United States Science Advisory Board EPA-SAB-EPEC-94-003
Environmental 1400F November 1993
Protection Agency Washington, DC
AN SAB REPORT:
EVALUATION OF DRAFT
TECHNICAL GUIDANCE
ON BIOLOGICAL
CRITERIA FOR STREAMS
AND SMALL RIVERS
PREPARED BY THE BIOLOGICAL
CRITERIA SUBCOMMITTEE OF
THE ECOLOGICAL PROCESSES
AND EFFECTS COMMITTEE
-------
U.S. Environmental Protection Agency
*
NOTICE
This report has been written as part of the activities of the Science
Advisory Board, a public advisory group providing extramural scientific
information and advice to the Administrator and other officials of the
Environmental Protection Agency. The Board is structured to provide balanced,
expert assessment of scientific matters related to problems facing the Agency.
This report has not been reviewed for approval by the Agency and, hence,-the
contents of this report do not necessarily represent the views and policies of the
Environmental Protection Agency, nor of other agencies in 2 Executive Branch
of the Federal government, nor does mention of trade name : or commercial
products constitute a recommendation for use.
-------
ABSTRACT
On May 13-14, 1993, the Biological Criteria Subcommittee of the Ecological
Processes and Effects Committee reviewed the draft document, "Biological Criteria:
Technical Guidance for Streams and Small Rivers." Biological criteria (biocriteria)
are numeric or narrative expressions that describe the biotic integrity (health) of
aquatic communities in minimally impaired reference areas. The Subcommittee
concluded that the options presented for selecting reference conditions (i.e., use of
reference sites in concert with historical data, empirical models, and expert
opinion/consensus) were appropriate. The Subcommittee also supported the use of
multiple metrics to evaluate the integrity of aquatic communities, but felt that
seasonal variability requires that sampling be conducted at various times of the
year. The report stresses the importance of consistent taxonomic identification of
biological specimens, use of established museum repositories for curation of
voucher specimens, and the importance of developing diagnostic tools to
differentiate probable causes of impairment. The Subcommittee also highlights the
important linkages between the biocriteria program and other Agency efforts,
including the Environmental Monitoring and Assessment Program (EMAP), the
Framework for Ecological Risk Assessment, and the Ecoregion Research Program.
KEYWORDS: biological criteria; reference condition; ecological risk
assessment; monitoring
11
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
ECOLOGICAL PROCESSES AND EFFECTS COMMITTEE
BIOLOGICAL CRITERIA SUBCOMMITTEE
CHAIR
Dr. William E. Cooper, Zoology Department, Michigan State University, East
Lansing, Michigan
MEMBERS/CONSULTANTS
Dr. Str-Ung L. Burks, Oklahoma State University, Water Quality Research
Lahore ~y, Stillwater, Oklahoma
Dr. Kenneth W. Cummins, University of Pittsburgh, Pymatuning Laboratory of
Ecology, Linesville, Pennsylvania
Dr. Kenneth L. Dickson, University of North Texas, Institute of Applied Sciences,
Denton, Texas
Dr. Charles P. Hawkins, College of Natural Resources, Utah State University,
Logan, Utah
Dr. Robert J. Huggett, Virginia Institute of Marine Science, College of William and
Mary, Gloucester Point, Virginia
Dr. Richard Kimerle, Monsanto Corporation, St. Louis, Missouri
Dr. Alan W. MaM, Exxon Company, USA, Houston, Texas
Dr. Dean B. Premo, White Water Associates, Inc., Amasa, Michigan
Dr. Stephen T. Ross, Dept. of Biological Sciences, University of Southern
Mississippi, Hattiesburg, Mississippi
Dr. Joseph Shapiro, Limnological Research Center, University of Minnesota,
Minneapolis, Minnesota
111
-------
SCIENCE ADVISORY BOARD STAFF:
Stephanie Sanzone, Designated Federal Officer, Science Advisory Board, U.S. EPA,
401 M Street, S.W. (1400F), Washington, D.C. 20460
Marcia K. Jolly, Staff Secretary, Science Advisory Board, U.S. EPA, 401 M Street,
S.W. (1400F), Washington, D.C. 20460
IV
-------
TABLE OF CONTENTS
1. EXECUTIVE SUMMARY ....................................
2. INTRODUCTION ......................................... 3
3. GENERAL COMMENTS ...... . ......................... ---- *
3.1 The Importance of Biocriteria .......................... •
3.2 Applications of Biocriteria ..............................
3.3 Diagnosing Causes of Impairment ........................
3.4 The Need for Real Case Studies ..........................
4. DEFINING THE REFERENCE CONDITION 7
4.1 Ecoregions and Other Classification Schemes
4.2 Approaches for Defining Reference Condition
7
4.3 Chemical Characterization of Sites
4.4 Protection of Designated Reference Areas
10
5. SAMPLING DESIGN AND IMPLEMENTATION JJ
5.1 Design and Statistical Considerations
5.2 Quality Assurance and Quality Control
5.3 Taxonomy and Reference Collections
6 CHARACTERIZING BIOLOGICAL COMMUNITIES: MULTIMETRIC
APPROACH 1?
6.1 Use ar Misuse of Indices ^
6.2 Select;, .1 of Metrics g
6.3 The Role of Biomarkers
7. TRAINING AND TECHNICAL ASSISTANCE 19
8. CONCLUSIONS AND RECOMMENDATIONS 20
23
9. REFERENCES CITED
-------
1. EXECUTIVE SUMMARY
The Biological Criteria Subcommittee of the Ecological Processes and Effects
Committee has reviewed the draft document, "Biological Criteria: Technical
Guidance for Streams and Small Rivers." The draft guidance is an important step
in the effort to develop scientifically-credible biological criteria (biocriteria) as a
tool for assessing the biotic integrity, or health, of stream communities.
