?>55 R 96,(00/
BIOLOGICAL CRITERIA
Technical Guidance for
Streams and Small Rivers
Revised Edition
Project Leader and Editor
Dr. George R. Gibson, Jr.
U.S. Environmental Protection Agency
Office of Science and Technology
Health and Ecological Criteria Division
401 M Street, SW (4304)
Washington, DC 20460
Principal Authors
Dr. Michael T. Barbour, Principal Scientist
Dr. James B. Stribling, Senior Scientist
Dr. Jeroen Gerritsen, Principal Scientist
Tetra Tech, Inc.
10045 Red Run Boulevard, Suite 110
Owings Mill, MD 21117
Dr. James R. Karr, Director
Institute for Environmental Studies .
Engineering Annex FM-12
University of Washington
Seattle, WA 98195
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Prepared by JT&A, inc., and Abt Associates for the U.S. Environ-
mental Protection Agency. Points of view expressed in this publica-
tion do not necessarily reflect the views or policies of the U.S.
Environmental Protection Agency, nor does mention of trade names
or commercial products constitute an endorsement or recommenda-
tion for their use.
Address comments or suggestions related to this document to
Dr. George R. Gibson, Jr.
U.S. Environmental Protection Agency
Office of Science and Technology
Health and Ecological Criteria Division
401 M Street, SW (4304)
Washington, DC 20460
/'/'
-------
Acknowledgments
Dr. George Gibson of the Office of Science and Technology's Health
and Ecological Criteria Division is project leader and main editor of
this document whose principal authors are consultants Drs. Michael Bar-
bour, James Stribling, Jeroen Gerritsen, and James Karr.
Dr. Phil Larsen of the U.S. Environmental Protection Agency's Envi-
ronmental Research Laboratory in Corvallis, Oregon; and Dr. David Cour-
temanch of the Department of Environmental Protection in Augusta,
Maine, also provided valuable insights and wrote portions of the docu-
ment. Staff from several program offices in the Office of Water provided
expert advice and made comments on the text, and Rachel Reeder of
JT&A,,inc., helped weave the text with its multiple contributions into a
more cogent document.
Many others also contributed to the writing of this document and de-
serve special thanks: first and foremost, the Streams Biocriteria Work-
group. The Workgroup, composed of state and EPA biologists, members of
academic institutions, and other consultants, helped provide the frame-
work for the basic approach and served as primary reviewers of the
manuscript. Next, our special thanks to those scientists who responded to
our request for peer review and to the members of the Ecological Proc-
esses and Effects Committee of the Science Advisory Board (SAB), who
also reviewed the manuscript and prepared an insightful critique. We sin-
cerely appreciate the contribution of their valuable time and constructive
advice. Their comments have greatly improved the final document.
Streams Biocriteria Workgroup
• George R. Gibson, Ph.D., Workgroup Chair, U.S. EPA Health and Ecological
Criteria Division
• Michael Barbour, Ph.D., Tetra Tech, Inc.
• Edward Berider, Ph.D., U.S. EPA Science Advisory Board
• Lawrence Douglas, Ph.D., University of Maryland
• Chris Faulkner, U.S. EPA Assessment and Watershed Protection Division
• James Karr, Ph.D., University of Washington, Institute for Environmental Studies
• D. Phil Larsen, Ph.D., U.S. EPA Environmental Research Laboratory, Corvallis
• James Lazorchak, U.S. EPA Environmental Monitoring Systems Laboratory,
Cincinnati
• Dave Penrose, North Carolina DEM, Environmental Services Laboratory
• James O. Peterson, Ph.D., University of Wisconsin
• Ron Preston, U.S. EPA Region 3, Wheeling Division
• Stephanie Sanzone, U.S. EPA Science Advisory Board
• Christopher Zarba, U.S. EPA Health and Ecological Criteria Division
iii
-------
Intentionally Blank Page
-------
Contents
Acknowledgments iii
List of Figures viii
List of Tables xi
CHAPTER 1: Introduction 1
The Concept of Biocriteria 2
Applications of Biocriteria 3
The Development, Validation, and Implementation
Process for Biocriteria . 4
Characteristics of Effective Biocriteria 9
Examples of Biocriteria 10
Narrative Biological Criteria 10
Numeric Biological Criteria 11
Other Biocriteria Reference Docqments 12
Suggested Readings 13
CHAPTER 2: Components of Biocriteria 15
Conceptual Framework and Theory 15
Components of Biological Integrity 16
Assessing Biological Integrity 18
Complex Nature of Anthropogenic Impacts 19
i
The Biocriteria Development Process 21
Suggested Readings 25
CHAPTER 3: The Reference Condition 27
Establishing the Reference Condition 27
The Use of Reference Sites 29
Characterizing Reference Conditions 32
Classification 32
Framework for Preliminary Classification 33
. Site Selection 39
Confirming Reference Conditions — Successful Classifications 41
Suggested Readings 44
CHAPTER 4: Conducting the Biosurvey 45
Quality Assurance Planning 46
Quality Management. . 47
Biocriteria Program Structure, Personnel, and Resources 47
Quality Control Elements in an Ecological Study 50
Data Quality Objectives 54
Study Design ; 55
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Biosurveys of Targeted Assemblages, 56
Attributes of Selected Assemblages ............. 56
Synthesis 59
Technical Issues 60
Selection of the Proper Sampling Periods 61
Selection of Habitat for Aquatic Assemblage Evaluations 67
Standardization of Techniques 72
Sample Collection 72
Sample Processing 73
Suggested Readings 74
CHAPTER S: Evaluating Environmental Effects — 77
Water Quality 77
Habitat Structure 81
Habitat Quality and Biological Condition 82
Development of Habitat Assessment Approach 83
Flow Regime 85
Energy Source 88
Biotic Interactions 90
Cumulative Impacts ............. 90
Suggested Readings 91
CHAPTER 6: Multlmetric Approaches for Biocriterla Development 93
Metric Evaluation and Calibration 94
Biocriteria Based on a Multimetric Approach 97
Potential Metrics for Fish and Macroinvertebrates 102
Index Development 106
Multivariate Approaches 109
Suggested Readings 109
CHAPTER 7: Biocriteria Development and Implementation. 111
Establishing Regional Biocriteria 111
Designing the Actual Criterion 112
Biocriteria for Significantly Impacted Areas 114
Selecting the Assessment Site 114
Evaluating the Assessment Site 116
Overview of Selected State Biocriteria Programs ,119
Costs for State Programs Developing Bioassessments and Biocriteria...... 124
Value of Biocriteria in Assessing Impairment 128
Suggested Readings 132
CHAPTER 8: Applications of the Biocriteria Process 133
Stream Characterization and Classification 133
Case Study — North Carolina 133
Refining Aquatic Life Uses 135
Judging Use Impairment 136
Case Study — Ohio 137
Diagnosing Impairment Causes 138
vi
-------
Case Study — Delaware 139
Problem Identification 141
Case Study — Maine 141
Other Applications of the Process 142
Suggested Readings —v.. 144
Contacts for Case Studies 144
Glossary.,.. 145
References 151
4
t
vii
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
List of Figures
Figure 1-1.—Model for biocriteria development and application 6
Figure 2-1.—Conceptual model showing the interrelationships of the primary
variables relative to the integrity of an aquatic biota. External refers to
features outside the stream system; internal to in-stream features (Karr,
1991). Terrestrial environment includes factors such as geology, topography,
soil, and vegetation 20
Figure 2-2.—Organizational structure of the attributes that should be
incorporated into biological assessments 21
Figure 3-1.—Approach to establishing reference conditions 30
Figure 3-2.—Reciprocal averaging ordination of sites by fish species in the
Calapooia River watershed, Oregon. The inset shows the correspondence
between fish assemblages in the rivers and ecoreglons 37
Figure 3-3.—Generalized box-and-whisker plots illustrating percentiles and
the detection coefficient of metrics 41
Figure 3-4.—Index of Biotic Integrity at Ohio reference sites 43
Figure 3-5.—Fish species richness as a function of the log of watershed
area. Bars to right indicate range of observations before regression and
range of residuals after regression. Residuals have smaller variance than
the original observations 43
Figure 4-1.—Organization chart illustrating project organization and lines of
responsibility. 50
Figure 4-2.—Summary of Data Quality Objective (DQO) process for
ecological studies (taken from Barbour and Thornley, 1990) 54
Figure 4-3.—Classification of U.S. elimatological regions 63
Figure 4-4.—Biological and hydrological factors for sampling period
selection in the Northeast (macroinvertebrates). The gray area is the overlap
between emergence and recruitment ; 65
Figure 4-5.—Biological and hydrological factors for sampling period
selection in the Northeast (fish) 66
Figure 5-1.—Five major classes of environmental factors that affect aquatic
¦ biota in lotic systems. Right column lists selected expected results of
anthropogenic perturbation (Karr et al. 1986) 78
Figure 5-2.—Decision matrix for application of rapid bioassessments in
Arkansas for permitted point source discharges (Shackleford, 1988). 80
Figure 5-3—Qualitative Habitat Evaluation Index (QHEI) versus Index of
Biotic Integrity (IBI) for 465 relatively unimpacted and habitat modified Ohio
stream sites (Rankin, 1991) 83
Figure 5-4.—Choptank and Chester rivers tributaries (Primrose et al. 1991) 83
viii
-------
Figure 5-5.—Relationship of the index of Biotic Integrity (IBI) to changes in
the quality of habitat structure through the Qualitative Habitat Evaluation
Index (QHEI) in channelized (triangles) and unchannelized (circles) (Ohio
EPA, 1990)
. 86
Figure 5-6.—Diagrammatic representation of the stream continuum to
illustrate variation in trophic structure of benthic invertebrates (adapted from
Cummins, 1983)
. 89
Figure 5-7.—Biological community response as portrayed by the Index of
Biotic Integrity (IBI) in four similarly sized Ohio rivers with different types of
point and nonpoint source impacts (Yoder, 1991)
. 91
Figure 6-1 a.—Metrics that decrease with impairment
• ?4
Figure 6-1 b.—Metrics that increase with impairment
. 95
Figure 6-2.—Total number of fish species versus stream order for 72 sites
along the Embarras River in Illinois (Fausch et al. 1984)
. 96
Figure 6-3.—Metrics plotted with a continuous covariate (hypothetical
example). :
. 96
Figure 6-4.—Box and whisker plots of metric values from hypothetical
stream classes. Shaded portions are above the median for each class. The
box represents a percentile, the vertical line is 1.5 times the interquartile
range, and the horizontal line is the median of each distribution
. 97
Figure 6-5a—Site discrimination for the number of Ephemeroptera,
Piecoptera, and Trichoptera (EPT index) in Florida streams. (Reference =
least impaired, other - unknown, impaired - determined impaired a priori.)
