ORD’s Environmental Monitoring and Assessment Program (EMAP) Sound Science for Measuring Ecological Condition

Download Report

Transcript ORD’s Environmental Monitoring and Assessment Program (EMAP) Sound Science for Measuring Ecological Condition

ORD’s Environmental
Monitoring and Assessment
Program (EMAP)
Sound Science for Measuring Ecological Condition
www.epa.gov/emap
Key EPA Monitoring Questions
• What are the current conditions of our
ecosystems?
• Where are the conditions
improving or declining?
• What stresses are associated
with declines?
• Are our management
programs and policies
working?
What’s at Stake?
• >$1B/y spent on monitoring
• Condition of estuaries, coastlines, streams,
rivers, wetlands and lakes are still unknown.
• Effectiveness of protection and restoration
programs and policies are often unknown
GOALS of EMAP
• Develop the scientific basis for consistent,
unbiased, cost-effective measurement of
the condition of the Nation’s aquatic
ecosystems
– Status
– Trends
• Build state and tribal capacity for
monitoring condition and
transfer our technology
• Make data generally available to all
stakeholders (STORET)
Integrated
Monitoring
States conduct
Probability survey
With suite of indicators
Condition
305(b)
Reports
Associated
Stressors
Point
Source
State of the
Environment
Reports
Dose Response
Non-point
Source
Likelihood
Criteria
Standards
303(d)
List
Comparison of # of
Expected 303(d)
Sites to known sites
Accept
State
303(d) list
Waterbody has high
Probability of
Impairment
Diagnosis
< or >
=
Probability of
Impairment
Assessment
Models
Waterbody has
Moderate Probability
of impairment
Intensive sampling
to confirm impairment
TMDL
Development
Remediation
Waterbody
Impairment
Confirmed
Waterbody
Not impaired
Waterbody has low
Probability of
Impairment
De-list
No additional
Sampling (continue
to Monitor as part of
5-year cycle)
Science Behind the Scenes
Designs
3%
Indicators
35%
Assessments
10%
15%
32%
32%
43%
=
+
30%
31%
Status (Trends)
Variance Estimation
Panel Rotation
Variable Density Approaches
Spatial Balance
Frame Development
Population Identification
Ecoregion Framework
Reference Condition
28%
23%
10%
15%
14%
37%
Analysis
Data
Index Regionalization
Field Sampling
Index Calibration
Training
Index Construction
Partnerships
44%
CWA: Resource Monitoring Needs
EMAP Extramural Research Areas
GRI
STAR Grants
Western Pilot
R-EMAP
•
•
•
•
•
Coastal Initiative
Coastal Initiative – 60% to State Co-ops
Western Pilot – 60% to State Co-ops
GRI – 60% to State and other Co-ops
R-EMAP – 100% to EPA Regions
STAR – 100% to Academic Research Institutions
EMAP Design Approach
• Probabilistic Design Framework – Randomized statistical designs
that allow interpretation of monitoring data with known uncertainty, extrapolation
to the entire population of interest with a small sample size, and the ability to
statistically aggregate similar data to larger geographic areas
• Classification - meaningful groupings within resource types and/or
ecosystem types to allow better statistical design and analysis
• Biological Indicators - Direct measures of aquatic ecosystem condition,
integrates stressors, and the public can relate to them
– Streams, rivers, estuaries, lakes, reservoirs, wetlands
Probabilistic Survey Design Advantages
•
•
•
•
•
•
Representative and allows inference to system of interest
Adaptable to resource characteristics
Adjusts sample sizes to meet precision requirements
Adaptable to temporal and spatial scales of interest
Condition of streams
Unbiased
Cost-effective
Fully
Supporting
13%
Not
Supporting
13%
Delaware
Fully
Supporting
87%
Traditional Targeted Monitoring
Not
Supporting
25%
Nebraska
Fully
Supporting
75%
Traditional Targeted Monitoring
Not
Supporting
87%
Probability Survey
Not
Support
ing
5%
Fully
Supporting
95%
Probability Survey
EMAP Uses Biological Indicators
• Historic Aquatic Indicators – Measured
physical/chemical characteristics and related them to
the biological condition of an aquatic system
• Aquatic Biological Indicators – Direct
measure of condition of aquatic ecosystem,
integrates stressors, and the public can relate
Effectiveness of Design
• Eutrophication of NE US lakes
Sampling costs
% Impaired Lakes
– 4219 mostly problem lakes sampled by states for 305(b)
– 2756 non-random lakes censused (Rohm et al. 1995)
– 344 lakes with EMAP probability design (11,076 lakes total)
• Alabama reduced the cost of estuarine monitoring by
~33%, and can now report on all estuarine waters
Stream Conditions in MAHA