Biocriteria offer the opportunity for EPA and the states to evaluate and
demonstrate the adequacy of current regulatory and management activities in
protecting aquatic ecosystems. The current state of the science is sufficient to
support application^ of biocriteria for assessment of site-specific impacts and
regional trends, but not for establishing point source criteria or permit limits. In
order for biocriteria to be more broadly useful, the Agency should support the
development of diagnostic tools to differentiate probable causes of observed
changes in biological metrics. These tools would be analogous to the Toxicity
Identification Evaluation procedures associated with the Whole Effluent Toxicity
program.
We agree that the assessment of biological integrity should rely on multiple
metrics, and multiple assemblages of organisms. However, we discourage the use
of a single aggregate criterion for any site, since important changes in individual
metrics may be masked by aggregation into a single index. To assist states with
analysis and aggregation of multiple metrics, the discussion of statistical methods
in the guidance document should be expanded. In addition, the guidance should
place greater emphasis ori the statistical design of the bioassessment sampling
program, rather than focusing primarily on post-monitoring data analysis.
The definition of reference condition using reference sites is appropriate
when used in conjunction with historical data, empirical models, and expert
opinion/consensus. Definition of the reference condition, and detection of impacts
at test sites, will require sampling at a variety of temporal and spatial scales in
order to account for the natural variability of biological systems.
Taxonomic identification of collected specimens will be a critical task. The
guidance should establish minimum levels of taxonomic resolution to be achieved
for various assemblages, and the Agency should take an active role in ensuring
that there is an adequate, consistent level of taxonomic training among various
state biocriteria programs. With regard to voucher specimens from the biological
monitoring, the Agency should strongly encourage states to use the established
network of federal, state, university and other museums for regionally-centralized
curation.
-------
The guidance should also include several real case studies of the application
of biocriteria for assessment on a watershed scale, additional information on costs
associated with development and implementation of biocriteria, and cost-saving
measures which have been adopted by state biocriteria programs.
-------
2. INTRODUCTION
At the request of EPA's Office of Science and Technology in the Office of
Water, the Biological Criteria Subcommittee of the Ecological Processes and
Effects Committee met on May 13-14, 1993, to review the draft document
"Biological Criteria: Technical Guidance for Streams and Small Rivers" (April,
1993). This document builds on the Agency's past efforts to assist states to
develop and apply narrative and numeric biological criteria (biocriteria), as tools
for water resource management which complement existing chemical and physical
criteria (see, for example: Plafkin et al., 1989; EPA, 1990; 1992b). The Agency is
currently developing additional biocriteria guidance documents for lakes/reservoirs
and estuaries, and plans to develop similar guidance for large rivers and wetlands
in the future.
In reviewing the draft guidance on biocriteria for streams and small rivers,
we were asked to consider four questions:
a) Are the options presented for selecting reference conditions
scientifically sound?
b) Does the recommended approach to sampling frequency adequately
account for seasonal variability?
c) Will the recommended multimetric approach to evaluating aquatic
communities adequately characterize the resource?
d) Does the recommended approach to selection and aggregation of
biological information provide a sound basis for the development of
biocriteria?
-------
3. GENERAL COMMENTS
3.1 The Importance of Biocriteria
Maintaining the integrity of community and ecosystem structure and
function should be the ultimate goal of environmental regulation. Stream and
small river habitats are integral components of the terrestrial ecosystem
(watershed) and must be managed within that context. The development and
implementation of biocriteria for streams and other aquatic systems will allow
EPA and state agencies to evaluate and to demonstrate the adequacy of current
regulatory and management activities in protecting aquatic ecosystems. We
ap laud the Agency's efforts in this direction.
Traditional end-of-pipe, command and control regulation has tended to focus
on single stressors at specific sites. In contrast, biocriteria are response variables
that provide an integrated picture of the effects from multiple stressors (chemicals,
sedimentation, exotic species, etc.) arising from point sources, nonpoint sources,
habitat alteration and hydrological modification. Ecological responses observed at
the community level of organization, for example, offer dependable and readily-
observable indicators that integrate the impacts of multiple, and often subtle,
stressors. In addition, the community responses reflect both the adaptive ability of
populations and the resilience of the community to perturbations, and thus can
provide insight into the ecological significance of observed changes in habitat and
water quality.
In a real sense, the application of biocriteria to streams and small rivers is
an assessment of risks to stream communities. The Agency has developed a
guidance document, "Framework for Ecological Risk Assessment," that is intended
to serve as the conceptual framework for currently evolving initiatives throughout
the Agency (EPA, 1992a). Within that framework, the biocriteria guidance should
more fully identify the linkages, both conceptual and procedural, between the two
documents to highlight the risk assessment nature of the use of biocriteria for
streams. The ecological risk assessment approach will be particularly important in
diagnosing the causes of ooserved changes in biological metrics, and separating
anthropogenic impacts from natural fluctuations in stream conditions (e.g.,
dissolved oxygen, water flow, and temperature). There are too many possible
permutations and combinations of stress/response pathways to allow trial and
error risk management. Rather, an adaptive management approach based on
hierarchical diagnostic analysis needs to be developed.
