Tracking the Primary Sources of Fecal Pollution in a Tropical Watershed: a One Year Study

Supplemental information

Tracking the primary sources of fecal pollution in a tropical watershed: a one year study

Carlos Toledo-Hernandez1, Hodon Ryu2, Joel Gonzalez-Nieves1, Evelyn Huertas3, Gary A. Toranzos1, and Jorge W. Santo Domingo2*

1Department of Biology, University of Puerto Rico, San Juan, PR 00936; 2Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, OH 45268

Introduction

Several molecular methods used for the identification of fecal pollution sources have been published recently. These methods have been developed using fecal and water samples from different geographical regions, but most of validation studies have been conducted in temperate regions. As a result, little is known on the applicability of MST methods in tropical regions. The goal of this study was to shed information on the performance of conventional MST PCR-based assays in a tropical watershed system. We collected watersamples from different sites in the Rio Grande of Arecibo watershed (Puerto Rico) and challenged water and fecal DNA extracts to evaluate several MST methods.In this supplemental file we provide additional details and data on the study sites, sampling dates andculturable and molecular methods used in this research project. We furtherexplain the rationale and the interpretation of the PCR inhibition tests.

Materials and Methods

Sampling sites. Ten sampling sites along the RGA watershed were selected according to previously recorded fecal pollution levels and presumed primary sources of fecal pollution. Sampling sites 1-4, 9 and 10 were headwater tributaries of the RGA, whereas sites 6, 7, and 8 were located at the RGA. Site 1 (Lago Garza) and 2 (Río Vaca) were located within or near the Guilarte National Forest, at elevations of 736 and 542m, respectively. Human development such as human residences, poultry yards, and cattle are either scarce or absence near these two sites and therefore these sites are considered to be receiving low fecal loadings. Sites 3 and 4 are located within the Cidra River, within the town center of the municipality of Adjuntas. Contamination from non-point sources such as septic tanks, domesticated animals, and wildlife is considerably low at site 3. Site 4 is located 100m downstream from sampling point 3 and 15m from the Adjuntas treatment plant effluent. Thus, site 4 is primary impacted by the treatment plant and to a minor extent by poultry from a small neighborhood chicken flock (i.e., two to three dozen birds) and by waterfowl (duck). Sites 5-8 are located at the Río Grande de Arecibo, downstream from sites 1-4. Site 5 is surrounded by woods and scattered houses that use septic tanks. Chickens and horses were sporadically spotted during the study period. Sites 6 and 7 were located within the urban area of the municipality of Utuado. Site 6 is located approximately 1.62 km upstream from site 7, next to a grassland with occasional cattle grazing at one side and human residences at the other side, most of which are connected to the Utuado sewage treatment plant. Site 7 is located 120 m downstream from Utuado sewage treatment plant. Cattle and poultry were occasionally sighted near this sampling point. Site 8 is located right before the RGA drains into the Atlantic Ocean at sea level and it is the most downstream of all sampling sites. This site is downstream of grasslands used for cattle grazing and is close to the town center of Arecibo. Site 9 is located between a fenced-farm with approximately 140 cattle on one side and human residence with septic tanks on the other side. The farm’s fence is 4-6 m from the river, whereas human residences are located as close as 100m away from the river. Site 10 is located at Río Caunillas, 15m downstream from the Jayuya Sewage Treatment Plant effluent and surrounded by grasslands were cows and horses frequently graze. Raw sewage water samples were collected from the Adjuntas sewage treatment plant and used as positive control for human-specific assay.

Sampling dates. A total of 54 sampling events were conducted in this study. The objective wasto sample each month as frequently as possible for at least 12 month. At the end, we sampled once mostweeks during approximately a 13 month period. Specifically we sample the following dates: 10/30/2009, 11/3/2009, 11/5/09, 11/12/09, 11/17/09, 11/23/09, 12/3/09, 12/8/09, 12/10/09, 12/17/09, 1/28/10, 2/2/10, 2/11/10, 2/14/10, 2/16/10, 2/24/10, 3/3/10, 3/8/10, 3/13/10, 3/18/10, 3/31/10, 4/8/10, 4/13/10, 4/24/10, 4/27/10, 5/6/10, 5/13/10, 5/24/10, 5/27/10, 6/3/10, 6/10/10, 6/15/10, 7/2/10, 7/18/10, 7/23/10, 7/26/10, 8/3/10, 8/5/10, 8/11/10, 8/20/10, 9/3/10, 9/8/10, 9/15/10, 9/23/10, 10/12/10, 10/14/10, 10/26/10, 10/28/10, 11/5/10, 11/11/10, 11/16/10, 11/30/10, 12/14/10, and 12/22/10.

