. Water. 2018;10(2).
- 50. Giller PS, Malmqvist B, editors. The biology of streams and rivers1998.
- 51.
Opsahl SP, Musgrove M, Mahler B, Lambert RB. Water-quality observations of the San Antonio segment of the Edwards aquifer, Texas, with an emphasis on processes influencing nutrient and pesticide geochemistry and factors affecting aquifer vulnerability, 2010–16. Report. Reston, VA; 2018. Report No.: 2018–5060.
- 52.
Pearce G R, Ramzan Chaudhry M, Ghulam S. A simple methodology for water quality monitoring. HR Wallingford.: Department for International Development; 1999.
- 53. Messier KP, Kane E, Bolich R, Serre ML. Nitrate Variability in Groundwater of North Carolina using Monitoring and Private Well Data Models. Environmental Science & Technology. 2014;48(18):10804–12. pmid:25148521
- 54.
Dubrovsky NM, Burow KR, Clark GM, Gronberg JM, Hamilton PA, Hitt KJ, et al. The quality of our nation’s waters: Nutrients in the nation’s streams and groundwater, 1992–2004. Report. 2010. Report No.: 1350.
- 55. Katz BG, Griffin DW, Davis JH. Groundwater quality impacts from the land application of treated municipal wastewater in a large karstic spring basin: Chemical and microbiological indicators. Science of The Total Environment. 2009;407(8):2872–86. pmid:19232432
- 56. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem. 2009;55(4):611–22. pmid:19246619
- 57. Ahmed W, Payyappat S, Cassidy M, Harrison N, Marinoni O, Besley C. Prevalence and abundance of traditional and host-associated fecal indicators in urban estuarine sediments: Potential implications for estuarine water quality monitoring. Water Res. 2020;184:116109. pmid:32818744
- 58.
Development and validation of same-day monitoring methods for recreational water [press release]. United States Environmental Protection Agency 2023.
- 59. Tanvir Pasha ABM, Hinojosa J, Phan D, Lopez A, Kapoor V. Detection of human fecal pollution in environmental waters using human mitochondrial DNA and correlation with general and human-associated fecal genetic markers. Journal of Water and Health. 2020;18(1):8–18. pmid:32129182
- 60.
Antonio CoS. FY 2019: Unrestrained Dog Population Study 2019.
- 61. Johnson S, Schindel G, Brahana J. Tracer testing in the Edwards Aquifer. 2019.
- 62. Malla B, Ghaju Shrestha R, Tandukar S, Bhandari D, Inoue D, Sei K, et al. Validation of host-specific Bacteroidales quantitative PCR assays and their application to microbial source tracking of drinking water sources in the Kathmandu Valley, Nepal. J Appl Microbiol. 2018;125(2):609–19. pmid:29679435
- 63. Raith MR, Kelty CA, Griffith JF, Schriewer A, Wuertz S, Mieszkin S, et al. Comparison of PCR and quantitative real-time PCR methods for the characterization of ruminant and cattle fecal pollution sources. Water Res. 2013;47(18):6921–8. pmid:23871256
- 64. Lu J, Santo Domingo JW, Lamendella R, Edge T, Hill S. Phylogenetic diversity and molecular detection of bacteria in gull feces. Applied and environmental microbiology. 2008;74(13):3969–76. pmid:18469128
- 65. Ohad S, Ben-Dor S, Prilusky J, Kravitz V, Dassa B, Chalifa-Caspi V, et al. The development of a novel qPCR assay-set for identifying fecal contamination originating from domestic fowls and waterfowl in Israel. Frontiers in Microbiology. 2016;7:145. pmid:26925034
- 66. Fern RR, Morrison ML. Mapping critical areas for migratory songbirds using a fusion of remote sensing and distributional modeling techniques. Ecological Informatics. 2017;42:55–60.
- 67.
Shackelford CE, Rozenburg ER, Hunter WC, Lockwood MW. Migration and the Migratory Birds of Texas: Who They Are and Where They Are Going. FOURTH EDITION ed: Texas Parks and Wildlife; 2005.