-------
3.2 Applications of Biocriteria
The guidance document needs a clear up-front definition of the Agency's
goals for the biocriteria program. Potential applications of biocriteria range from
use as an assessment arid screening tool, to use in defining and evaluating permit
limits for point source dischargers, to inclusion of the criteria in permits
themselves. We feel strongly that the current state of the science underlying the
identification of reference sites and the selection of suitable biological measures of
stream communities limits the utility of biocriteria at this time to two critical
applications. First, the site-specific assessment of ecological degradation using a
valid reference site as a baseline is justifiable. This would include the use of
biocriteria as a site-specific measure of ecosystem response to remediation or
mitigation activities. Convergence of the biocriteria at the test site with the
reference site will indicate when the recovery is sufficient and the corrective
activity has been successful. Second, biocriteria may be used to assess biological
resource trends in well characterized watersheds. Representative watersheds from
different regions could be used to conduct trend analyses similar to those
incorporated into the Agency's Environmental Monitoring and Assessment
Program (EMAP).
With regard to the application of biocriteria in a regulatory context, we see
several factors as limiting at this time: the degree to which we can detect subtle
impacts; the current lack of diagnostic tools to determine the causes of observed
impacts; and the state-of-the-science in defining ecoregions and reference areas.
We address the need for diagnostic tools and improved classification schemes in
subsequent sections of this report. In general, however, we feel that the
methodologies have not yet undergone a sufficiently rigorous scientific evaluation,
from a toxicological "cause and effect" perspective, to be used for establishing point
source criteria or permit limits. This does not preclude the use of biocriteria as a
way of judging the combined effectiveness of current point and nonpoint source
controls (e.g., numerical chemical limits, whole effluent toxicity tests, and total
maximum daily load~TMDL~assessments) in protecting biological integrity.
3.3 Diagnosing Causes of Impairment
While the draft document provides detailed guidance on how to establish
biocriteria, it includes very little discussion of how to determine the probable
causes of impacts on biological communities (stress/response relationships).
Observed biological conditions at assessment sites can have multiple causes. Thus,
it is essential that guidance on identifying and prioritizing the probable causes be
developed. We suggest that an approach analogous to the Toxicity Identification
Evaluation (TIE) procedures associated with the Whole Effluent Toxicity program
-------
be developed for the biocriteria program. Such an approach could include
activities that enhance the interpretation of data collected in the biological
assessment (e.g., use of biomarkers) and guidance on sampling design required for
diagnosing probable causes of impairment. We recommend that the section
"Identification of Impact Types" in Chapter 8 of the draft document be augmented
to include guidance on diagnosing probable causes. However, the Agency should
consider developing a separate (companion) guidance document on diagnosing
probable causes of impairment in streams and small rivers. The diagnostic
approach must embody the elements of the Agency's Ecological Risk Assessment
paradigm, including evaluation of background chemical monitoring data, habitat
alteration, and introduction of exotic species, to ensure that observed biological
changes can be related to specific stressors vs. natural fluctuations.
3.4 The Need for Real Case Studies
We agree with the need to include case studies that illustrate the
application of biocriteria on a watershed basis. However, the draft guidance
contains only a hypothetical example. The inclusion of one or more "real world
case studies would be most effective in underscoring the utility of the approach
and demonstrating to potential users that it has been successfully applied in the
field. The case studies selected should represent a geographic balance (the current
document has an "Eastern bias" in terms of examples and literature cited), and
should include detailed information on implementation costs. In the current era of
ever-dwindling budgets faced by most state agencies, one of the most important
obstacles facing the adoption of a new program is the incremental cost, or cost
savings vs. the status quo. Therefore, the guidance should include information on
start-up and unit operating costs, incremental costs to attain increasingly refined
levels of taxonomic accuracy, and cost-saving measures that have been adopted by
states where a biocriteria program has been developed.
-------
4. DEFINING THE REFERENCE CONDITION
4.1 Ecoregions and Other Classification Schemes
Effective classification and pairing of reference sites with test sites is
essential to define reference condition and to detect impairment. Past research by
the Agency to define ecoregions based on land use, land form, natural vegetation,
and soil type (Omernik, 1987) has advanced our ability to identify areas having
similar biotic assemblages and communities. Such approaches have application not
only to the development of biocriteria, but also to broader Agency concerns such
as remediation and restoration of damaged aquatic and terrestrial sites, watershed
protection programs, and ecosystem approaches to environmental management.
However, we are concerned that the Agency is no longer supporting
research to refine ecoregional classification techniques. With regard to streams
and small rivers, more attention should be paid to factors that vary longitudinally
along stream ecosystems (e.g., riparian corridors, temperature, hydrology, and
channel geomorphology) and that may cross ecoregion boundaries and have more
direct effects on stream communities than general ecoregion characteristics
(Cummins et al., 1989; Statzrer and Higler, 1986; Vannote et al, 1980). In some
parts of the country, the distribution of aquatic organisms, particularly
invertebrates, tends to transcend traditional ecoregion boundaries. Therefore,
classification of stream reference sites based on currently used ecoregion criteria
may be inappropriate. We recommend that the Agency explore other ways of
classifying reference sites that explicitly recognize biogeographic and distributional
patterns of stream and riparian biota. In addition, the guidance should encourage
states to overlay airsheds, since many pollutants are transported into watersheds
by air.
4.2 Approaches for Defining Reference Condition
Defining the reference condition is one of the most critical aspects of
biocriteria development since the reference condition describes the baseline against
which test sites will be evaluated. The draft guidance document identifies four
approaches to establishing the reference condition: use of reference sites; use of
historical data; use of empirical models; and use of expert opinion/consensus. The
draft guidance has a detailed discussion of the use of reference sites, describes
criteria for their selection, and reviews approaches for classifying resources.