Culturable methods. Thermotolerant coliforms and enterococci were enumerated using the culture-based methods described by Santiago-Rodriguez et al (1) and recommended by published guidelines (2).

Molecular methods targeting fecal bacteria. Methods used to determine the presence of fecal bacterial groups and host-specific targets were determined using conventional PCR assays as described in papers cited in Table S1. Additionally, enumeration of enterococci was performed using qPCR assay as described by Ryu et al (3). This method was originally developed by Ludwig and Schleifer (4) and first applied with environmental waters by Hauglang et al (5).

PCR inhibition studies.To address PCR inhibition we followed the guidelines established by Bustin et al (6). In brief, we applied three surrogate tests with ten-fold diluted and undiluted DNA extracts of every fecal and water sample. Two of the assays (Bac32F, and Ccoc) target fecal bacteria often present in the GI tract of mammals and other animals that can be sources of fecal pollution (7,8). The third assay (Eub8F) is used in bacterial diversity studies (9-11). We also used a quantitative PCR (qPCR) assay that targets enterococci (Entero1) (12), a bacterial group that inhabits the gut of a wide variety of hosts and that is used as an indicator of fecal pollution (13-15). We challenged Entero1 to all of the DNA extracts (n=357 for fecal and WWTP samples, n=530 for water samples) and their respective 10-fold dilutions.

Bayesian statistics. In order to better understand the value of each assay we applied Bayesian statistics. We calculated probabilities as described elsewhere (16, 17).

Results and Discussion

Culturable enterococci and thermotoletarant coliforms were detected in all study sites (Fig S1-2). Similar results were obtained with an enterococci qPCR assay(Table S2), further indicating the presence of fecal indicators throughout this watershed.

While a relative number of fecal and water samples were positive with any of the three general fecal bacterial assays (Eub8F, Bac32F, and Cococ) (Table S3, Table 1 and 2 – manuscript), the presence of some of the MST markers was lower than expected (Fig S3). One interpretation is that currently availableMSTmethods developed withtemperate regionsamples might not be fully applicable in tropical settings or that the targeted populations are below detection limits. However, another explanation is that some of the samples contain PCR inhibitors.

To determine the importance of PCR inhibition in the samples used in this study we challenged DNA extracts to three surrogate assays that are normally positive for feces and fecally-impacted water samples (4, 5, 7, 8, 15, 17) and combined the results as a first approach to determine if many samples have significant PCR inhibitors. Specifically, if samples are negative with the Eub8F assay we suspect the potential presence of inhibitors withinDNA extract or other problems with the PCR assay. The Bac32 and Ccoc assaysfurther help us determine PCR inhibition by using the following rationale: if any given sample is negative with both of the latter assays PCR inhibition could be relevant when applying the MST host-specific assays. Although negative signalscould also suggest any of these two bacterial groups are not present in a sample, we seldom observe this in fecal samples and waters with history of fecal pollution. In addition, we dilute every single DNA extract 10-fold and use an aliquot in all PCR assays (which is recommended by MIQE guidelines, 6). A sample has been inhibited if the diluted sample is positive while the undiluted is negative. Thus, the percent of samples inhibitedis defined as the sum of DNA extracts that are negative for the undiluted DNA extracts but positive for the diluted DNA extracts divided by the total number of different samples tested. For qPCR assays, we also include in the equation the number of DNA extracts that have higher Ct values for the diluted samples compared to the undiluted samples.

Approximately79.5% of the fecal samples were positive for the three general assays (Table S3). Most of the other samples showed inhibition in only one of the assays, and therefore we concluded that only a small number of DNA extracts had inhibitors. As determined by the Entero1 qPCR assay, cattle fecal DNA extracts had the greatest number of PCR inhibited samples (Table S4). PCR inhibition was also observed for swine, pigeon, chicken, duck, and turkey samples. Dilution removed inhibition in samples that were originally considered negative when using undiluted DNA.

Inhibition in water samples was considerably lower than with the fecal samples. Judging by the fact that none of the samples showed more than 6% inhibition, it is reasonable to conclude that inhibition water DNA extracts contained less PCR inhibitors. More importantly, PCR inhibition was removed when we use diluted DNA extracts.We have tried several commercially available kits in our laboratory the MoBio kits consistently remove more PCR inhibitors than others. However, the data in this study suggest that it is important to process both undiluted and diluted DNA extracts when using PCR assays in source tracking.

