Use of Bacterial Source Tracking to Characterize Texas Watersheds

Transcription

1 Use of Bacterial Source Tracking to Characterize Texas Watersheds Terry Gentry 1, George Di Giovanni 2, Kevin Wagner 3, Elizabeth Casarez 2, Pauline Wanjugi 1 1 Texas A&M University; 2 University of Texas School of Public Health, Houston, El Paso Branch Campus; 3 Texas Water Resources Institute

2 Where did the Bacteria (E. coli) Come From? Potential sources Humans Domesticated animals Wildlife Methods for determining sources Source survey Modeling Bacterial source tracking

3 What is Bacterial Source Used to determine the sources of fecal contamination Tracking (BST)? Based on uniqueness of bacteria from individual sources A variety of different methods are used Often works best as part of a toolbox approach

4 BST Target Organisms Bacterial v. Microbial Source Tracking Different targets: E. coli Bacteroidales Bacteriophage Human viruses Chemicals

5 History of BST Use in Texas Lake Waco/Belton Project initiated Sep Funded by TSSWCB Evaluated utility & methods Completed Feb. 2006

6 History of BST Use in Texas Lake Waco/Belton Project Findings 4-method composite performed better than individual methods 2-method composites appeared promising ERIC-ARA = lower cost but more sample & data processing ERIC-RP = higher cost but automated TMDL Task Force Report 2007 Confirmed ERIC-RP as recommended method

7 Methods: Library-Dependent BST Methods 1 2 M 3 M4 1M M5 76 M8 7M M M M DNA fingerprinting Enterobacterial repetitive intergenic consensus sequence-polymerase chain reaction (ERIC-PCR) RiboPrinting (RP) Advantages/Disadvantages: More discriminating Allows (A) ranking of sources More expensive (B) (B)

8 Development of Texas E. coli BST Library Sources M 3 4 M 5 6 M 7 8 M M M Isolate E. coli DNA Fingerprint Add to Library (B)

9 Texas E. coli BST Library (v. 6-13) Contains 1,524 E. coli isolates from 1,358 different human and animal samples Developed by collecting over 3,500 domestic sewage, wildlife, livestock, and pet fecal samples and screening over 6,000 isolates for clones and host specificity Samples from 13 watersheds across Texas for BST including: Waco / Belton Lake San Antonio Lake Granbury Oyster Creek / Trinity River Buck Creek Little Brazos River Tributaries Attoyac Bayou Human 25% Domestic Animals 34% Wildlife 41% Additional isolates being added from ongoing and future BST projects in other areas of Texas

10 Use of Texas E. coli BST Library for Identifying Water Isolates 3 4 M 5 6 M 7 8 M M M 7 8 Isolate E. coli DNA Fingerprint Compare to Library Source ID (B)

11 Comparison to Texas E. coli BST Library Best match approach with 80% minimum similarity cutoff based on laboratory QC data Water isolate must match library isolate 80% similarity or it is considered unidentified Identification to single library isolate with highest similarity max similarity approach Similarity: 96.94% Similarity: 95.82% RP ERIC-PCR

12 Three-way v. Seven-way Split of Results Using the results Is it from human sources? Is it from livestock? Is it from wildlife? Biology Large variety of wildlife Cosmopolitan strains Geographical and temporal differences Statistics Number of isolates collected May only use three-way split for limited studies (1) Human (2) Livestock & Pets (3) Wildlife vs. Human (1) Pets (2) Cattle (3) Other livestock, avian (4) Other livestock, non-avian (5) Wildlife, avian (6) Wildlife, non-avian (7)

13 Texas E. coli BST Library Composition & Rates of Correct Classification (RCC)

14 Approach: Library-Independent BST Genotypic detection of microorganisms based on marker genes (DNA) Does not require known-source library Most common approach targets Bacteroidales Presence/absence Relative abundance M

15 What are Bacteroidales? More abundant in feces than E. coli Not pathogens Obligate anaerobes less likely to multiply in environment Subgroups appear to be host specific Markers available for humans, ruminants, horse, hog Others being tested (e.g., poultry) site/_images/bacteroidetes.jpg Limited wildlife markers

16 Use of BST Results Reconcile with: Land use Watershed source survey Modeling Stakeholder input Common sense

17 13 Studies Completed

18 BST for Attoyac Bayou Limited, library-dependent Analyze E. coli from ~100 water samples from across the study area using both ERIC-PCR and RP fingerprinting Add ~100 known-source E. coli isolates from the area to the Texas E. coli BST Library Wastewater, poultry, cattle, wildlife, etc. Library-independent Analyze ~250 water samples from across the study area using Bacteroidales PCR for human, ruminant, hog, and horse markers

19 Bacteroidales BST Results Base Flow Samples (n=225)

20 Bacteroidales BST Results Base Flow vs Storm Flow Human Ruminant Hog Horse Attoyac Bayou at SH 21 Big Iron Ore at FM 354

21 E. coli BST Results Base + Storm Samples 3-Way Split Unidentified (n=13) 13% Livestock and Domesticated Animals (n=21.5) 21% Humans (n=6) 6% Wildlife (n=63.5) 61%

22 E. coli BST Results Base + Storm Samples (7-Way Split) Humans (n=6) 6% Unidentified (n=13) 13% Cattle (n=10.5) 10% Other livestock, avian (n=3) 3% Other livestock, non-avian (n=3) 3% Wildlife, non-avian (n=47.5) 46% Pets (n=5) 5% Wildlife, avian (n=16) 15%

23 BST Summary Library-Independent Analysis Ruminant and hog (feral) markers most common Spike in ruminant and hog hits during storm events Limited Library-Dependent Analysis Major E. coli sources in watershed appear to be wildlife (feral hogs, small mammals, avian wildlife) as well as domesticated animals (cattle) Texas E. coli BST Library additions from Attoyac Bayou Significant effort to include isolates from poultry litter

24 E. coli BST Results - Lampasas (Monthly 4-way Split All Sites Combined)

25 Summary of Texas BST Studies (n=11) 3-Way Split Unidentified 12% Livestock and [CATEGORY NAME] [PERCENTAGE] Wildlife 51% Human 10%

26 Summary of Texas BST Studies (n=7) 7-Way Split Unidentified 12% Human 10% Non-Avian Wildlife 32% Cattle 13% Avian Livestock 5% Other Non-Avian Livestock 5% Pets 5% Avian Wildlife 18%

27 Future Methods & Approaches 1. Identify the Unidentified Continue expansion of BST library Evaluating naturalized E. coli 2. Improve Library-Independent BST Limited markers, but new markers being developed Geographic stability of markers Quantification?

28 BST for Plum Creek? What is the Goal of BST? Characterize watershed or monitor specific sources? How many potential sources? All, most numerous One or a few (e.g., human) What level of resolution is needed? Individual species Groups (e.g., humans, domesticated animals, and wildlife) Presence/absence, relative ranking, or absolute number for various sources

29 BST for Plum Creek? Current BST costs: ERIC-RP = $250/isolate Bacteroidales PCR General + one specific marker = $300/sample General + four specific markers = $350/sample Possible initial approach for Plum Creek BST: Three sites Samples collected monthly for one year ERIC-RP five isolates per sample 3 sites x 12 sampling events x 5 isolates/sample x $250/isolate = $45,000 Initial sample processing = $65/sample [$2,340] Does not include sample collection and transport to lab

30 Questions? Terry Gentry Texas A&M University 2474 TAMU College Station, TX Phone: (979)