. 98
Figure 6-5b.—Site discrimination for the number of Chironomidae taxa in
Florida streams. (Reference = least impaired, other = unknown, impaired =
determined impaired a. priori.)
. 98
Figure 6-6.—Tiered metric development process (adapted from Holland,
1990)
. 99
Figure 6-7.—The conceptual process for proceeding from measurements to
indicators to assessment condition (modified from Paulsen et al. 1991)
100
Figure 6-8.—Invertebrate stream index scores for Florida streams
108
Figure 7-1.—Hierarchy of statistical models used in Maine's biological
criteria program (taken from Davies et al. 1993)
113
*
Figure 7-2.—The process for proceeding from measurements of fish
assemblage to indicators such as the Index of Biotic Integrity (IBI) or Index
of Well Being (IWB) — as used to develop criteria and apply those criteria to
streams (modified from Paulsen et al. 1991)
118
Figure 7-3a—Biological criteria in the Ohio WQS for the Warmwater Habitat
(WWH) and Exceptional Warmwater Habitat (EWH) use designations
arranged by biological index, site type for fish, and ecoregion (Ohio EPA,
1992)
124
Figure 7-3b.—Biological criteria in the Ohio WQS for the Modified
Warmwater Habitat (MWH) use designation arranged by biological index,
site type for fish, modification type, and ecoregion (Ohio EPA, 1992)
125
Figure 7-4.—Comparison of ambient toxicity and fish richness surveys at
eight sites in various parts of the United States (taken from U.S. EPA, 1991). ...
, 129
. /X
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Figure 7-5.—Comparison of effluent toxicity of receiving water impact using
Ceriodaphnia dubia chronic toxicity tests and freshwater receiving stream
benthic invertebrates at 43 point source discharging sites in North Carolina
(taken from U.S. EPA, 1991). 130
Figure 7-6.—Comparison of chemical criteria exceedances and biosurvey
results at 645 stream segments in Ohio 130
Figure 7-7.—Assessment of nontidal stream aquatic life use attainment in
Delaware (taken from the state's 395[b] report, 1994) 131
Figure 8-1.—EPT Index (number of taxa of Ephemeroptera, Plecoptera, and
Trichoptera) for two locations on the South Fork of the New River, North
Carolina 134
Figure 8-2.—Examples from some states using biological assessments to
determine aquatic life use support in rivers and streams. Failure to sustain
fish and aquatic life is defined with respect to the reference condition in that
state 136
Figure 8-3—Temporal trends in the improvement of the Upper Hocking River
1982-1990 138
Figure 8-4.—Assessment summary, Kent and Sussex counties, Delaware,
1991 140
Figure 8-5.—State of Delaware 1994 305(b) report, aquatic life use
attainment — all nontidal streams 140
Figure 8-6.—Macroinvertebrates in the Piscataquis River, Maine, 1984 -
1990 143
X
-------
List of Tables
Table 2-1.— Components of biological integrity (modified from Karr, 1990) 17
Table 3-1.— Hierarchical classification of stream riparian habitats (from
Minshall, 1993; after Frissell et al. 1986) 36
Table 4-1.— Quality control elements integral to activities in an ecological
study in sequence 51
Table 4-2.— Common benthic habitats 70
Table 4-3.— Proposed minimal levels of taxonomic resplution for stream
macroinvertebrates (taken from Sci. Advis. Board, 1993) 74
Table 5-1.— Parameters that may be useful in evaluating environmental
conditions and their relationship to geographic scales and the environmental
factors influenced by human actions . 82
Table 6-1.— Sequential progression of the biocriteria process 101
Table 6-2.— Index of Biotic Integrity metrics used in various regions of North
America 103
Table 6-3.— Examples of metric suites used for analysis of
macroinvertebrate assemblages 104
Table 6-4.— Index of Biotic Integrity metrics and scoring criteria based on
fish community data from more than 300 reference sites throughout Ohio
applicable only to boat (i.e., nonwadable) sites. Table modified from Ohio
EPA (1987) 107
Table 6-5.— Ranges for Index of Biological Integrity values representing
different narrative descriptions of fish assemblage condition in Ohio streams.
Site category descriptions — wading, boat, and headwaters — indicate the
type of site and style of sampling done at those sites. Modified from Ohio
EPA (1987) 108
Table 7-1.— Sequential process for assessment of test sites and
determination of the relationship to established biocriteria 117
Table 7-2.— Maine's water quality classification system for rivers and
streams, with associated biological standards (taken from Davies et al.
1993) 120
Table 7-3.— Bioclassification criteria scores for EPT taxa richness values for
three North Carolina ecoregions based on two sampling methods 122
Table 7-4.— The investment of state water resource agency staff to develop
bioassessment programs as a framework for biocriteria 128
Table 7-5.— Costs associated with retaining consultants to develop
bioassessment programs as a framework for biocriteria. Dash indicates work
done by state employees or information not available; FTE costs for
contractors and state employees are not equivalent 128
xi
-------
Intentionally Blank Page
-------
CHAPTER 1.
Introduction
The goal of this document is to help states develop and use biocriteria
for streams and small rivers. The document includes a general strat-
egy for biocriteria development, identifies steps in the process, and pro-
vides technical guidance on how to complete each step, using the
experience and knowledge of existing state, regional, and national surface
water programs.
This guidance document is designed primarily for water resource
managers and biologists familiar with standard biological survey tech-
niques and similarly familiar with the EPA guidance document "Rapid
Bioassessment Protocols for Use in Streams and Rivers: Benthic Macroin-
vertebrates and Fish" (Plafkin et al. 1989). It should be used in conjunction
with that earlier text.
The biosurvey-biocriteria process provides a way to measure the con-
dition of a water resource, that is, its attainment or nonattainment of bio-
logical integrity. In turn, biological integrity is a conceptual definition of
the most robust aquatic community to be expected in a natural condition
— in a water resource unimpaired by human activities. Thus, biological
criteria are the benchmarks for water resource protection and manage-
ment; they reflect the closest possible attainment of biological integrity. It
follows that any criterion representing less than achievable biological in-
tegrity is an interim criterion only, since the use of biocriteria are intended
to improve the nation's water resources.
