Human source identification by using a human-associated Escherichia coli genetic marker in the Mae Klong River, Thailand

Fecal contamination of receiving water from point and non-point sources causes waterborne diseases. This is a major concern in many countries, especially in low- and middle-income countries (Pandey et al. 2014). A high level of fecal bacteria in environmental water is worrisome, and identifying the sources of bacteria is crucial for mitigating fecal pollution and preventing waterborne diseases (Gomi et al. 2014). Pathogens are transmitted to humans via poorly managed water and sanitation systems. Moreover, waterborne bacteria can be associated with diarrhea, which is caused by the fecal–oral transmission of pathogens through direct human contact with fecal matter or indirect ingestion through food, fingers, and environmental water contaminated with feces (Penakalapati et al. 2017).

Fecal indicator bacteria have been selected to monitor fecal pollution in environmental water because of their high levels in sewage and feces and because they are easy to measure (Griffith et al. 2009). In addition, Escherichia coli has been investigated with respect to water quality and has been used as an indicator organism of fecal pollution in freshwater for decades (Anderson et al. 2005); however, the limited information does not provide original data from wastewater owing to the nature of this species (found in cold and warm-blooded animals and human feces) (McLellan 2004; Field & Samadpour 2007). Fecal source tracking has been important in understanding the source of fecal contamination for better water quality management and to decrease the hazard of waterborne diseases in the environment. Along with other microbial source-tracking markers, a human-associated genetic marker for E. coli H8 has recently been developed and used in Japan, Australia, and Bangladesh (Gomi et al. 2014; Warish et al. 2015; Harada et al. 2018). The H8 marker has significantly higher sensitivity and specificity (>90%), which may consequently lead to minimum false-positive and false-negative results. It has been effectively used for evaluating environmental water samples (Warish et al. 2015; Senkbeil et al. 2019). The performance and effectiveness of the marker was also tested with the fecal and wastewater samples in Thailand (Nopprapun et al. 2020). However, the marker has not been applied widely in other areas especially in surface water where other factors could be different.

The Mae Klong River is situated in western Thailand. The lower reaches of the river receive water from irrigation canal systems, industries, and domestic wastewater. Water from the river is intensively used to produce tap water by the people in the western part (Khalil et al. 2018), and the river has now been polluted by fecal matter of which the source has been unknown. Therefore, this study aimed to apply the H8 marker to the Mae Klong River and understand the sources of fecal contamination in the river by using it in association with land use data in the watershed.

Water samples were collected from the Mae Klong River during dry weather flow as fecal contamination in each location might not emanate from runoff but could be from the direct discharge of sewage. There were 10 sampling sites throughout the Mae Klong River – MK1 (downstream) to MK10 (upstream). Samples were taken on two separate days (March 13 and 18) based on the similar weather conditions and tide level. The samples on the first day were taken from MK1 to MK5 (Samut Songkhram to Ratchaburi province), whereas the samples on the second day were taken from MK6 to MK10 (Ratchaburi to Kanchanaburi province) to analyze fecal contamination sources. The grab sample of volume 1,000 mL was collected in a sterilized bottle from the middle portion of the river flow. The following water quality parameters were evaluated for all samples: electrical conductivity (EC), temperature, dissolved oxygen (DO), pH, turbidity, and total organic carbon (TOC). The samples were transported in a cooling box to the laboratory, kept in the dark at 4 °C, and processed within 12 h. Detailed information of each sampling site and its location is shown in Table 1 and Figure 1.

A total of 100 mL of river water samples was filtered using a sterilized funnel and disposable filtration devices (0.45 μm, white MCE membrane, Merck, Germany) for each sample. Five replicates were considered for each of the sampling sites. The filter papers were put on HiCrome Chromogenic Coliform Agar (HiMedia, Mumbai, India) for the detection of E. coli. After incubation for 22 h at 37 °C, blue colonies indicating E. coli on the filters were counted to determine the E. coli concentration at each sampling site of the Mae Klong River. E. coli isolates were used for source tracking with a maximum of 10 isolates from each replicated plate and a total of 50 isolates from five replicated plates. Each colony was picked using sterilized toothpicks and transferred into a 96-well plate filled with 50 μL MilliQ water. The 96-well plates were stored at −20 °C for a maximum of 24 h before PCR analysis.