However, the document gives only minimal guidance on how to use or apply the
other approaches for establishing the reference condition. This issue will be
particularly important when biocriteria are developed for large rivers, lakes and
-------
reservoirs, and estuaries, because it will be increasingly difficult to identify
minimally impacted reference sites. In such cases, greater reliance will have to be
placed on historical data, empirical models and expert opinion to define the
reference condition. In addition, we suggest that the four approaches be used in
combination where possible to provide convergent evidence of the reference
condition. As currently written, the guidance suggests independent application of
the approaches.
When using reference sites to establish the reference condition, the natural
variability of communities (over seasonal, annual and longer time scales) must be
considered. Such variation can be of high magnitude and often appears to be
stochastic (e.g., McElravey et al., 1989). Thus, the final guidance document should
discuss the feasibility and costs of establishing the long term collection records
which may be needed to establish the range of conditions characteristic of an
unperturbed reference site.
While we support the use of historical data to help define biological
expectations for an area, the discussion on limitations of this approach should be
strengthened. Data sets often contain species biases depending on the purpose
for which the information was originally collected. For example, fish data sets
compiled by agencies interested in managing sport or game fisheries often do not
identify forage and other non-game fish, museum collections of fish often have a
bias against game or larger non-game fish, and natural heritage inventories are
often biased toward rare species. In addition, sampling efficacy and thoroughness
are not always known for these data sets. Quite often, historical distributions do
not accurately reflect where organisms occur, but rather where people collect (e.g.,
at bridges). This means that historical data may actually be from more disturbed
(i.e., easily accessible) sites of the r. vis.
The authors of the guidance document may find it desirable to address
"natural biodiversity" of the reference condition and adopt the concept as the ideal
reference condition. This condition can be estimated by using our best data on
what organisms were historically present in or near the reference area. This
historical information may be best for fish and is often available in archival
collections. Our "best historical data," combined with more recent survey data, and
scientific judgement, may provide the most appropriate and firm baseline for a
reference condition. It should be acknowledged, however, that past societal
decisions (e.g., introduction of "desirable exotic species" or land use/urban
development) may make the goal of achieving the natural biodiversity ideal
condition impossible.
8
-------
The issue of the introduction (intentional or unintentional) of exotic species
is controversial, and should be dealt with more specifically in the guidance. The
guidance currently excludes exotic species from the definition of reference
condition. Defining the reference condition based on naturally occurring species is
one way of avoiding a potential downward spiral of reference sites over time. On
the other hand, the reality in many regions is that the dominant species in the
community are in fact non-native introduced species. To discount such dominant,
pervasive species in the definition of reference condition may be to ignore the
reality of past introductions. We agree, however, that decisions to introduce exotic
species may be taken as the result of societal decisions, but cannot be justified on
ecological grounds.
In addition, most biocriteria indices do not incorporate a measure of the
subjective aesthetic or perceived value of populations of specific organisms.
Whereas the concept of ecologically equivalent species (e.g., with respect to trophic
level or niche) may reflect a scientifically objective approach to community
structure/function, it may be unrealistic from a socio/political perspective. For
example, recreational fishermen may not view ecologically equivalent species as
equally desirable. This concept of the aesthetic value of populations may need to
be addressed in the selection of biological metrics.
4.3 Chemical Characterisation of Sites
Another consideration when selecting reference sites is the extent to which
the sites may be adversely impacted by anthropogenic chemicals transported by
water or air and concentrated primarily in the substrate. Chemical contamination
could seriously skew the baseline definition of biotic integrity reflected in the
reference condition. The final document should include guidance on how to
determine whether or not a potential reference site is "chemically appropriate."
Some minimum set of chemical data (perhaps a full scan of halogenated organic
compounds and a suite of petroleum hydrocarbons and trace elements) should be
required from all sites, both reference and test locations. The guidance document
should also discuss the influence of total hardness and pH upon the relative
tqxicity of various metals, pH upon chemical species of ammonia, and other
attenuating factors (presence of organic chelating agents, suspended solids, etc.) It
may be possible to use chemical data collected by the Agency's Environmental
Monitoring and Assessment Program (EMAP) to satisfy some of these data
requirements.
-------
4.4 Protection of Designated Reference Areas
Once reference areas have been selected that are representative of a region,
the biotic integrity of these areas should be protected. This means, in part, that
management of the areas should be coordinated with other government agencies
(both federal and state). For example, while the draft guidance document states
that forested ridge tops are usually the least impacted sites in the Appalachian
region, the U.S. Forest Service specifically targets these areas for chemical
application (e.g., Dimilin spraying for gypsy moth control). In addition, the
natural variability of biological communities (over seasonal, annual, and longer
time scales) may require long-term collection records to establish the range of
conditions characteristic of reference sites. For these reasons, reference sites
should be selected, wherever possible, from dedicated research areas (e.g., National
Science Foundation's Long Term Ecological Research (LTER) sites, Department of
Energy facilities, U.S. Forest Service research areas, and national parks) where the
integrity of the watershed and the availability of baseline data will be supported
over the long term. In addition, states should notify other agencies and academic
institutions of the location of reference sites, perhaps through an Agency-
sponsored Register of Biocriteria Reference Sites. Even with attempts to protect
designated reference areas, there will inevitably be "drift" in the baseline reference
condition. It will be important, therefore, to monitor trends at reference sites.
10
-------
5. SAMPLING DESIGN AND IMPLEMENTATION
5.1 Design and Statistical Considerations
The draft guidance document does not adequately address the up-front
planning steps necessary to design an appropriate biological sampling program.
These steps include a decision on the goal of the program (site-specific impact
assessment vs. trends assessment), definition of the universe to be assessed, and
the level of precision needed for management decisions. The answers to these
questions will affect the intensity and pattern of sampling, the level of quality
assurance and quality control (QA/QC) required, the type of sample analysis, and
ultimately, the cost of implementation.