Potential Stressors
Fish IBI
Good
(Insufficient
Data)
17%
25%
Fair
17%
36%
Poor
Sedimentation
Riparian Habitat
Mine Drainage
Acidic Deposition
Tissue Contamination
Phosphorus
Nitrogen
Acid Mine Drainage
31%
Proportion of
Stream Length
0%
24%
14%
11%
10%
5%
5%
1%
10%
20%
40%
30%
% of Stream Length
34%
Introduced Fish
0%
10%
20%
30%
40%
Estuarine Conditions
Louisianian Province
Virginian Province
Degraded
18 ± 8%
Degraded
30 ± 6%
Undegraded
82 ± 8%
Undegraded
70 ± 6%
Condition
Unknown
10%
Habitat 14%
Metals 42%
Unknown
39%
Low Dissolved
Oxygen 49%
Low D.O.
Contaminants 28%
Contaminants 10%
Toxicity 4%
Both
2%
Stressors Associated with Degraded Condition
Statistical Change Detection
Change in Percent Area of Chesapeake Bay
with Impaired Benthic Community
% area with impaired benthos
50
*
40
30
20
10
0
1991-93
1997-98
EMAP National Demonstrations
• Estuaries – All 24 marine
coastal states monitoring
with core EMAP design and
indicators
• Streams – Mid-Atlantic
States and 12 Western States
• Great Rivers – Mississippi
River Basin
States conduct
Probability survey
With suite of indicators
Condition
3
1
2
Integrated Monitoring
and Assessment
5
8
Associated
Stressors
305(b)
Reports
Point
Source
4
State of the
Environment
Reports
Dose Response
Comparison of # of
Expected 303(d)
Sites to known sites
Non-point
Source
9,10
Thresholds of
Impairment
Standards
Accept
State
303(d) list
10
303(d)
List
10
< or >
=
6,7
Waterbody has high
Probability of
Impairment
Diagnosis
Probability of
Impairment
Assessment
Models (2 levels)
Waterbody has
Moderate Probability
of impairment
Intensive sampling
to confirm impairment
10
TMDL
Development
10
Remediation
Waterbody
Impairment
Confirmed
10
10
Waterbody
Not impaired
Waterbody has low
Probability of
Impairment
De-list
No additional
Sampling (continue
to Monitor as part of
5-year cycle)
1
Example of Integrated Monitoring and Assessment
with Maryland Biological Stream Survey Data
MBSS probability survey for benthic IBI and fish IBI measures of
stream condition (impairment for BIBI < 3, FIBI < 3),
chemical and physical measurements taken, land cover data
available
Analysis:
cumulative distribution functions (cdfs)
conditional probabilities
conditional cdfs
2
Condition of Streams in Maryland
54% of 1st order stream miles are impaired (BIBI < 3)
40% of 2nd order stream miles are impaired (BIBI < 3)
47% of 1st order stream miles are impaired (FIBI < 3)
24% of 2nd order stream miles are impaired (FIBI < 3)
0.8
0.6
0.4
0.0
2
3
4
5
1
2
3
Benthic IBI
stream order = 1
stream order = 2
4
0.8
0.6
0.4
0.2
0.0
0.0
0.2
0.4
0.6
0.8
Fract ion of St ream Miles
1.0
Benthic IBI
1.0
1
Fract ion of St ream Miles
0.2
Fract ion of St ream Miles
0.8
0.6
0.4
0.2
0.0
Fract ion of St ream Miles
1.0
stream order = 2
1.0
stream order = 1
1
2
3
Fish IBI
4
1
2
3
Fish IBI
4
5
3
4
Associated Stressors
5
Thresholds of Impairment
MBSS-derived thresholds of impairment:
• pH < 5
• ANC < 200 μeq/l
• Nitrate-nitrogen > 2 mg/l
• DO < 5 ppm
• Sulfate > 24 mg/l
• DOC > 8.0 ppm
Conditional probability thresholds of impairment:
1st order steams:
DO < 5, DO > 12
pH < 6, pH > 8
NO3 < 5, <15
SO4 < 40-50
Temp <5, temp > 25
Hilsenoff <1, Hilsenhoff > 6
2nd order steams: DO < 3, DO > 11
pH < 5, pH > 8.5
NO3 < ?
SO4 < 75
Temp <10, temp > 28
Hilsenoff <2,
7
6
7
0.8
0.6
0.4
0.2
British Columbia
Conditional value
0.0
Probability of Benthic Impact in Streams
1.0
Percent Fines in Substrate
0
20
40
60
Percent Fines (< 2mm) in Substrate
80
100
Impaired Streams in Maryland
8800 stream miles stream miles in MD
66% 1st order - 5808
17% 2nd order
7304 miles in 1st and 2nd
order streams
3725 miles of 1st and 2nd order
streams should be on 303(d) List
based on benthic impairment
8
Probability of Impairment Models
9
Combine condition information with land cover
data to predict probability of impairment
Agriculture on >3% Slopes
0.8
0.6
0.2
0.0
40
60
80
0
20
40
60
URBAN
URBAN
stream order 1
stream order 2
80
0.8
0.6
fish ibi
0.0
0.2
0.4
0.0
0.2
0.4
0.6
0.8
1.0
20
1.0
0
fish ibi
Spatial Models
for Probability
of Impairment
0.4
benthic ibi
0.6
0.4
0.0
0.2
benthic ibi
0.8
1.0
stream order 2
1.0
stream order 1
0
20
40
URBAN
60
80
0
20
40
URBAN
60
80
Data to Drive
Modeling
10
Probability of Stream Benthic Impairment for
Exceeding Levels of Catchment Urbanization
1.0
0.8
0.6
0.4
0.2
0.0
Probability of Benthic Impairment when Urban Exceeded
Maryland Biological Stream Survey
1995-97 2nd order streams
0
20
40
60
Urban = Percent of Catchment Area Urbanized
80
10