SYBR green-based real-time PCR assays were performed on each E. coli isolate with the H8 marker, following Gomi et al. (2014) and Harada et al. (2018). For the real-time PCR assays, the total 15 μL PCR mixture was composed of 7.5 μL of QuantiFast SYBR green PCR (QIAGEN, Germany), 0.3 μL each of forward and reverse primer, 2 μL of E. coli samples, and 4.9 μL of MilliQ water. All PCR reactions were performed in a 96-well plate for real-time PCR with a CFX96 Touch real-time PCR Detection System (BIO RAD, Singapore). Positive (DNA from control strains) and negative (sterile water) controls were included for each PCR assay. The real-time PCR conditions were set at 95 °C × 5 min + (95 °C × 10 s + 60 °C × 30 s) × 40 cycles + melting curve analysis. The H8 genetic primer sets used for SYBR green-based PCR assays are shown in the primer and target DNA sequences in Table 2.

Land use data were provided by the Land Department Development in Thailand (Land Department Development 2015). Moreover, the population of livestock in the research area was obtained from the Department of Livestock Development in Thailand (Department of Livestock Development 2017) and the Department of Provincial Administration in Thailand (Department of Provincial Administration 2018), respectively. These land use data were processed using ArcGIS 10.1.

Statistical analyses were conducted using RStudio (R 3.4.0), a programming language for statistical analysis and data science. Proportion tests were used to assess the positive proportion between the locations. The significant differences among the sampling sites were tested using the chi-square test. The p-value was set at a significance level of p < 0.001.

The results of the water quality parameter tests, particularly EC, TOC, and DO, revealed a gradual decline in water quality along the river flow. This finding highlights the pollution caused by organic and inorganic substances, which are discharged from the upstream point, and non-point sources accumulate in the downstream region as shown in Supplementary Material, Figure S1. Particularly, for E. coli concentration, the results show a slightly increased polluted trend from upstream to downstream, with the exception of MK10, MK7, and MK5, which may be largely affected by domestic and animal waste contamination, and MK1, which is affected by seawater mixing. From MK10 to MK1, the geometric mean of E. coli concentration was 349, 60, 55, 455, 95, 218, 92, 149, 375, and 134 CFU/100 mL, respectively, as shown in Table 3. Only in the case of MK9, MK8, MK6, and MK4, the E. coli concentrations were below the maximum allowable limit (126 CFU/100 mL) set by USEPA (2012). This finding indicates that the Mae Klong River water is polluted by human and animal waste at different places and is not suitable for recreational or irrigation purposes. Pathogenic microbes present in the water may spread through contact or consumption, thereby adversely affecting human health. The E. coli concentration indicated that fecal microorganisms comprised a high level of contamination, especially at MK10, MK7, MK5, and MK2, they were too numerous to count using the standard plate count method. The results of the PCR assay with the H8 marker showed that of 500 E. coli isolates, 112 (22%) were H8 positive for all sampling sites. Gomi et al. (2014) reported that 27% of 549 E. coli isolates were H8 positive for the river water in Japan; Warish et al. (2015) reported 26% of H8-positive isolates from 307 E. coli isolates from environmental water in Australia. The H8-positive proportion was in a similar range to that in these previous studies.

The transitions of H8-positive percentages along the river were 46, 14, 10, 18, 14, 38, 12, 26, 32, and 14% at MK10 (upstream) to MK1 (downstream), respectively, as shown in Figure 2. The positive proportions along the Mae Kong River were significantly different among the locations (p < 0.001). These results indicated that human-associated E. coli highly contributed to the E. coli concentration at MK10, MK5, and MK2, which are Mueang districts in Kanchanaburi province, Ratchaburi, and Samut Songkhram province, respectively. This suggests that the fecal pollution at these three sites was caused by human sources more so than at other sites.

Built-up land, including villages, cities, towns, commercial and industrial land, communication, and utility and institutional land, are located along the river (Land Department Development 2015). The percentages of built-up land surrounding the sampling sites were 22, 31, 34, 43, 33, 32, 21, 28, 29, and 18% at MK10–MK1, respectively. Table 3 shows the comparison of the average concentration of E. coli, H8-positive percentages, and urban area percentages along the Mae Klong River. The E. coli concentrations were averaged from five replicates for each sampling site. Supplementary Material, Table S1 can be referred for complete data.

