Transcript Microbial Source Tracking in Two Southern Maine Watersheds

Microbial Source Tracking in Two
Southern Maine Watersheds
NEERS/SNECAFS Joint Meeting, May 9, 2003
Project Author: Kristen Whiting-Grant, Maine Sea Grant
Cayce Dalton*, AmeriCorps/Maine Conservation Corps
Fred Dillon*, U Southern Maine-Muskie School for Public Service
Steve Jones, Jackson Estuarine Lab, U New Hampshire
Michele Dionne, Wells National Estuarine Research Reserve
* presenters
Project Partners
Funded by a grant from:
Cooperative Institute for
Coastal and Estuarine
Environmental
Technology (CICEET)
In cooperation with
partners from:
What is Bacterial Pollution?
Indicates presence of fecal matter in water.
Risk of illness from water contact.
Grounds for shellfish bed closure.
Sources of Bacterial Pollution?
• Unmanaged livestock/pet waste
• Leaking sewer pipes/storm overflows
• Wildlife (incl. mammals and birds)
• Malfunctioning septic systems
Importance of Bacteria as Water
Quality Indicators
• Inexpensive surrogate for fecal
pathogens.
• Countable, not just presence/absence.
• Regulatory standards for shellfish
areas and recreational waters.
What is Microbial Source Tracking?
• ID strains of indicator bacteria
or virus
• Phenotypic or Genotypic
methods
• Unknown strains from
environment compared to
strains found in host animals
• Close matches are a basis for
source identification
• Experimental technique,
gaining attention
Why Use MST?
• Addresses biggest weakness of conventional bacterial
tests: not source specific.
• Knowing sources means corrective measures focused,
saving public resources and reducing frustration.
• Example: Expensive sewer extension in Wells, Maine, did
not significantly reduce fecal coliform levels in Little
River Estuary.
Deer Scat
Water Sample
Study Area
Local Need: History of fecal contamination in So. ME results
in closed clam flats. Mirrors state & national issue.
Wells Beach enjoys significant tourism (despite shark).
Bacteria = public health issue.
MST Step 1: Intensive Water Sampling
Sample freshwater tributaries of estuary
during winter clamming season.
Half of sampling during post-storm, snowmelt
or high flow conditions.
Test is membrane filtration using
mTEC + urea for E. coli.
MST Step 2: Collect Fecal Samples For
Reference Library
E. coli obtained from about 10 fecal samples.
Includes 3 human (septic, sewage, direct fecal).
Includes other probable sources: dog, deer, cow,
raccoon, etc.
MST Step 3: Select and Save Bacteria
Samples with high E. coli are identified.
10 bacteria are isolated on TSA and refrigerated.
Isolates transported to Jackson Estuarine Lab
for ribotyping within about two weeks.
All water samples tested for E. coli
Over ten months, 390 water samples were
collected and filtered.
From samples with high E. coli,
bacteria are isolated
Isolates were made from 136 of 390 water samples.
(Ten bacterial isolates per water sample.)
From isolates, a few representative
samples are ribotyped
Genetic analysis conducted on 27 of 390 water samples.
We counted 22,856 bacterial colonies. 159 were ribotyped.
Source Species Database
Species
Pets
Cat
Dog
Humans
Human
Septage
Wastewater
Wildlife
Coyote
Deer
Grey Fox
Muskrat
Raccoon
Red Fox
Species
ME
6
-6
40
10
17
13
23
10
3
3
-4
3
ME & NH
17
2
15
86
14
17
55
116
15
41
3
3
28
26
ME
Livestock/chickens
Cow
-Horse
-Chicken
-Birds
2
Cormorant -Duck
-Goose
-Grouse
2
Pigeon
-Robin
-Seagull
--
ME & NH
46
30
14
2
48
13
4
19
2
2
3
5
TOTAL
317
75
159 bacterial colonies from streams/estuary ribotyped.
Compared to 75 bacterial colonies from watershed.
Compared to 317 bacterial colonies from ME-NH region.
Results of MST in
Maine:
MST on small
subset of
samples.
Detailed view of
geographic
distribution of
bacteria.
Ribotyping:
Lab Procedures
•
•
•
•
•
•
DNA extracted & purified.
DNA digested w/restriction enzyme.
DNA separated via gel electrophoresis.
DNA denatured & blotted onto membrane.
Hybridization with E. coli rRNA DNA probe.
DNA exposed to a chemiluminescent
substrate & digitally imaged.
• Image enhanced & optimized in computer.
Ribotyping:
Data Analysis
•
DNA patterns are analyzed by cluster
analysis and by computing a
similarity coefficient.
•
Source species identification for
sample patterns based on degree of
matching to source species patterns.
Matching ribotype gel banding patterns:
Unknown
Which of these patterns is most similar to the unknown?
Duck
Human
Seagull
Human
Unknown
Human
Human
Duck
Seagull
Human = 93% similarity
Researcher Sets Similarity Criteria
Researcher must decide what tolerance and percent
similarity qualify as a match:
Higher standard means greater certainty
but potentially too few matches.
Of 159 ribotyped isolates:
at >80%, 112 had source-ID
at >85%, 70 would have source-ID
at >90%, 30 would have source-ID
at >95%, 10 would have source-ID
Results of Ribotyping in Webhannet
Watershed: Local Source Library
Wildlife 24%
Birds 0%*
Livestock 0%*
Unknown 47%
Pets 11%
HUMAN 18%
* not in library
Largest single source: Human (18% of 53% known)
Largest category: Wildlife (24% of 53% known)
Results of Ribotyping in Webhannet
Watershed: Regional Source Library
Livestock 11%
Birds 3%
Unknown 30%
Wildlife 29%
Pets 9%
HUMAN 18%
Largest single source: Human (18% of 70% known)
Largest category: Wildlife (29% of 70% known)
CONCLUSIONS
• Webhannet report being drafted now. Awaiting
steering committee input re: management ideas.
• Results provide data in place of speculation.
• As more regional MST is conducted, library size
should expand and % unknown diminish.
• Perhaps most useful when one or two major sources
are suspected and impact is severe.
• Driven by need and interest, the basic science
underlying MST is being researched further.
More info at web site:
www.umseagrant-mst.org