The guidance in this document is designed so that users may tailor the
methods to their particular biocriteria development needs. Chapters 1 and
8 are inclusive of the methodology — at different levels of complexity —
while chapters 2 through 7 explore the process step by step. Thus, the
document is organized as follows:
¦ Chapter 1: Introduction. An overview of the process.
¦ Chapter 2: Components of Biocriteria. An exploration of the basic re-
lationship between biological integrity and biocriteria, the complex
nature of human disturbances, and the definition of biological ex-
pectations.
¦ Chapter 3: The Reference Condition. Selection of reference sites and
the role of the reference condition in biocriteria development.
Purpose:
To provide conceptual
. guidance on
how and when to
use the biosurvey-
biocriteria process to
evaluate streams and
small rivers.
1
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
¦ Chapter 4: Conducting the Biosurvey. An investigation of the de-
sign, management, and technical issues related to biocriteria-bio-
assessment programs, the various biosurvey methods and their
standardization.
¦ Chapter 5: Evaluating Environmental Effects. Factors that affect
water resource integrity.
¦ Chapter 6: Multimetric Assessment Approaches for Biocriteria De-
velopment. Emphasis on the community composition element of
biological surveys.
¦ Chapter 7: Biocriteria Development and Implementation. Designing
and developing biocriteria from the data and precautions for some
site selections.
¦ Chapter 8: Applications of the Biosurvey-Biocriteria Process. Case
Studies from North Carolina, Ohio, Delaware, and Maine.
Each chapter concludes with a list of readings containing supplemen-
tal information on the specific topic treated in that chapter. An extensive
glossary and full reference list appear at the end of the document. Future
documents will be oriented to other waterbody types: lakes and reser-
voirs, rivers, estuaries near coastal marine waters, and wetlands.
The Concept of Biocriteria
Early efforts to monitor human effects on waterbodies in the 18th century
were limited to physical observations of sediment and debris movement
resulting from land settlement, and commercial activities (Caper et al.
1983). Later, as analytical methods became increasingly available for meas-
uring microchemical conditions in the waterbody (Gibson, 1992), chemical
measurements became the most commonly employed source of water
quality criteria. However, investigators and resource managers have long
recognized that such water column measurements reflect conditions only
at the time of sampling.
To understand fully the effects of human activities on water resources,
biological sampling is an important supplement to chemical sampling.
Biological measurements reflect current conditions as well as temporal
changes in waterbodies, including the cumulative effects of successive dis-
turbances.
Three aspects of water resource management (chemical, physical, and
biological) are recognized in the National Clean Water Act as amended by
the Water Quality Act of 1987 (U.S. Gov. Print. Off. 1988). Section 101a
states that the Act's primary objective is to "restore and maintain the
chemical, physical, and biological integrity of the nation's waters."
The development and widespread use of formal biological criteria
(biocriteria) has lagged behind chemical-specific, in-stream flow, or toxic-
ity-based water quality criteria in waterbody management (U.S. Environ.
Prot. Agency, 1985a,b; 1986). Biological criteria are numeric values or nar-
rative expressions that describe the preferred biological condition of
aquatic communities based on designated reference sites. The conditions
of aquatic life found at these reference sites are used to help detect both
the causes and levels of risk to biological integrity at other sites in the
Biocriteria are
developed from
expectations for the
region or watershed,
site-specific
applications, and
consensus definitions
by regional experts.
The biological
sampling for this
process requires
minimally impaired
reference sites
against which the
study area may be
compared.
2
-------
CHAPTER 1: '
Introduction
same region. In keeping with the policy of not degrading the resource, the
reference conditions — like the criteria — are expected to be upgraded
with each improvement to the water resource. Thus, biocriteria contribute
directly to water management programs, and recent recommendations
(U.S. Environ. Prot. Agency, 1987a,b) on monitoring strategies for aquatic
resources have emphasized the need to accelerate the development of bio-
logical sampling as a regular part of surface water programs.
Biocriteria are developed from expectations for the region or water-
shed, site-specific applications, and consensus definitions by regional
authorities. The biological sampling for this process requires minimally
impaired reference sites against which the study area may be compared.
Minimally impaired sites are not necessarily pristine; they must, however,
exhibit minimal disturbance (i.e., human interference) relative to the over-
all region of study.
Applications of Biocriteria
Biocriteria applications are presented in some detail in chapter eight. Here,
a brief description of these applications is sufficient to demonstrate the
usefulness of the concept.
¦ Aquatic Life Designated Uses. The States and Tribes together with EPA
identify the most appropriate uses of our water resources and then man-
age or restore these waters accordingly. Some aquatic life uses are cold
water fisheries, warm water fisheries, unique natural systems, and sys-
tems including rare or endangered species. Biological assessments and
subsequent criteria are essential to the development and refinement of
these designations and the management necessary to support them.
¦ Problem Identification. Biological surveys and their comparison to es-
tablished biological criteria, in addition to traditional chemical and physi-
cal investigations, often provide insights into problems not otherwise
identifiable. For example, new compounds or synergistic reactions be-
tween existing waterborne chemicals may affect the biota even though in-
dividual chemical tests show no rise in historic concentrations; hydrologic
modifications such as installed impoundments may restrict species distri-
bution and recruitment; increased watershed sealed surfaces may change
flow regimes, cause more scouring, and destroy habitat for essential com-
munity assemblages.
¦ Regulatory Assessments. Much of the work done by EPA is regulatory
in nature and involves the use of permits to regulate the discharge of vari-
ous substances into the waters. The Agency does not require the use of
biocriteria as numeric regulatory limits in National Pollution Discharge
Elimination system (NPDES) permits. It does, however, strongly recom-
mend that states develop and use biocriteria as a permit assessment tool
and as a mechanism for evaluating the success of pollution control efforts.
Concurrence of biotic data with established biocriteria can be a key meas-
ure of permit effectiveness and of regulatory compliance.
¦ Management Planning. Water resource managers can use the relative
relationships of a series of similar streams, as ranked by their compliance
with biocriteria, as a means of assigning priorities to their management ef-
3
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
forts. In this way budgets and manpower can be applied most effectively
because the manager is better informed about the most pressing problems
and about those streams most likely to respond to restorative efforts.
¦ Water Quality Project Evaluations. The measurement of the resident
stream biota before, during, and after implementation of pollution man-
agement efforts is an excellent way to evaluate the success or failure of
those techniques.
¦ Status and Trends of Water Resources. As states and tribes gather more
biological data in support of their biocriteria, their knowledge of the wa-
ters becomes more refined. The condition of the nation's waters will be
better understood and the direction of change in the various regions will
be more evident and better addressed.
To achieve these objectives for the use of biocriteria, EPA is evaluating
not only the role of biocriteria in the permit process but also the inde-
pendent application of various criteria to determine water resource qual-
ity. Presently chemical, physical, and biological criteria — when used in a
regulatory context — are applied to a waterbody independently. Compli-
ance or lack of compliance with one criterion does not influence the appli-
cation of another. As biological and other types of criteria, such as
sediment criteria (now being investigated) are more widely implemented
in state programs, the Agency will continue to investigate the usefulness
of weight of evidence approaches as an alternative.
Thus, biocriteria expand aquatic life use designations and improve
water quality standards, help identify impairment of beneficial uses, and
help set program priorities. Biological surveys (or biosurveys) in conjunc-
tion with biocriteria are valuable because they provide
• a direct measure of the condition of the water resource at the site,
• early detection of problems that other methods may miss or
underestimate,
• a systematic process for measuring the effectiveness of water resource
management programs,
• an evaluation of the adequacy of permits, and
• a measurement of the status and trends of streams over time and space.
The Development, Validation, and
Implementation Process for Biocriteria
Three processes are part of the overall implementation plan to incorporate
biocriteria into the surface water programs of regulatory agencies: the de-
velopment of biocriteria and associated biological survey methods, the
validation of the reference condition and survey techniques, and the im-
plementation of the program at various sites within watersheds with sub-
sequent determinations of impairment.
The development of biocriteria by regulatory agencies partly depends
on bioassessment to evaluate or compare ecosystem conditions. Bioassess-
Biocriteria expand
aquatic life use
designations and
improve water quality
standards, help
identify impairment of
beneficial uses and
help set program
priorities.
4
-------
CHAPTER 1:
Introduction
ment contains two types of data: toxicity tests and field biological surveys
of surface waters. Toxicity tests are described elsewhere (U.S. Environ.
Prot. Agency, 1985a,b; 1988; 1989) and are not the subject of this document.