- 68. Yuan R, Li Z, Guo S. Health risks of shallow groundwater in the five basins of Shanxi, China: Geographical, geological and human activity roles. Environ Pollut. 2023;316(Pt 1):120524. pmid:36309298
- 69. Castaldo G, Visser A, Fogg GE, Harter T. Effect of Groundwater Age and Recharge Source on Nitrate Concentrations in Domestic Wells in the San Joaquin Valley. Environmental Science & Technology. 2021;55(4):2265–75. pmid:33507730
- 70. Flood MT, Hernandez-Suarez JS, Nejadhashemi AP, Martin SL, Hyndman D, Rose JB. Connecting microbial, nutrient, physiochemical, and land use variables for the evaluation of water quality within mixed use watersheds. Water Research. 2022;219:118526. pmid:35598465
- 71. Malla B, Ghaju Shrestha R, Tandukar S, Bhandari D, Inoue D, Sei K, et al. Identification of Human and Animal Fecal Contamination in Drinking Water Sources in the Kathmandu Valley, Nepal, Using Host-Associated Bacteroidales Quantitative PCR Assays. Water. 2018;10(12):1796.
- 72. Minaya V, McClain ME, Moog O, Omengo F, Singer GA. Scale-dependent effects of rural activities on benthic macroinvertebrates and physico-chemical characteristics in headwater streams of the Mara River, Kenya. Ecological Indicators. 2013;32:116–22.
- 73. Luo K, Hu X, He Q, Wu Z, Cheng H, Hu Z, et al. Using multivariate techniques to assess the effects of urbanization on surface water quality: a case study in the Liangjiang New Area, China. Environmental Monitoring and Assessment. 2017;189(4):174. pmid:28324277
- 74. Voutilainen A, Hartikainen S, Sherwood PR, Taipale H, Tolppanen A-M, Vehviläinen-Julkunen K. Associations across spatial patterns of disease incidences, socio-demographics, and land use in Finland 1991–2010. Scandinavian Journal of Public Health. 2015;43(4):356–63. pmid:25743878
- 75. Wei G, Wang C, Niu W, Huan Q, Tian T, Zou S, et al. Occurrence and risk assessment of currently used organophosphate pesticides in overlying water and surface sediments in Guangzhou urban waterways, China. Environmental Science and Pollution Research. 2021;28(35):48194–206. pmid:33904132
- 76. El-Kholy RA, Zaghlool E, Isawi H, Soliman EA, Khalil MMH, El-Aassar A-hM, et al. Groundwater quality assessment using water quality index and multivariate statistical analysis case study: East Matrouh, Northwestern coast, Egypt. Environmental Science and Pollution Research. 2022;29(43):65699–722.
- 77. Rappaport SM. Implications of the exposome for exposure science. Journal of exposure science & environmental epidemiology. 2011;21(1):5–9. pmid:21081972
- 78. Wade TJ, Sandhu SK, Levy D, Lee S, LeChevallier MW, Katz L, et al. Did a severe flood in the Midwest cause an increase in the incidence of gastrointestinal symptoms? Am J Epidemiol. 2004;159(4):398–405. pmid:14769644
- 79. Brown KI, Graham KE, Boehm AB. Risk-Based Threshold of Gull-Associated Fecal Marker Concentrations for Recreational Water. Environmental Science & Technology Letters. 2017;4(2):44–8.
- 80. Brown KI, Boehm AB. Comparative decay of Catellicoccus marimmalium and enterococci in beach sand and seawater. Water Res. 2015;83:377–84. pmid:26196307
- 81. Feng Y, Alderisio KA, Yang W, Blancero LA, Kuhne WG, Nadareski CA, et al. Cryptosporidium genotypes in wildlife from a new york watershed. Appl Environ Microbiol. 2007;73(20):6475–83. pmid:17720824
- 82. Stein RA, Katz DE. Escherichia coli, cattle and the propagation of disease. FEMS Microbiology Letters. 2017;364(6). pmid:28333229
- 83. Solo-Gabriele H, Neumeister S. US outbreaks of cryptosporidiosis. Journal AWWA. 1996;88(9):76–86.