References

1. Santiago-Rodriguez TM, Tremblay RL, Toledo-Hernandez C, Gonzalez-Nieves JE, Ryu H, Santo Domingo JW, Toranzos GA. 2012. Microbial quality of tropical inland waters and effects of rainfall events. Appl Environ Microbiol. 78:5160-5169.
2. US Environmental Protection Agency. 2002. Method 1600: membrane filter test method for enterococci in water, EPA-821-R-02-022. Office of Water, US Environmental Protection Agency, Washington, DC.
3. Bae S, Wuertz S. 2009. Rapid decay of host-specific fecal Bacteroidales cells in seawater as measured by quantitative PCR with propidium monoazide. Water Res. 43:4850-4859.
4. Bernhard AE, Field KG. 2000. Identification of nonpoint sources of fecal pollution in coastal waters by using host-specific 16S ribosomal DNA genetic markers from fecal anaerobes. Appl. Environ. Microbiol. 66: 1587-1594
5. Lamendella R, Santo Domingo JW, Oerther DB, Vogel JR, Stoeckel DM. 2007. Assessment of fecal pollution sources in a small northern-plains watershed using PCR and phylogenetic analyses of Bacteroidetes 16S rRNA gene. FEMS Microbiol. Ecol. 59:651-660.
6. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, Mueller R, Nolan T, Pfaffl MW, Shipley GL, Vandesompele J, Wittwer CT.2009. Minimum Information for Publication of Quantitative Real-Time PCR Experiments (MIQE) guidelines. Clin Chem. 55:611-622.
7. Simpson JM, Santo Domingo JW, Reasoner DJ. 2004. Assessment of equine fecal contamination: the search for alternative bacterial source-tracking targets. FEMS Microbiol Ecol. 47:65-75.
8. Matsuki T, Watanabe K, Fujimoto J, Miyamoto Y, Takada T, Matsumoto K, Oyaizu H, Tanaka R.2002. Development of 16S rRNA-gene-targeted group-specific primers for the detection and identification of predominant bacteria in human feces. Appl. Environ. Microbiol.68:5445-5451.
9. Revetta RP, Pemberton A, Lamendella R, Iker B, Santo Domingo JW. 2010. Identification of bacterial populations in drinking water using 16S rRNA-based sequence analyses. Water Res. 44:1353-1360.
10. Keinänen-Toivola MM, Revetta RP, Santo Domingo JW. 2006. Identification of active bacterial communities in a model drinking water biofilm system using 16S rRNA-based clone libraries. FEMS Microbiol Lett. 257:182-8.
11. Revetta RP, Matlib RS, Santo Domingo JW. 2011. 16S rRNA gene sequence analysis of drinking water using RNA and DNA extracts as targets for clone library development. Curr Microbiol. 63:50-59.
12. Ludwig W, Schleifer KH. 2000. How quantitative is quantitative PCR with respect to cell counts? Syst Appl Microbiol. 23:556-562.
13. Haugland RA, Siefring S, Lavender J, Varma M. 2012. Influences of sample interference and interference controls on quantification of enterococci fecal indicator bacteria in surface water samples by the qPCR method. Water Res. 46:5989-6001.
14. Ryu H, Henson M, Elk M, Toledo-Hernandez C, Griffith J, Blackwood D, Noble R, Gourmelon M, Glassmeyer S, Santo Domingo J. 2013. Development of quantitative PCR assays targeting 16S rRNA gene of Enterococcus spp. and their application to the identification of Enterococcus species in environmental samples. Appl. Environ. Microbiol.doi:10.1128/AEM.02802-12 .
15. Ryu H, Tran H, Ware M, Iker B, Griffin S, Edge T, Newmann N, Villegas E,Santo Domingo J. 2011. Application of leftover sample material from waterborne protozoa monitoring for the molecular detection of Bacteroidales and fecal source tracking markers. J. Microbiol. Methods. 86:337-343
16. Ryu, H, Griffith J, khan I, Hill S, Edge T, Toledo-Hernandez C, Gonzalez J, Santo Domingo J. 2012. Comparison of gull feces-specific assays targeting the 16S rRNA gene of Catellicoccus marimammalium and Streptococcus spp. Appl. Environ Microbiol. 78:1909-1916.
17. Lamendella R, Santo Domingo JW, Yannarell AC, Ghosh S, Di Giovanni G, Mackie RI, Oerther DB. 2009. Evaluation of swine-specific PCR assays used for fecal source tracking and analysis of molecular diversity of swine-specific "bacteroidales" populations. Appl Environ Microbiol. 75:5787-5796.
18. Ryu H, Lu J, Vogel J, Elk M, Chávez-Ramírez F, Ashbolt N, Santo Domingo J. 2012. Development and evaluation of a quantitative PCR assay targeting sandhill crane (Grus canadensis) fecal pollution. Appl Environ Microbiol.78:4338-4345.
19. Lamendella R, Santo Domingo JW, Yannarell AC, Ghosh S, Di Giovanni G, Mackie RI, Oerther DB. 2009. Evaluation of swine-specific PCR assays used for fecal source tracking and analysis of molecular diversity of swine-specific "bacteroidales" populations. Appl Environ Microbiol. 75:5787-5796.
20. Lu J, Santo DomingoJ, Shanks OC. 2007. Identification of chicken-specific fecal microbial sequences using a metagenomic approach. Water Res. 41:3561-3574.