The discussion on conducting biosurveys (Chapter 4) does not provide
sufficient guidance on: a) how to identify the universe of concern (i.e., the test
site or reference location); b) the a priori determination of the level of difference
between a reference and test site one wants to detect, or can expect to detect with
the use of biocriteria; and, c) the determination of sampling sufficiency. These
issues have both a spatial and temporal component. While the temporal
component is discussed in some detail in the Technical Issues section, issues
related to spatial variability are less well developed. Much of the discussion
focuses oh choice of a particular habitat or substrate type, and avoids the larger
issue of sample representativeness or the units of sampling.
Given that there will always be species biases of different sampling gears,
some basic criteria need to be included to determine when a stream section has
been sufficiently sampled. Because species richness and representation of trophic
categories (e.g., herbivores, carnivores) are critical to the determination of the
Index of Biotic Integrity (IBI), this is not a trivial issue. While the intent of
sampling should not be the total documentation of species richness, which would
require many samples over a long period of time, sampling should be sufficient to
reveal the majority of species. Due to the annual dynamics of streams,
characterization of both reference and test sites should be based on sampling at
various times of the year to include the major components of the fall-winter and
spring-summer (or wet season-dry season) communities.
Ideally, sampling should be based on random, stratified, or fixed interval
arrays that will generate an unbiased estimate of the assemblage structure that
occurs in the universe in question. A sufficient number of samples should be
taken to ensure that within-site variance is sufficiently low to allow detection of
11
-------
whatever a priori defined level of between-site difference is required. Units of
sampling could be linear stream sections or time blocks of sampling.
The guidance document should refer to available methods for determining
the appropriate sample size. For example, simple approaches to determining
sampling sufficiency include: a) fitting of species abundance data to lognormal
distributions (Ludwig and Reynolds, 1988); b) construction of a species sample
curve that plots cumulative number of species against cumulative number of
samples (Brower et al., 1990); and, c) use of Elliott's precision estimate (e.g.,
number of samples required to achieve a variance of ± 40% of the mean) (Elliott,
1977).
Clearly, if sampling is restricted to one (or only a few) habitat types and to
one time of the year, the resultant bia? representation of tiie community will
compromise the ability of any analysis reveal true differences (or lack thereof)
between test and reference sites. Some examples illustrating the consequence of
selective sampling would be useful. For example, if human use has resulted in a
shift in the abundance and frequency of major channel geomorphic forms (e.g.,
riffles and pools), sampling a single habitat type (e.g., riffles) would reveal little of
the actual system-level differences in biotic structure between sites and little of the
real change in structure that would occur over time with either future degradation
or site improvement. These issues are especially problematic in western and/or
mountainous streams that are geomorphically and hydrologically complex (chaotic
according to geomorphologists), and in which the primary stressors are associated
with landscape alterations. For specific guidance on sampling design, consult
Elliott (1971), Green (1979), Hurlbert (1984), and Waters and Erman (1990).
A related question is that of determining when statistically significant
changes in biocriteria are ecologically significant with respect to long-term stability
of the ecosystem. It is well understood that biological communities are always in
a state of flux, oscillating around some general community mean level of
abundance and organization. Care must be taken to fully describe these natural
oscillations and delineate them from those changes brought about by degradation
of water quality, habitat, etc. The guidance document should reference the series
of papers being developed by the Agency's Risk Assessment Forum which address
the issue of ecological significance.
The guidance document should also reference recent advances in the
ecological impact assessment field to provide more meaningful statistical
descriptors of populations and communities than have traditionally been available
through normal theory statistics. The application of binomial theory tests and
packages such as GLIM (Generalized Linear Interactive Model) have significantly
12
-------
advanced our ability to test for effects of confounding variables and covariates that
have traditionally plagued ecologists attempting to compare structural and
functional measures between impacted and unimpacted communities. A state-of-
the-science review of these techniques, and examples of their application, should be
considered for addition to the document.
5.2 Quality Assurance and Quality Control
We appreciated the inclusion of the quality assurance/quality control
(QA/QC) section in the guidance and the discussion of elements that should be
included in all field biological assessments. However, a few issues are presented
here for consideration. First, many of the biologists that will be implementing the
biocriteria program at the state level may not be familiar with a fully-documented
field assessment QA program. Therefore the guidance should include some specific
examples of a biological sampling QA program, including cost estimates. The QA
Plan developed for biological data by EMAP may provide relevant examples.
Second, it is critically important to ensure consistent quality in the
taxonomic identifications in all field biological assessment programs. We agree
with the statement on page 63 of the draft guidance that "a major factor
influencing the comparability of field ecological projects is the type and intensity of
appropriate training and professional experience for all personnel." The frequent
reference in the guidance to the use of "volunteer" or "lay" personnel, however,
seems to contradict the focus on adequate training. We urge that references to
the use of lay personnel be removed from the guidance, or that a full discussion be
added on the impacts thsit using nonprofessional personnel will have on the
QA/QC of data collected.
Lastly, we agree with the need to tailor the depth of the QA program to the
objectives of the bioassessment program. We recommend that the guidance offer
suggestions on tiers of effort, as can be found in the EPA's Good Laboratory
Practices (GLP) program,, which would allow QA/QC Plans to be customized
according to the individual needs of the state programs. For example,
bioassessment programs that will support significant economic and regulatory
decisions may need a more rigorous QA program than ones used as an assessment
or screening tool.