The use of bioassessments to investigate potential impairment, evalu-
ate the severity of problems, ascertain the causes of the problems, and de-
termine appropriate remedial action is a step-by-step process.
Inherent in the process for implementation of biocriteria is the as-
sumption that bioassessment methods have been developed. However,
the actual development of biocriteria is the most difficult step in the whole
process. A conceptual model for biocriteria development was presented by
the U.S. Environmental Protection Agency (1990) to streamline the major
elements in the process. This model has been refined for presentation here
(Fig. 1-1).
Each component of the model is numbered so that it can be identified
and discussed more easily as an important part of the biocriteria develop-
ment process. Nevertheless, these steps are not sequential. The following
paragraphs describe the model process in more detail and identify areas of
simultaneous development.
Components 1 through 8 describe the development of biocriteria, prior
to their use in regulatory programs.
1. Investigate the Biocriteria Program Concept. The biocriteria proc-
ess involves the selection of several program elements that contrib-
ute to effective biocriteria. Each state agency will have its own
program objectives and agenda for establishing biocriteria; how-
ever, the underlying characteristics for effective biocriteria will be
the same in all states.
2. Formulate the Biocriteria Approach. Defining biological integrity
is the first step in the formulation of a biocriteria program. The ac-
tivities important to this step are planning the biocriteria process;
designating the reference condition; performing the biosurveys;
and establishing the biocriteria.
3. Select Reference Sites or Conditions. The attainable biological
status of an aquatic system is primarily described by the reference
condition. If we understand the water resources's biological poten-
tial, we can judge the quality of communities at various sites rela-
tive to their potential quality. Natural environmental variation
contributes to a range in expected conditions; deviations from this
range help to distinguish perturbation effects.
Historical datasets existing from previous studies are also an
element of the derived biocriterion. These data range from hand-
written field notes to published journal articles; however, biologi-
cal surveys of present reference sites that are minimally impaired is
key to the defined reference condition.
The selection of reference "sites is key to the success of biocrite-
ria development. Various spatial scales can be used, but reference
conditions must be representative of the resource at risk and must,
therefore, be of the same or similar ecological realm or bio-
geographic region (i.e., an area characterized by a distinctive flora
or fauna).
The selection of .
reference sites is key
to the success of
biocriteria
development. Various
spatial scales can be
used, but reference
conditions must be.
representative of the
resource at risk and
must, therefore, be of
the same or similar
ecological realm or
biogeographic region.
5
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Reference
Discussion
Model
Ch.1
Ch.3
Ch.3
Ch.4
Ch.4
Ch.4
Ch.2
Ch. 2,3,5,7
Ch.4,5,6
Ch. 4,5,6, 7
Ch. 6,7
Ch. 5,6, 7
Ch. 7
Figure 1-1.—Model for blocrlteria development and application.
Establish Biocriteria
Address Technical Issues
Test Protocol Sensitivity
Modify/Refine Protocols
Develop Standard Protocols
Define Expected Conditions
Impaired Condition Detected
No Impaired Condition Detected
Diagnose Cause of Impairment
Formulate Biocriteria Approach
Implement Control and
Continued Monitoring
If Needed, Revise Approach
Based on Evaluation of Data
Evaluate Biocriteria Program Concept
No Action Required; Continue
Monitoring Recommended
Characterize Biological Integrity of
Reference Conditions from Database
Select Reference Sites and/or Condition
Appropriate to Targeted Assemblages
Evaluate both the Biological
and Physicochemical Data
Within an Ecological Contort
Conduct Biosuivays at Test Sites (Determine
Impairment Within the Revised Framework)
-------
CHAPTER 1:
Introduction
Candidate reference sites can be selected in a number of ways,
. but must meet some requirements established on the basis of over-
all habitat and minimally impaired status in a given region. The
reference condition is best described by including data collected
from several reference sites representing undisturbed watersheds.
Such biological information can be combined for a more accurate
assessment of the reference condition and its natural variability.
The reference condition approximates the definition of biological
integrity unless the reference sites were selected in significantly al-
tered systems.
4. Select Standard Protocols. The development of standard protocols
requires consensus building relative to the biological and ecological
endpoints of interest. The primary goal is to develop measures to
assess the biological integrity of aquatic communities in specified
habitats, that is, to assess the integrity of the aquatic community as
measured by the activities that maintain communities in equilib-
rium with the environment. There is no correct method to use or
biological assemblage to sample; rather, a number of possibilities
exist, including the Index of Biotic Integrity (IBI) for fish, and the
Rapid Bioassessment Protocols (RBPs) for benthos.
The process of applying these and other indices across widely
differing systems is not a straightforward "process and best profes-
sional judgment should be exercised before applying them to spe-
cific problems. For example, the IBI must be modified for
northwestern assemblages since it was developed in the Midwest
for midwestern assemblages. These indices measure a structural or
functional attribute of the biological assemblage that changes in
some predictable way with increased human influence. Combina-
tions of these attributes or metrics provide valuable synthetic as-
sessments of the status of water resources. As the basic theoretical
framework and approach should remain consistent, the use of
these indices should occur only after rigorous review arid evalu-
ation of their documentation. Such reviews are available in a vari-
ety of peer-reviewed publications.
5. Modification and Refinement of the Protocols. The refinement
process is an important step before large-scale biosurveys are con-
ducted. The sensitivity of the protocols should be tested to deter-
mine whether differences in community health resulting from
anthropogenic activities are discernible from changes caused by
other impacts or natural variation. An impact is any change in the
chemical, physical, or biological quality or condition of a water-
body caused by external sources. This process applies to all aspects
of the protocol from sampling to data analysis and may be re-
peated as often as necessary.
6. Address Technical Issues. Certain technical issues — for example,
natural seasonal variability, the aquatic assemblages selected for
evaluation, the procedure for selecting sampling sites, and the type of
sampling gear or equipment — affect the derivation of biocriteria.
The process of
applying indices
across widely
differing systems is
not a straightforward
process, and best
professional judgment
should be exercised
before applying them
to specific problems:
7
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
7. Characterize Biological Integrity. Analyze biological databases to
establish the range of values within the reference condition that
will characterize biological integrity. Characterization depends on
the use of biological surveys in concert with measurements of habi-
tat structure.
8. Establish Biocriteria and a Biological Monitoring Program. Once
biological integrity has been characterized and the geographic area
regionalized, biological information can be equated to the water
quality expectations of the state, and biocriteria can be established
for these regions. Biocriteria may vary within a state depending on
the region's ecological structure and the type of monitoring used in
its water quality programs. Sources for the derived biocriteria are
reference sites, historical records, in some instances empirical mod-
els of the systems (especially if significantly altered), and the con-
sensus of a representative panel of regional experts evaluating this
information.
Step 9 describes the validation of the biocriteria developed in the pre-
vious components.
9. Evaluate and Revise as Needed. Biocriteria are revised whenever
"better information is available, natural conditions have changed,
and/or the waters of interest have improved. This process includes
statistical analyses, of biological, physical, and chemical data to es-
tablish natural variability and the validity of existing biocriteria.
Regional frameworks should be adjusted if biological and geo-
graphical data support the need to do so. Reasons for these adjust-
ments and. the data used to determine them should be clearly
documented.
Steps 10 through 14 describe the use of biocriteria for water resource
management, that is, for the assessment, protection, remediation, and
regulation of water quality.
10. Conduct Biosurveys, Biosurveys conducted at test sites help to de-
termine whether and to what extent a site deviates from the nor-
mal range of values observed for the reference condition and from
the regional biocriteria. Candidate test sites are any locations along
the stream or river in which the conditions are not known but are
suspected of being adversely affected by anthropogenic influence.
11. Detect Impaired and Nonimpaired Conditions. Decisions on
whether adverse or impaired conditions exist must be made, but
whether these conditions are socially tolerable may be beyond sci-
ence. Scientists and resource managers are, however, obliged to de-
termine the relative impairment of the water resource as a
precondition for any subsequent decisions.
12. Review Other Data Sources for Additional Information. The use
of additional data to complement the biological assessment is im-
portant in the decision-making process. As part of an integrated
approach, whole effluent toxicity (WET) testing, chemical-specific
analyses, and physical characteristic measurements can be used to
make a comprehensive evaluation.