Supplementary Tables

Table S1. Summary of oligonucleotide primers for PCR analyses

* Bac708R primer was used for all Bacteroidales 16S rRNA gene-based assays as a reverse primer.

** Optimum annealing temperatures determined using temperature gradient PCR

Table S2. Results from Entero1 assays for water samples

* Determined using the number of positive samples with diluted and undiluted DNA extracts.

** Determined using the number of negative undiluted DNA extracts that turn positive when extracts were diluted.

Table S3. Inhibition results from Eub8F, Bac32F, and Ccoc assays

* Determined using the undiluted DNA extracts.

** All of the diluted samples were positive.

Table S4. Results from Entero1 assay using fecal samples

* Determined using the number of negative undiluted DNA extracts that turn positive or that showed less than an equivalent of ten-fold change with Ct value when extracts were diluted.

** Determined using the number of negative undiluted DNA extracts that turn positive when extracts were diluted.

*** Number of samples that were positive but did not show inhibition based on Ct values.

Table S5. Bayesian statistics on each of the host-specific assays used in this study

Figures

Figure S1. Average of enterococci per month for each study site.

Figure S2. Average of thermotolerant coliforms per month for each study site.

Figure S3. Monthly detection frequencies of water samples for human-specific marker HF183 at each study sites: Largo Garza (A, site 1); Río Vaca (B, site 2); Río Cidra, before the Adjuntas treatment plant effluent (C, site 3); Río Cidra, after treatment plant effluent (D, site 4); Río Grande de Arecibo at Adjuntas (E, site 5); Río Grande de Arecibo before the Utuado treatment plant (F, site 6); Río Grande de Arecibo after the Utuado treatment plant (G, site 7); Rio Grande de Arecibo mouth, Arecibo (H, site 8); Río Criminales (I, site 9); Río Caunillas, after Jayuya treatment plant (J, site 10).

Figure S4. Monthly detection frequencies of water samples for cattle-specific marker CF128 at each study sites: Largo Garza (A, site 1); Río Vaca (B, site 2); Río Cidra, before the Adjuntas treatment plant effluent (C, site 3); Río Cidra, after treatment plant effluent (D, site 4); Río Grande de Arecibo at Adjuntas (E, site 5); Río Grande de Arecibo before the Utuado treatment plant (F, site 6); Río Grande de Arecibo after the Utuado treatment plant (G, site 7); Rio Grande de Arecibo mouth, Arecibo (H, site 8); Río Criminales (I, site 9); Río Caunillas, after Jayuya treatment plant (J, site 10).

Figure S5. Monthly detection frequencies of water samples for swine-specific marker PF163 at each study sites: Largo Garza (A, site 1); Río Vaca (B, site 2); Río Cidra, before the Adjuntas treatment plant effluent (C, site 3); Río Cidra, after treatment plant effluent (D, site 4); Río Grande de Arecibo at Adjuntas (E, site 5); Río Grande de Arecibo before the Utuado treatment plant (F, site 6); Río Grande de Arecibo after the Utuado treatment plant (G, site 7); Rio Grande de Arecibo mouth, Arecibo (H, site 8); Río Criminales (I, site 9); Río Caunillas, after Jayuya treatment plant (J, site 10).