EPA841-F-05-002B
Goals
Produce a report
on the condition
of wadeable
streams of the U.S.
by December 2005
•
Promote
collaboration
across
jurisdictional
boundaries in the
examination and
assessment of
water quality
•
Build State
capacity through
use of survey
design and
comparability of
methods or
indicators
For more information contact:
Susan Holdsworth, USEPA
202-566-1187
holdsworth. susan@epa.gov
Developing a Data Analysis Plan for the
Wadeable Streams Assessment (WSA)
Key Questions WSA will Examine
The results of the Wadeable Streams Assessment will be used to characterize the
ecological condition of small streams throughout the U.S. The study is designed
like an opinion poll: 1100 sites were selected at random to represent the condition of
all streams in regions that share similar ecological characteristics. This is the first
time a national monitoring study of streams has been conducted using this approach.
Question 1: What % of the Nation's
wadeable streams resource is in good
condition ?
Addressing this question requires input
from WSA Partners regarding:
S What are the appropriate ecological
indicators for describing condition of
the resource based on the data
collected?
S How do we set expectations for these
indicators for least-disturbed settings?
S What are the thresholds or
benchmarks for judging condition as
(good, fair, poor)?
Mid-Atlantic stream condition using macroinvertebrate data
Question 2: What is the relative importance ofstressors as evaluated in the WSA?
Addressing this question requires input from WSA Partners regarding:
S What WSA measures are best for describing stressors?
S What is the linkage between stressors such as nutrients, sedimentation, habitat
alterations, etc., and biological indicators?
S What is the relative risk to the ecological indicators from the stressors?
Stressor effects on stream condition
Relative risk to macroinvertebrates
Non-Native Fish
Sedimentation
Large Wood
Riparian Habitat
Nitrogen
Phosphorous
Mine Drainage
Acidic Deposition
Acid Mine Drainage
10%
20% 30% 40%
% of Stream Length
50%
0.0
0,5
1.0 1.5
Relative Risk
2.0
Percent of stream length in poor condition for each of
the stressors to streams in the mid-Atlantic with 90%
confidence intervals around each estimate.
The length of the bar represents the increased
likelihood of encountering poor macroinvertebrate
assemblage when the stressor is also ranked poor.
-------
Key Issues in Data Interpretation
The central focus of data interpretation is to differentiate among aquatic conditions ranging from high quality natural
conditions to low quality severely altered conditions. A collaborative effort among the various partners will include
evaluation of several approaches for analyzing and reporting the assessment results at the ecoregion level II scale, and
then aggregating up to a regional and national scale. It is envisioned that partners will build on existing efforts of
states, EPA, USGS and other organizations. Because of the large-scale and multijurisdictional nature of this effort,
the key issues for data interpretation are unique and include:
Scale of Reporting
Many of this project's partners generally select monitoring sites that represent
assessing conditions for a small stretch of streams, usually in response to
specific problems. For the WSA, sites were randomly selected across large-
scale reporting units to be representative of conditions of all the waters in that
unit. Using a probability-based design, about 50 sites were randomly selected
throughout each potential reporting unit, i.e., level II ecoregion, EPA region,
and major river basin. The data from these sites will be aggregated to describe
the range of the conditions throughout the reporting unit.
Selecting the best ecological indicators
Every state and tribal agency has ecological indicators that are used as a basis
for assessing condition. In the WSA, these indicators will be evaluated for use
on regional and national scales. It is anticipated that only a few candidate
indicators will be universally applicable for all of the reporting units that
constitute the continental US. The primary biological indicator will be derived
from the benthic macroinvertebrate data collected at each stream site.
Defining least-impacted condition as reference
Each state provided a list of candidate reference sites from their monitoring and
assessment program. A subset of these sites was selected to represent a
regional reference condition for each of the ecoregions. These sites plus higher
quality sites from the probability data set will be used to develop expectations
for the ecological indicators. The regional reference condition will serve to
anchor the best quality of the indicators expected to be found throughout the
reporting units.
Determining thresholds for judging condition
A decision framework exists for each agency for how to judge the condition of
its aquatic resources. The condition is normally presented as a value system of
"good", "fair", and "poor." The thresholds that differentiate these condition
qualifiers will be determined through evaluation of current state-derived
thresholds, analyses of the data along a biological condition gradient, and in
conjunction with discussions among the partners.
1.5
o 1.0
0.5
0.0
Reference sites
Test sites
50 40 30 20 10
Count
0 10 20 30 40 50
Count
Distribution approach (comparing ambient
and reference distributions)
0 10 20 30 40 50 60 70 80 90 100
Indicator Score
Distribution function to determine ecological
condition of the water resource
Consensus-based Process to Develop the Data
Analysis Plan
The data analysis plan will be developed via:
1. Convene a workshop of about 20 experts, including researchers, state and
EPA biologists, and managers, to discuss key questions and data analysis
options, perform exploratory analyses, and prepare a detailed plan for review
and discussion among the states and other partners.
2. Convene a national meeting of states and other partners to reach consensus on
the analysis and presentation of the data for a summary report at regional and
national scales.
3. Convene regional workshops to implement the data analysis plan.
------- |