5.3 Taxonomy and Reference Collections
The guidance needs further discussion of the training requirements for
personnel conducting field sampling, and guidelines for sample identification and
handling. Successful implementation of the biocriteria program is dependent upon
13
-------
accurate taxonomic identification. Personnel involved in field sampling and work-
up of collections should have training in the systematics of the group(s) for which
they are responsible. It is unlikely that resource agency personnel with skills hi
identifying macroinvertebrates or fish will have competence in identifying birds by
visual or auditory cues, or mammals by the common signs of tracks and scat.
(The statement on page 51 of the draft document that "tracks and droppings also
provide easily attainable survey data" is misleading; in fact, very few field staff can
reliably identify these mammal signs).
Although field sorting and observations are important to record behavioral
and nonstructural color pattern characters that are lost after preservation, major
groups (aquatic plants, periphyton, invertebrates and fish) should generally be
identified in the laboratory, following field collection. Exceptions might include
some game fish, many mammal species, and most birds. Collection of protected
soecies (threatened, endangered, or otherwise protected) of vertebrates where field
identification is possible should be avoided and conducted only with appropriate
collecting permits. Many protected plant species would have to be properly
collected for identification by botanists (again, only with appropriate permits).
The guidance should also recommend varying levels of taxonomic
identification for different levels of intensity (different applications of the
biosurvey results), for example:
Level 1-Screening Assessment Protocol: field identification to the generic
level for fish and familial or generic level for macroinvertebrates, as in the
rapid bioassessment protocol for streams (Plafkin et al., 1989).
Level 2-Site Assessment Protocol: identification of juvenile/adult fish to
species, and benthic macroinvertebrates at least to the species level
taxonomic levels indicated in Table 1.
In addition, the guidance should include recommendations on reasonable
levels of taxonomic resolution for various categories of organisms. Field collection
and systematics should be driven by protocols, established in the peer-reviewed
literature, that are realistic. That is, the taxonomic resolution should be set at a
level achievable by appropriately trained state personnel. Stream ecology research
over the last decade indicates that a specific minimal level of resolution should be
set ("lowest achievable taxonomic level" is not a helpful criterion) and that
additional refinement should be left to individual state groups as their capabilities
permit. Proposed levels of intensity and taxonomic resolution must receive a
thorough evaluation by the scientific research community.
14
-------
In most cases, adult and juvenile fish should be identifiable to species.
While identification of larval fish may also provide useful information, it may only
be feasible to identify to the generic or familial levels. Reasonable candidate levels
for stream macroinvertebrates are given in Table-1.
Table 1. Propiosed Minimal Levels of Taxonomic Resolution for
Stream Macroinvertebrates
Taxonomic Level
Genus
Tribe
Subfamily
Family
Order
Groups
Plecoptera (in part), Ephemeroptera, Odonata,
Trichoptera, Megaloptera, Neuroptera,
Lepidoptera, Coleoptera (in part, larvae and
adults), Hemiptera, Diptera (Tipulidae and
Simuliidae), Crustacea, Mollusca *
Chironominae
Chironomidae
Diptera (other than Tipulidae and Simulidae),
Oligochaeta, Plecoptera (in part), Coleoptera
(in part)
Other non-insect groups
The discussion of sample collection and processing should be expanded. Aa
noted in the guidance, all specimens of a single species from each collection area,
from a single date, should be assigned a unique field number. The number should
be placed inside the container with the specimens, rather than being written or
attached to the outside of the container. Information for each collection should be
recorded in a field notebook containing, at a minimum, the following: field
number, date, collectors, site location (state, county, township, range, section, and
a verbal description of sampling location), sampling gear used, time of sample, and
habitat characteristics.
Once the samples have been analyzed (identifications, enumerations,
measurements, etc.), reference (voucher) material should be placed in the well
established network of federal, state and university museums for regionally
centralized curation. This ensures a second level of quality control in terms of
specimen identification. Preferably, collection and identification of voucher
specimens would be coordinated with taxonomic experts in regional museums.
These repositories, which have always been the centers for systematics, should'
15
-------
continue to be used for this function. Funds must be made available to these
institutions to insure their participation in the process. The Agency should work
with the U.S. Department of Interior as the new National Biological Survey is
established to ensure that samples collected by state biocriteria programs are
deposited in established museum repositories.
Once the information on the samples has been entered into a data base and
verified, the repository institutions should be encouraged to conduct additional
systematic studies on the material. Information from these additional analyses can
then be made available to state biocriteria programs.
16
-------
6. CHARACTERIZING BIOLOGICAL COMMUNITIES: MULTIMETRIC
APPROACH
6.1 Use and Misuse of Indices
We agree that the use of multiple metrics in biocriteria will ensure a more
balanced assessment of biotic integrity than any single measure. However, we
caution the Agency about the loss of information that can result from aggregation
of metrics, and the possible misuse of indices. One danger is the use of
judgmental labels to characterize different value ranges for a metric or index. For
example, the long-needed relationship between nutrient loading and trophic state
of lakes provided by Vollenweider (1966) was grossly misused because of his use of
the terms "permissible" and "dangerous." Because of this terminology, arguments
ensued both in and out of court over whether additional nutrient loading would be
"permissible" (i.e., would not cause a body of water to become eutrophic), or would
be "dangerous" (i.e., would lead to a eutrophic condition). In fact, all nutrient
additions contribute to eutrophication, and there is no threshold effect.
A second possible misuse of indices is for users to presume too high a
precision. For example, the Carlson Trophic Scale Index (TSI) for lakes (Carlson,
1977) was originally conceived as a 1-10 scale, but was converted to a 10-100+
scale when it was pointed out that an index of 5.4 implied a greater degree of
precision than 54 (i.e., once a decimal is introduced, people suspect greater
precision than may exist).