-------
13. Diagnose Causes of Impairment. Once impairment has been de-
termined, its probable causes must be identified before remedial
action can be considered and implemented. Probable "causes" may
include alteration of habitat structure, energy source, biological in-
teractions, flow characteristics, or water quality. The "source" of
the disturbance may be point or nonpoint source contamination or
other human activities. Thus, if impairment is detected, the data
should be evaluated to determine its probable causes; the site and
surrounding area should be investigated for other probable causes;
additional data should be collected; and either remedial action
should be formulated (if the actual causes have been determined)
or the investigation should be continued.
14. Implement Remedial Actions and Continue Monitoring. If prob-
able causes have been identified so that an action plan can be de-
veloped, the last step is. to begin remedial measures and continue
monitoring to assess the stream's recovery. This step can be used to
evaluate management programs and to determine cost-effective
methods. The relative success of the measures depends on the se-
lection of appropriate remedial actions to reduce or eliminate im-
pairments and to attain the designated uses that the biocriteria
protect.
If no impairment is found, no action is necessary except contin-
ued monitoring at some interval to ensure that the condition does
not change adversely.
Characteristics of Effective Biocriteria
Generally, effective biocriteria share several common characteristics. Well-
written biocriteria
• provide for scientifically sound evaluations,
• protect the most sensitive biota and habitats,
• protect healthy, natural aquatic communities,
• support and strive for protection of chemical, physical, and biological
integrity,
• include specific assemblage characteristics required for attainment of
designated uses,
• are clearly written and easily understood,
• adhere to the philosophy and policy of nondegradation of water
resource quality, and
• are defensible in a court of law.
In addition, well-written biocriteria are set at levels sensitive to an-
thropogenic impacts; they are not set so high that sites that have reached
their full potential cannot be rated as attaining, or so low that unaccept-
ably impaired sites receive passing scores. The establishment of formal
biocriteria warrants careful consideration of planning, management, and
regulatory goals and the best attainable condition at a site. Stringent crite-
9
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and.Small Rivers
The best balance is
achieved by
developing biocriteria
that closely represent
the natural biota,
protect against further
degradation, and
stimulate restoration
of degraded sites.
States may draft
general narrative
biological criteria
early in their program
— even before they
have designated
reference sites or
refined their approach
to biological surveys.
ria that are unlikely to be achieved serve little purpose. Similarly, biocrite-
ria that support a degraded biological condition defeat the intentions of
biocriteria development and the Clean Water Act. Balanced biocriteria will
incorporate multiple uses so that any conflicting uses are evaluated at the
outset. The best balance is achieved by developing biocriteria that closely
represent the natural biota, protect against further degradation, and
stimulate restoration of degraded sites.
Additional general guidance regarding the writing of biocriteria is pro-
vided in U.S. Environ. Prot. Agency (1990). Several kinds of biocriteria are
possible and vary among state programs. Both narrative and numeric
biocriteria have been effectively implemented. Both should be supported by
effective operational guidelines and adequate state resources, including
people, materials, methods, historical data, and management support.
Narrative biocriteria consist of statements such as "aquatic life as it
should naturally occur" or "changes in species composition may occur,
but structure and function of the aquatic community must be maintained."
An aquatic community, the association of interacting assemblages in a
given waterbody, is the biotic component of an ecosystem. Numeric val-
ues, such as measurements of community structure and function, can also
serve as biocriteria. The numeric criterion should be a defined range
rather than a single number to account for a measure's natural variability
in a healthy environment. It may also combine several such values in an
index. General examples of actual narrative and numeric biocriteria from
selected state programs, are presented in the following section; the infor-
mation was taken from Biological Criteria: State Development and Implemen-
tation Efforts (U.S. Environ. Prot. Agency, 1991a).
Examples of Biocriteria
Five states have adopted definitive biocriteria for water quality manage-
ment. Maine and North Carolina use narrative criteria; Ohio and Florida
have implemented combined narrative and numeric criteria. Delaware has
defined biocriteria for estuarine waters, and most other states have pro-
grams in various stages of development.
Narrative Biological Criteria
States may draft general narrative biological criteria early in their program
— even before they have designated reference sites or refined their ap-
proach to biological surveys. This haste does not mean that having refer-
ence sites and a refined system for conducting surveys is unimportant; it
means that a biocriteria program begins with writing into law a statement
of intent to protect and manage the water resources predicated on an ob-
jective or benchmark, for example, "aquatic life shall be as naturally oc-
curs."
When the objective to restore and protect the biological integrity of the
water resources has been formally mandated, then the operational mean-
ing of the statement and the identification of the agency responsible for
developing the necessary procedures and regulations can be stipulated as
the state's first steps toward the development of narrative and numeric
biological criteria. The key point is that natural or minimally impaired
water resource conditions become the criteria for judgment and manage-
10
-------
CHAPTER 1:
Introduction
ment. For more specific information on this concept and its implementa-
tion, see the EPA guidance document "Procedures for Initiating Narrative
Biological Criteria" (Gibson, 1992).
Narrative biological criteria form the legal and programmatic basis for
expanding biological surveys and assessments and for developing sub-
sequent numeric biological criteria.
Maine and North Carolina are examples of the practical development
and use of narrative biological criteria. Maine incorporated the general
statement "as naturally occurs" into its biocriteria, but also developed
supporting statements that specified collection methods to survey aquatic
life. Maine uses narrative biocriteria defined by specific ecological attrib-
utes, such as measures of taxonomic equality, numeric equality, and the
presence of specific pollution tolerant or intolerant species.
North Carolina uses narrative criteria to evaluate point and nonpoint
source pollution and to identify and protect aquatic use classifications. In
North Carolina, macroinvertebrate community attributes are used to help
define use classifications. These attributes include taxonomic richness and
the biotic indices of community functions and numbers of individuals.
They are also used in conjunction with narrative criteria to determine
"poor," "fair," "good-fair," "good," and "excellent" ratings for the desig-
nated uses.
Narrative biological criteria specify the use designations established
by the state and describe the type of water resource condition that repre-
sents the fulfillment of each use. Conversely, when adopted by the state
and approved by EPA, they become one of the standards by which water
resource violations are determined.
Nevertheless, narrative biological criteria cannot be fully implemented
without a quantitative database to support them. Quantitative data pro-
vide a responsible rationale for decision making and assure resource man-
agers a degree of confidence in their determinations. In fact, some states
have elected to develop narrative biocriteria and to use this legislative
mandate to establish administrative authority for their quantitative imple-
mentation in a state natural resources agency. In this manner, future im-
provements in scientific methods and indicators can be accommodated
through the administrate process rather than the more cumbersome and
expensive method of amending state laws.
These data are similar to the data used to formulate numeric biological
criteria; they can and should include the determination of reference condi-
tions and sites. Thus, when the survey process for narrative biocriteria is
well developed and refined, the program can easily begin the develop-
ment of numeric biocriteria. While not an essential precursor, the narrative
process is an excellent way for states to begin expanding their stream re-
source evaluation and management procedures to include more definitive
numeric biocriteria.
Numeric Biological Criteria
Although based on the same concept as narrative biocriteria, numeric
biocriteria include discrete quantitative values that summarize the status
of the biological community and describe the expected condition of this
system for different designated water resource uses.
Narrative biological
criteria cannot be fully
implemented without
a quantitative
database to support
them.
Numeric biocriteria
include discrete
quantitative values
that summarize the
status of the
biological community
and describe the
expected condition of
this system for
different designated
water resource uses.
11
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
The key distinction between narrative biocriteria supported by a quan-
titative database and numeric biocriteria is the direct inclusion of a spe-
cific value or index in the numeric criteria. This index allows a level of
specification to water resource evaluations and regulations not common to
narrative criteria.
To develop numeric criteria, the resident biota are sampled at mini-
mally impaired sites to establish reference conditions. Attributes of the bi-
ota, such as species richness, presence or absence of indicator taxa, and
distribution of trophic groups, help establish the normal range of the bio-
logical community as it would exist in unimpaired systems.
Ohio combines narrative and numeric biocriteria and uses fish and in-
vertebrates in its stream and river evaluation programs. Its numeric
biocriteria are defined by fish community measurements, such as the In-
dex of Weil-Being (IWB) and the Index of Biotic Integrity (IBI). Ohio also
employs an Invertebrate Community Index (ICI). All three measures pro-
vide discrete numeric values that can be used as biocriteria.
Ohio's numeric criteria for use designations in warmwater habitats are
based on multiple measures of fish and benthic macroinvertebrates in dif-
ferent reference sites within the same ecoregion. Macroinvertebrates are
animals without backbones that are large enough to by seen by the un-
aided eye and caught in a U.S. Standard No. 30 sieve. Criteria for this use
designation are set at the 25th percentile of each biological index score re-
corded from the established reference sites within the ecoregion. Excep-
tional warmwater habitat criteria are set at the 75th percentile from the
statewide set of reference sites (Ohio Environ. Prot. Agency, 1987). Use of
the 25th and 75th percentiles, respectively, portrays the minimum biologi-
cal community performance described by the narrative use designations.