The percentage of urban area results was not significantly different, and the river water seemed to be highly polluted through the water flow. In addition, we could not identify human pollution sources from upstream to downstream and these were not comparable with H8-positive percentages (p = 0.687) and E. coli (p = 0.416). Gentry-Shields et al. (2012) reported that land use changes have been significantly associated with higher concentrations of human sources of fecal indicator microorganisms, in addition to affecting the water quality. These results indicated that in addition to GIS data, fecal source-tracking methods and other information were needed to identify fecal pollution sources.

E. coli comprises gram-negative bacteria that can be found in the intestines of a variety of warm-blooded animals and humans (Belanger et al. 2011). These bacteria can represent animal and human fecal pollution from water discharge, which can affect environmental water (Ishii & Sadowsky 2008). According to Figure 2, a higher E. coli concentration resulted in higher H8-positive percentages in the Mae Klong River, except at MK7. This relationship indicated that fecal pollution mainly increases owing to the human-originated fecal matter. In contrast, MK7 showed a different relationship between the E. coli concentration and percentage of the H8 marker. Whereas the E. coli concentration was high, the H8-positive proportion was low. This implies that there is a possible discharge source of fecal matter of non-human origin causing the increase in the E. coli concentration in this area. This possibility can be supported by the number of livestock in Kanchanaburi (2014), Ratchaburi (2015), and Samut Songkhram (2012) provinces, as shown in Figure 3. It was found that dairy cattle, buffaloes, and ducks were the most bred species at Ban Pong district (MK7) in Ratchaburi province. This large number of livestock and resultant animal feces could be a possible reason for the higher E. coli concentration at MK7.

Figure 4 shows the H8-positive percentages with the populations in each district along the Mae Klong River. The results showed that the population of the three provinces, Kanchanaburi, Ratchaburi, and Samut Songkhram, where MK10, MK5, and MK2 of the Mueang district are respectively located, were high. This population trend visually corresponds to the trend in the H8-positive proportion. The Spearman's rank correlation between H8-positive percentage and the total population at each sampling site was 0.63 (p-value = 0.06) for sampling sites in Kachanaburi and Ratchaburi provinces, where the effect of salt concentration was not observed. The higher positive proportion of a human-associated E. coli marker seems to be associated with anthropogenic activities in more populated areas. Though MK2 and MK1 are located in the same district, MK1 is in the coastal area with higher salt concentration, resulting in reduced E. coli growth (Omotoyinbo & Omotoyinbo 2017), as shown in Figure 2. Previous studies also reported the high concentration of fecal indicator bacteria associated with high populations along environmental water (DiDonato et al. 2009; Rowny & Stewart 2012). The results of this study were consistent with those of Tanaka (2018). The sampling site with a significant relationship between E. coli concentration and H8 positivity can be used to identify fecal contamination from humans, whereas at the sampling site with an insignificant relationship between E. coli concentration and H8 marker. The human-associated E. coli genetic marker data can be extremely beneficial for environmental organizations or agencies in classifying the source of contamination and focusing on the most efficient and appropriate method for contamination cleanup.

The H8 marker was applied to investigate human sources of fecal pollution in the Mae Klong River, Thailand. River water samples were collected during dry weather flow, indicating that fecal contamination might originate from the direct discharge of sewage. Our results indicated the significance of fecal source tracking. E. coli concentrations alone was insufficient to identify the source of contamination. The highest E. coli concentration at MK7 was found to have a different relationship with the percent positivity of the H8 marker. According to the number of livestock, it was found that dairy cattle, buffalo, and duck comprised the greatest number of animals at MK7. Furthermore, population data in each district along sampling sites showed that MK10, MK5, and MK2 had high population numbers, which resulted in a similar trend with that of the H8-positive percentage in upstream, middle stream, and downstream areas. Therefore, this study showed that higher prevalences and concentrations of E. coli along the Mae Klong River with the H8 marker detection indicate that domestic wastewater was largely discharged owing to a high population and anthropogenic activity in those areas. Moreover, the H8 marker is recommended as a suitable method of fecal source tracking to identify human sources of fecal pollution and for proper water quality management.

The authors would like to thank the special audit program organized by the Graduate School of Global Environmental Studies, Japan.

No competing financial interests exist.

This study was supported by the Faculty of Graduate Studies, Mahidol University, Thailand, JSPS KAKENHI Grant Number JP18K18881 and JP19H02274, and the On-site Laboratory Initiative by Kyoto University, Japan. The funders had no role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.

All relevant data are included in the paper or its Supplementary Information.