A third misuse can occur when individual metrics are aggregated into a
single index value. Metrics derived for different assemblages (e.g., fish,
macroinvertebrates and periphyton) may respond in different ways to the same
stressor. Therefore, important changes in individual metrics may be masked by
aggregation into a single index. For these reasons, we recommend that the
guidance discourage attempts to develop a single aggregate biocriterion for any
particular site.
6.2 Selection of Metrical
In the discussion on selecting assemblages to measure, the draft guidance
document appears to focus primarily on fish and benthic macroinvertebrate
community structure and function as indicators of water quality. The case studies
should emphasize the importance of including the periphyton and decomposer
communities as well. For example, the potential impact of nutrient enrichment, a
major problem affecting streams (as well as lakes/reservoirs and estuaries), may be
17
-------
underestimated if the periphyton community is not considered. In order to assess
the temporal and spatial effects of nutrient enrichment all trophic levels should be
evaluated. In addition, the physical-chemical features of the habitat <£.. substrate
type, flow, depth, stream order, shading, water hardness alkalinity pH,
temperature, conductivity, and dissolved oxygen) must also be considered when
interpreting metrics or indices.
6.3 The Role of Biomarkers
The draft guidance document refers to the potential value of biomarkers
when comparing the ecological health of potentially impacted sites with that of
reference areas. Biomarkers, in this context, are defined as biochemical
physiological or histological markers of anthropogenic stress. We agree that
biomarkers (e.g., protein induction or tissue lesions in finfish) may provide useful
data to augment more classical measures of ecosystem health such as diversity,
species richness, IBI, etc. However, care must be exercised to ensure that
biomarker assay results are not misinterpreted or misused. The current state ot
the science is not advanced enough that biomarkers can be routinely used in
biocriteria, or as a diagnostic tool.
Biomarkers can be divided into two broad categories: those that reflect or
indicate exposure to stress, and those that denote an effect of exposure(s). For
Estence, Eduction of the enzyme P450IA1 in the liver of a finfish can be mdic^ve
of exposure to polynuclear aromatic hydrocarbons, PCB, etc. However, induction
of PXKATAI does not necessarily mean that the organism has been adversely
impacted. On the other hand, biomarkers of immune function, such as
macrophage phagocytosis, may indicate a harmful biological response to
STthropogemc stress since a reduced phagocytic activity relates, at some level, to
the animal's ability to cope with infectious particles, etc.
We encourage the Agency to pursue, with appropriate caution, the feasibility
of using biomarkers in the establishment of biocriteria. Additional research will
likely enhance the utility of biomarkers in the biocriteria program.
18
-------
7. TRAINING AND TECHNICAL ASSISTANCE
In order for the implementation of biocriteria programs to be successful, the
Agency must play an active role in providing training and technical assistance to
state agencies. At the current time, there is a wide spectrum of abilities and
expertise in the state agencies that will be responsible for developing biocriteria.
In recognition of this fact, the Agency should prepare a reference document on
how to: a) assess the educational needs of state agency staff; b) develop in-house
educational programs; and, c) fund outside educators (universities, consultants, or
other agencies) for training programs. This companion document could also be a
source book of museums, reference collections, and people with scientific expertise
relevant to specific taxonomic or ecological needs. Without some education
component and source of information and help for resource managers and
regulators, development and implementation of biocriteria could overwhelm state
agency technical resources.
We also recommend that the Agency consider the following steps to support
states in the development of biocriteria programs:
a) continue development of the BIOS and STORET data management
systems to provide centralized storage and access for state biological
data;
b) encourage states to use the existing repository network of academic,
state and federal museums for curation of type specimens of
macroinvertebrates and fish;
c) establish a network list of taxonomic experts that could "verify"
identifications (these experts should be readily available to confirm
identifications, and therefore an equitable fee should be established
for appropriate level of effort, i.e., graduated fee depending upon
degree of difficulty in identifying species); and,
d) provide workshops to assist training of state personnel.
19
-------
8. CONCLUSIONS AND RECOMMENDATIONS
We compliment the Agency on the quality of this initial guidance on
developing biocriteria for streams and small rivers. As the latest in a series of
documents from the Office of Water on this topic, the document clearly shows the
continuing maturity of the Agency's approach to assessment of biological integrity
as a measure of the effectiveness of water quality regulation. Our response to the
four questions posed by the Agency in the charge to the Subcommittee are as
follows:
a) Are the options presented for selecting reference conditions
scientifically sound?
Yes, although the document may overemphasize the role of reference
sites as the primary basis for defining the reference condition. The
final document should include greater detail on the use of other
approaches (historical data, empirical models, and expert
opinion/consensus) for affirming reference condition based on
reference sites, as well as for establishing the reference condition in
degraded areas.
b) Does the recommended approach to sampling frequency adequately
account for seasonal variability?
No. Because of the dynamic nature of streams, characterization of
both reference and test sites should be based on sampling at various
times of the year to include the major components of the fall-winter
and spring-summer (or wet season-dry season) communities.
c) Will the recommended multimetric approach to evaluating aquatic
communities adequately characterize the resource?
Yes. Assuming that the appropriate assemblages and metrics are
chosen, we agree that the multimetric approach is an appropriate way
to characterize the variety of responses at the individual, population,
and community level.
d) Does the recommended approach to selection and aggregation of
biological information provide a sound basis for the development of
biocriteria?
20
-------
Yes. However, we discourage the use of a. single aggregate criterion
for any site, since important changes in individual metrics may be
masked by aggregation into a single index.