Such applications require an extensive database and multiple reference ar-
eas across the stream and river sizes represented within each ecoregion.
To develop the most broadly applicable numeric biological criteria,
careful assessments of biota in multiple reference sites should be con-
ducted (Hughes et al. 1986). The status of the biota in surface waters may
be assessed in numerous ways. No single index or measure is universally
recognized as free from bias. Evaluating the strengths and weaknesses of
different assessment approaches is important, and a multimetric approach
that incorporates information on species richness, trophic composition,
abundance or biomass, and organism condition is recommended (see
Chapter 6).
Other Biocriteria Reference Documents
EPA has developed several program and technical guidance documents
for implementing biocriteria beginning with a preliminary discussion of
biocriteria program development issues: legislative authority, steps in de-
veloping biocriteria, and the application of biocriteria to surface water
management (U.S. Environ. Prot. Agency, 1990).
A survey of existing state programs was conducted in 1990 to deline-
ate the status of bioassessment implementation on a national basis (U.S.
Environ. Prot. Agency, 1991a). In addition,-a reference guide to the techni-
cal literature pertaining to biocriteria has been developed (U.S. Environ.
12
-------
CHAPTER 1:
Introduction
Prot. Agency, 1991b). The latter contains cross-references to technical pa-
pers that develop the concepts, approaches, and procedures necessary to
implement habitat assessment and biological surveys in the development
and use of biocriteria. In December 1990, a symposium on biological crite-
ria provided a forum for discussing technical issues and guidance for the
various surface waterbody types. The proceedings from this conference
are presented in U.S. Environ. Prot, Agency (1991d). Most recently, the
agency has developed guidance to help states initiate narrative biological
criteria (Gibson, 1992).
Suggested Readings
Gibson, George. 1992. Procedures for Initiating Narrative Biological Criteria, EPA-822-B-
92-002. U.S. Environ. Prot. Agency, Washington, DC.
U.S. Environmental Protection Agency. 1987a. Report of the National Workshop on In-
stream Biological Monitoring and Criteria. In-stream Biol. Criteria Comm. Reg. 5,
Environ. Res. Lab., Off. Water Reg. Stand., Corvallis, OR.
, 1987b. Surface Water Monitoring: A Framework for Change, Off. Water, Off. Pol.
Plann. Eval,, Washington, DC.
. 1991a. Biological Criteria: State Development and Implementation Efforts. EPA
440/5-91-003. Off. Water, Washington, DC.
—. 1991b. Biological Criteria: Guide to Technical Literature. EPA 440/5-91-004. Off.
Water, Washington, DC.
—. 1991c. Technical Support Document for Water Quality-based Toxics Control.
EPA 505/2-90-001. Off. Water, Washington, DC,
—, 1991d, Biological Criteria: Research and Regulation. EPA 440/5-91-005. Off.
Water, Washington, DC.
To develop numeric
biocriteria, the
resident biota are
sampled at minimally
impaired sites to
establish reference
conditions. Attributes
of the biota such as
species richness,
presence or absence
of indicator taxa, and
distribution of trophic
groups are useful for
establishing the
normal range of
biological community
components as they
would exist in
unimpaired systems.
1
-------
Intentionally Blank Page
-------
CHAPTER 2.
Components of
Biocriteria
Water resource legislation is usually designed to protect the resource
and to ensure its availability to present and future generations.
Over the past two decades, legislative and regulatory programs have es-
tablished goals such as "fishable and swimmable, antidegradation, no net
loss, and zero discharge of pollutants." However, actions to meet those
goals do not always accomplish the mandate of the Clean Water Act,
which is to restore and maintain biological integrity. The purpose of this
chapter is to provide managers with a basic conceptual understanding of
the relationship between biological integrity and biocriteria and to de-
scribe more fully the biocriteria process.
Conceptual Framework and Theory
Biological integrity was first explicitly included in water resource legisla-
tion in the Water Pollution Control Act Amendments of 1972 (Pub. L. 92-
500); and the concept, which was retained in subsequent revisions of that
act, is now an integral component of water resource programs at state and
federal levels (U.S. Environ. Prot. Agency, 1990).
The goal of biological integrity, unlike fishable and swimmable goals,
encompasses all factors affecting the ecosystem. Karr and Dudley (1981;
following Frey [1975]) define biological integrity as "the capability of sup-
porting and maintaining a balanced, integrated, adaptive community of
organisms having a species composition, diversity, and functional organi-
zation comparable to that of the natural habitat of the region." That is, a
site with high biological integrity will have had little or no influence from
human society.
Edwards and Ryder (1990) recently used the phrase "harmonic com-
munity" in a similar context to describe the goal of restoring ecological
health to the Laurentian Great Lakes. The sum of balanced, integrated,
and adaptive chemical, physical, and biological data can be equated with
ecological integrity (Karr and Dudley, 1981). Such healthy ecological sys-
tems are more likely to withstand disturbances imposed by natural envi-
ronmental phenomena and the many disruptions induced by human
society. These systems require minimal external support from manage-
ment (Karr et al. 1986).
Purpose:
To provide managers
with a basic
conceptual
understanding of the
relationship between
biological integrity
and biocriteria, and to
describe more fully
the biocriteria
process.
15
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
It is important to
distinguish between
the attributes of
natural systems that
we intend to protect
(assessment
endpoints) and the
attributes that we can
measure
(measurement
endpoints). Success
in protecting
biological integrity
depends on the
development of
measurement
endpoints that are
highly correlated with
assessment endpoints.
The adjective "pristine" is often invoked in such discussions; however,
in the late 20th century, it is almost impossible to find an area that is com-
pletely untouched by human actions. Thus, the phrase "minimally im-
paired" is more appropriate than the word "pristine" for describing
conditions expected at sites exhibiting high biological integrity.
Degradation of water resources comes from pollution, which is de-
fined in the Clean Water Act of 1987 as "manmade or man-induced altera-
tion of the chemical, physical, biological, or radiological integrity of
water" (U.S. Gov. Print. Off. 1988). This comprehensive definition does not
limit societal concern to chemical contamination. It includes any human
action or result of human action that degrades water resources. Humans
may degrade or pollute water resources by chemical contamination or by
altering aquatic habitats; they may pollute by withdrawing water for irri-
gation, by overharvesting fish, or by introducing exotic species that alter
the resident aquatic biota. The biota of streams, rivers, lakes, and estuar-
ies, unlike other attributes of the water resource (e.g., water chemistry or
flow characteristics), are sensitive to all forms of pollution. Thus, the de-
velopment of biological criteria is essential to protect the integrity of water
resources.
Components of Biological Integrity
While these definitions of integrity establish broad biological goals to sup-
plement more narrowly defined chemical criteria, their use depends on
the development of rigorous biological criteria. The challenge is to define
biological integrity clearly, identify its components, and develop methods
to evaluate a water resource and its surrounding environment based on
these conditions.
Evaluating the elements or components of biological integrity will in-
volve direct or indirect evaluations of biotic attributes. Indirect evalu-
ations are appropriate if direct approaches are prohibitively expensive or
in other ways difficult to implement. It is important to distinguish be-
tween assessment and measurement endpoints. Attributes of natural sys-
tems that we intend to protect, for example, the health of a fish
population, are assessment endpoints; and attributes that we can measure,
for example, age and size classes of the fish population, are measurement
endpoints. Success in protecting biological integrity depends on the devel-
opment of measurement endpoints that are highly correlated with assess-
ment endpoints.
Important components of biotic integrity have been measured before.
Toxicologists have long recognized the importance of individual health in
evaluating the extent to which human actions have degraded a water re-
source, and ecologists have long used the kinds and relative abundances
of species as indicators of condition. More recently, and in many ways less
insightfully, theoretical measures of diversity have been used to assess
species richness, that is, to determine if the number of species or relative
abundances of species have been altered. Fish biologists, for example, use
a variety of measures to assess the health of populations of targeted spe-
cies, such as game fish. However, none of the attributes used in the past
are comprehensive enough to cover all components of biological integrity.
In recent years, a broader conceptual foundation has been developed
to convey the breadth of biotic integrity. The original Index of Biotic Integ-
16
-------
CHAPfEfi 2.
Components of Biocriteria
rity (IBI) consisted of 12 metrics or attributes in three major groups: spe-
cies richness and composition, trophic structure, fish abundance and con-
dition. Another way of describing biotic integrity contrasts the elements of
the biosphere with the processes but argues that both are essential to the
protection of biological integrity (Table 2-1). The most obvious elements
are the species of the biota, but additional critical elements include the
gene pool among those species, the assemblages, and landscapes.
Table 2-1.—Components of biological integrity.
ELEMENTS
PROCESSES
Genetics
Mutation, recombination
Individual
Metabolism, growth, reproduction
Population/species
Age specific birth and death rates
Evolution/speciation
Assemblage (community
Interspecific interactions
and ecosystem)
Energy flow
Landscape
Water cycle
Nutrient cycles
Population sources and sinks
Migration and dispersal
Modified from Karr, 1990.
Processes (or functional relationships) span the hierarchy of biological
organization from individuals (metabolism) to populations (reproduction,
recruitment, dispersal, speciation) and communities or ecosystems (nutri-
ent cycling, interspecific interactions, energy flow). For example, an im-
portant process in streams is an interaction of fish and mussels in which
larval stages of the mussel (glochidia) attach to fish gills, presumably to I
enhance dispersal and to avoid predation.
Other approaches are available, but the important issue is not which
classification is the best approach. Rather, efforts to assess biological integ-
rity must be broadly based to cover as many components as possible.
The challenge in implementing biocriteria is to develop reliable and
cost-effective ways to exploit the insight available through biological
analyses. It is not necessary to sample the entire biota. Rather, carefully se-
lected representative taxa should be sampled. The selection should com-
bine as many attributes as possible with precision and sampling efficiency,
but all elements and processes are not necessarily covered in standard bio-
logical sampling.
Recent efforts to develop such integrative approaches include Karr's
IBI later expanded to apply to a wide geographic area (Ohio Environ. Prot.
Agency, 1987; Lyons, 1992; Oberdorff and Hughes, 1992), and to taxa other
than fish, for example, benthic invertebrate assemblages (Ohio Environ.
Prot. Agency, 1987; Plafkin et al. 1989). The Nebraska Department of Envi-
ronmental Control (Bazata, 1991) has proposed indices that combine fish
and invertebrate metrics, and the Ohio Environ. Prot. Agency (1987) has
calculated several indices separately (fish and invertebrates) but uses
them in combination to determine use attainment status.
biological integrity
must be broadly
based to cover as
many components as
possible.
bib
° 17
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
The choice of
attributes to be
assessed and
measured is critical to
the success of any
biological monitoring
and criteria program.
The best approach
to assessing
biological integrity
seems to be an
integrative one that
combines assessment
of the extent to which
either the elements or
the processes of
biological integrity
have been altered;
that is, efforts to
protect biotic integrity
should include
evaluation of a broad
diversity of biological
attributes.
Assessing Biological Integrity
A sound monitoring program designed to assess biological integrity
should have several attributes. A firm conceptual foundation in ecological
principles is essential to a multidimensional assessment that incorporates
the several elements and processes of biotic integrity. The use of the con-
cept of a reference condition, a condition against which a site is evaluated,
is also important.
In addition, the general principles of sound project management or To-
tal Quality Management (TQM), such as Quality Assurance and Quality
Control, are as critical as the use of standard sampling protocols. Quality
assurance (QA) includes quality control functions and involves a totally in-
tegrated program for ensuring the reliability of monitoring and measure-
ment data; it is the process of reviewing and overseeing the planning,
implementation, and completion of environmental data collection activi-
ties. Its goal is to assure that the data provided are of the quality needed
and claimed.
Quality control (QC) refers to the routine application of procedures for
obtaining prescribed standards of performance during the monitoring and
measurements process; it focuses on the detailed technical activities
needed to achieve data of the quality specified by the Data Quality Objec-
tives (DQOs). Quality control is implemented at the laboratory or field
level. Finally, biological monitoring must go beyond the collection and
tabulation of high quality data to the creative analysis and synthesis of in-
formation about relevant biological attributes.
Numerous attributes of the biota have been used to assess the condi-
tion of water resources. Some are difficult and expensive to measure while
others are not. Some provide reliable evaluations of biological conditions
while others, perhaps because they are highly variable, are more difficult
to interpret. Thus, the choice of attributes to be measured and assessed is
critical to the success of any biological monitoring and criteria program.
Historically, most biological evaluations were designed to detect a nar-
row range of factors degrading water resources. For example, the biotic in-
dex (Chutter, 1972; Hilsenhoff, 1987) is designed to detect the influence of
oxygen demanding wastes ("organic pollution") or sedimentation, as is
the Saprobic Index developed early in this century (Kolkwitz and Mars-
son, 1908).
With increased understanding of the complexity of biological systems
and the complex influences of human society on those systems, more inte-
grative approaches for assessing biological integrity have been developed.
Some (Ulanowicz, 1990; Kay, 1990; Kay and Schneider, in press) advocate
the use of thermodynamics, while others concentrate on richness or diver-
sity (Wilhm and Dorris, 1968). The best approach seems to be an integra-
tive assessment of the extent to which either the elements or the processes
of biological integrity have been altered; that is, efforts to protect biotic in-
tegrity should include evaluation of a broad diversity of biological attrib-
utes.
Because the goal of biocriteria-bioassessment programs is to evaluate
water resource systems stressed by or potentially destroyed by human ac-
tion, the selection of the monitoring approach is critical. Indicators and
monitoring design should be structured so that the same monitoring data
18
-------
CHAPTER 2:
Components of Biocriteria
can serve a multitude of needs. This openness requires a reasonable level
of sophistication for long-term status and trends monitoring. The more
complicated the water resource problem, the larger the number of attrib-
utes that should be measured. Finally, programs to monitor the effects of
human activity on the environment should have especially broad perspec-
tives to ensure sensitivity to all forms of degradation.
Complex Nature of Anthropogenic Impacts
A number of human activities strain the integrity of water resource sys-
tems and the cumulative impacts of these actions create even greater com-
plexity. Thus, it is useful, perhaps even necessary, to develop an
organizational framework within which factors responsible for degrada-
tion in biotic integrity can be evaluated.
A major weakness of past approaches to protect water resources has
been a narrow focus on the factors responsible for degradation. Specifi-.
cally, past approaches focused on reducing the chemical contamination of
the water on the assumption that clean water would produce high quality
water resources. Overall, the determinants of the biological integrity of the
water resource are complex, and the simplistic approach of making water
cleaner, though important, is inadequate.
Biological monitoring and the use of biocriteria to assess biotic integ-
rity provides a more comprehensive evaluation of the status of the
resource. Such evaluations, enhance our ability to identify the factors re-
sponsible for degradation and to treat the problem in the most
cost-effective manner. Monitoring specific and ambient (background) con-
ditions offers unique opportunities to detect, analyze, and plan the
treatment of degraded resources.
Because human actions may impact a wider range of water resource
attributes than water chemistry alone, a broader framework is necessary
to identify and reverse the specific factors responsible for the degradation
of biotic integrity. Degradation may begin in an area of the watershed or
catchment that is external to the reference or test site simply because it is
often the result of human actions that alter the vegetative cover of the land
surface. These changes combined with the alteration of stream corridors
degrade the quality of water delivered to the stream channels and attack
the structure and dynamics of those channels and their adjacent riparian
environments.
Human activities at the site affect five primary classes of variables —
all of which may result in further degradation of water resources (Karr,
1991). These five internal variables should be placed in a larger context as
illustrated in Figure 2-1:
1. Water Quality: Temperature, turbidity, dissolved oxygen, acidity,
alkalinity, organic and inorganic chemicals, heavy metals, toxic
substances.
2. Habitat Structure: Substrate type, water depth and current veloc-
ity, spatial and temporal complexity of physical habitat.
3. Flow Regime: Water volume, temporal distribution of flows.
19
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Riparian Corridor
EXTERNAL
INTERNAL
Water
^ Quality
Weather/
Climate
Integrity
of Aquatic
Biota
Energy
Source
Flow
Regime
Terrestrial
Environment/
Land Use
Biotic
Interactions
Habitat
Structure
Figure 2-1,—Conceptual model showing the Interrelationships of the primary vari-
ables relative to the Integrity of aquatic biota. External refers to features outside the
stream system; Internal to in-stream features {Karr, 1991). Terrestrial environment In-
cludes factors such as geology, topography, soli, and vegetation.