In addition to the questions in the charge, we have the following
conclusions and recommendations:
a) The Agency should ensure an adequate, consistent level of taxonomic
training among various state biocriteria programs. Within the
academic community and state and federal agencies, there exists a
wealth of expertise and data on the ecology and systematics of lotic
organisms (e.g., there are 200 inland field stations in the U.S. and 15
LTER sitess, to name a few). We strongly believe that the primary
role of the Agency should not be to recreate this expertise within the
EPA, but rather to ensure that this expertise is made available to the
states through collaboration with existing centers of expertise in
stream ecology, taxonomy and systematics. This collaboration may
take the form of support for systematic training in universities and
museums, on-site state training workshops, expert networks to verify
specimen identifications, etc.
b) Development of Diagnostic Techniques: current state of the science is
sufficient to support application of biocriteria for assessment of site-
specific impacts and regional trends, but not for establishing point
source criteria or permit limits. In order for biocriteria to be more
broadly useful, the Agency should support the development of
diagnostic tools to differentiate probable causes of observed changes
in biological metrics. This may include research on biomarkers and
indicator species.
c) The final guidance should strengthen the linkages, both conceptual
and procedural, between the Ecorisk Framework and the application
of biocriteria.
d) The Agency should refine existing classification schemes to include
biogeographic and distributional patterns of stream and riparian biota.
e) The final guidance should include several real case studies of the
application of biocriteria on a watershed scale, including cost
estimates for development and implementation of the program.
21
-------
f) Chemical characterization should be required at all sites, both
reference and test locations.
g) Reference areas, once designated, should be managed in cooperation
with other government agencies to protect the baseline reference
condition. One option for communication among states and academic
institutions would be an Agency-sponsored Register of Biocriteria
Reference Sites.
h) The final guidance should place greater emphasis on statistical design
of the bioassessment sampling program, rather than focusing
primarily on post-monitoring data analysis. Important issues include
identifying the universe to be sampled, defining the level of desired
precision, and determining sampling sufficiency.
i) The final guidance should suggest reasonable levels of taxonomic
resolution for various assemblages that a) are achievable by
appropriately trained personnel, and b) provide an adequate level of
information for the intended applications of the biosurvey results.
j) The Agency should strongly encourage states to use the established
network of federal, state and university museums for regionally
centralized curation of voucher specimens.
22
-------
9. REFERENCES CITED
Brower, J.E., J. H. Zar, and C. N. von Ende. 1990. Field and laboratory methods
for general ecology. Wm. C. Brown, Dubuque, Iowa. 237 p.
Carlson, R. 1977. A trophic state index for lakes. Limnol. and Oceanog. 2g:361.
Cummins, K. W., M. A. Wilzbach, D. M. Gates, J. B. Perry, and W. B. Taliaferro.
1989. Shredders and Riparian Vegetation; BioScience 39(1):24-30.
Elliott, J. M. 1977. Some methods for the statistical analysis of samples of
benthic invertebrates. Sci. Publ. No. 25, Freshw. Biol. Assoc., U.K.
Green, R. H. 1979. Sampling design and statistical methods for environmental
biologists. John Wiley and Sons, New York. 257 p.
Hurlbert, S. H. 1984. Pseudoreplication and the design of ecological field
experiments. Ecol. Monographs 54:187-211.
Ludwig, J. A., and J. F. Reynolds. 1988. Statistical Ecology. John Wiley and
Sons, New York. 337 p.
McElravy, E. P., G. A. Lamberti, and V. H. Resh. 1989. Year to year variation in
the aquatic macroinvertebrate fauna of a Northern California stream. J.
North Amer. Bentholog. Soc. 8:51-63.
Omernik, J. M. 1987. Ecoregions of the Coterminous United States. Ann. Assoc.
Amer. Geogr. 7701:118-125.
Plafkin, J. L., M. T. Barbour, K. D. Porter, S. K. Gross, and R. M. Hughes. 1989.
Rapid Bioassessment Protocols for Use in Streams and Rivers: Benthic
Macroinvertebrates and Fish. EPA/444/4-89-001.
Statzrer, B. and B. Hijler. 1986. Stream hydraulics as a major determinant of
benthic invertebrate zonation patterns. Freshwater Biol. 16:127-139.
U.S. Environmental Protection Agency. 1990. Biological Criteria: National
Program Guidance for Surface Waters. EPA-440/5-90-004.
U.S. Environmental Protection Agency. 1992a. Procedures for Initiating
Narrative Biological Criteria. EPA-822-B-92-002.
23
-------
U.S. Environmental Protection Agency. 1992b. Framework for Ecological Risk
Assessment. EPA/630/R-92/001.
Vannote, R. L., G. W. Minshall, K. W. Cummins, J. R. Sedell, and C. E. Gushing.
1980. The river continuum concept. Can. J. Fish. Aquafc. Sci. 37:130-137.
Vollenweider, R. A. 1968. Scientific fundamentals of the eutrophication of lakes
and flowing waters, with particular reference to nitrogen and phosphorus as
factors in eutrophication. Tech. Rpt. to OECD, Paris. DAS/CSI/68:27, 192 p.
mimeo.
Waters, W. E. and D. C. Erman. 1990. Research methods: concept and design.
In, C. B. Schreck and P. B. Moyle (eds.). Methods for fish biology. Amer.
Fish. Soc., Bethesda, MD 684 p.
24
-------
DISTRIBUTION LIST
Administrator
Deputy Administrator
Assistant Administrators
Deputy Assistant Administrator for Research and Development
Deputy Assistant Administrator for Water
EPA Regional Administrators
EPA Laboratory Directors!
EPA Regional Libraries
EPA Laboratory Libraries
-------
------- |