4. Energy Source; Type, amount, and particle size of organic material
entering stream, seasonal pattern of energy availability.
5. Biotic Interactions: Competition, predation, disease, parasitism,
and mutualism.
From this conceptual framework, at least four components of the biota
should be evaluated: structure, composition, individual conditions, and
biological processes (Fig. 2-2). Sample attributes for each component in-
clude the following:
. Community Structure: Species richness, relative abundances,
including the extent to which one or a few species dominates.
¦ Taxonomic Composition: Identity of the species that make up the
biota.
¦ Individual Condition: Health status of individuals in selected
species.
¦ Biological Processes: Rates of biological activities across the
biological hierarchy (from genes to landscapes).
Comprehensive assessments of these attributes ensure that all the
components of biotic integrity are protected. For each component, one or
more attributes should be assessed.
Successful metrics represent the expression of the influence of human
activities on the resident biota. For example, the presence of a few hardy
species of fish in abundance may be a response to sewage in the waters.
As human disturbance increases, total species richness, the number of in-
tolerant species, and the number of trophic specialists usually decline,
while the number of trophic generalists increases. Generalists are organ-
isms that can use a broad range of habitat or food types. Exceptions exist:
for example, when coldwater streams are warmed, species richness in-
creases, although this process must be viewed as a degradation of the bi-
otic integrity of a coldwater system.
20
-------
CHAPTER 2:
Components of Biocriteria
PROCESSES
STRUCTURE
BIOLOGICAL ASSESSMENT
RELATIVE
ABUNDANCE
DISEASE
Figure 2-2.—Organizational structure of the attributes that should be incorporated
into biological assessments.
Use of biocriteria to evaluate and protect biotic integrity focuses di-
rectly on the condition of the resource. The development of biological
monitoring is driven by the need for rigorous standardized evaluations of
point and nonpoint source pollution and other circumstances in which up-
and downstream evaluations may be inappropriate. In short, development
of biocriteria is driven by the need for a comprehensive approach to the
study and remediation of human effects on water quality.
The Biocriteria Development Process
Biocriteria must be developed with a clear understanding of several im-
portant concepts. Foremost is the basic premise underlying biocriteria de-
velopment: understanding the condition of the biota in a given waterbody
provides a baseline for an integrative and sensitive measure of water qual-
ity. Biocriteria are operational narrative or numeric expressions that char-
acterize and, if properly used, protect biological integrity.
Biocriteria can be used to protect biological, integrity and to establish
an aquatic life use classification. Following the definition of biocriteria,
field surveys are conducted to determine whether particular sites meet the
biocriteria or whether they have been affected by human activity. This de-
termination is made by comparing the aquatic biota at potentially dis-
turbed sites with minimally impaired reference conditions. Natural events
Understanding the
condition of the biota
in a given waterbody
provides a baseline
for an integrative and
sensitive measure of
water quality.
21
-------
biological criteria
Technical Guidance (or Streams and Small Rivers
The basic premise,
that biota provide a
sensitive screening
tool for measuring the
condition of a water
resource, depends on
the assumption that
the greater the
anthropogenic impact
in a watershed, the
greater the
impairment of the
water resource.
Once defined,
biocriteria for a
stream or river will
describe the best
attainable condition.
not initiated by or exacerbated by human actions (e.g., fire, beavers) are
not considered disturbances in this sense.
The basic premise, that biota provide a sensitive screening tool for
measuring the condition of a water resource, depends on the assumption
that the greater the anthropogenic impact in a watershed, the greater the
impairment of the water resource. A corollary is that streams and rivers
not subject to anthropogenic impact contain natural communities of
aquatic organisms that reflect unimpaired conditions. These assumptions
provide the scientific basis for formulating hypotheses about impairments
— departures from the natural condition result from human disturbances.
Natural disturbances, such as floods or drought, may also affect the
aquatic biota as part, of normal ecological processes, and these responses
vary among ecoregions and stream sizes. For example, relatively stable
structure is characteristic of fish communities in the eastern United States
but stable fish communities in the Great Plains streams may reflect human
disturbance (Bramblett and Fausch, 1991). Molles and Dahm (1991) pro-
vide additional cautions on the need to consider natural events in inter-
preting data from biological systems. Thus, natural disturbances must be
considered, but they are not considered as impairments because they are
not the result of human activity.
Ideally, biocriteria are reflective of the natural biological integrity of the
particular region under study, that is, of the region as it would be had it not
become impaired. Depending on the refinement of the biosurvey method,
the degree of impairment can often be established as part of the biocriteria
development process. Once defined, biocriteria for a stream or river will de-
scribe the best attainable condition. The best attainable conditions represent
expected conditions and are directly compared to the observed conditions.
Each state needs to formulate appropriate definitive descriptors (i.e.,
biocriteria) for the aquatic organisms in its streams, and these descriptors or
biocriteria should support the state's designated use classifications or other
resource protection and management objectives.
Successful implementation of biocriteria requires a systematic pro-
gram to collect and evaluate complex scientific information and translate
that information into an effective planning tool to protect water resources.
This effort must be systematic as well as conceptually and scientifically
rigorous; it must also be logical and easily understood. The components of
a program to implement biocriteria may be divided in a variety of ways.
The four primary steps to develop and implement biocriteria are intro-
duced here and will be discussed in greater detail in later sections of this
document. The four steps are
1. planning the biocriteria development process,
2. designating the reference condition for biosurvey sites,
3. performing the biosurveys to characterize reference condition, and
4. establishing biocriteria based on reference biosurvey results.
Each step must be considered in the context of regulatory policy, the
scientific method, and the practical aspects of fieldwork involving biosur-
veys. Further, acceptable biocriteria for streams and rivers can be devel-
22
-------
CHAPTER 2:
Components of BiocrJferJs
oped in various ways. Therefore, biocriteria development should be based
on a set of flexible procedures derived from management, the regulatory
process, or both. When properly implemented, the procedures lead to self-
defined biocriteria that will protect the unique characteristics of streams
and rivers. When not properly implemented, water resources continue to
be degraded. Although the general concepts and procedures of biocriteria
development can be adapted to any stream or river, the development of
useful biocriteria requires individual planning for different waterbodies.
¦ Planning Biocriteria. Planning includes the classification of surface
water types and the definition of designated uses; however, the planning
process necessarily extends beyond stream and river use classification. To
be effective, planning must ensure that program objectives are clearly de-
fined and that the scientific information generated to meet program objec-
tives is appropriate for making environmental management decisions.
The planning phase assumes the interaction of environmental manag-
ers (staff involved in policy, budgeting, and resource management) and
technical staff (those involved in data collection and interpretation) to en-
sure that the environmental data to be collected are acceptable and meet
the state's needs. To facilitate interaction, a formal quality assurance and
quality control plan that includes the formulation of data quality objec-
tives should be included in the biocriteria development process. Complete
data quality objectives describe the decisions to be made, the data re-
quired and why, the calculations in which the data will be used, and time
and resource constraints. They are used to design data collection plans
and to specify levels of uncertainty. Levels of uncertainty pertain to the
confidence that decision makers can realistically have that collected data
will actually support particular conclusions.
Finally, interagency cooperation (within and among states) should be a
critical component of the planning process. Time spent on developing
good relations with other groups improves biocriteria and their use.
¦ Designating Reference Condition. Designating the reference condition
for biosurvey sites is the second major activity in biocriteria development.
This continuation of the planning process shifts attention to the specific
data needed to define the biotic conditions that would be expected to oc-
cur in the study stream in the absence of human impact. Issues requiring
consideration at this stage of the process include
¦ the database to be formed and evaluated (e.g., the taxonomic
assemblages or other biological attributes to be used to describe
biological condition);
¦ the habitat types to be included in the survey (e.g., runs, riffles,
pools, and snags);
¦ the type of reference conditions needed for the program or study
being formulated (e.g., regional, ecoregional, or site-specific);
¦ the geographical scale to which the biocriteria are applicable (e.g.,
specific river reach, watershed, ecoregion, or other parameters);
¦ the temporal scale for which biocriteria are being considered (e.g.,
seasonal, annual, or multiyear);
The development of
useful biocriteria
requires individual
planning for different
waterbodies.
Planning must
ensure that program ¦
objectives are clearly
defined and that the .
scientific information
generated to meet
program objectives is
appropriate for
making environmental
management
decisions.
Interagency
cooperation should
be a critical
component of the
planning process.
23
-------
Biological criteria
Technical Guidance for Streams and Small Rivers
Definition of the
reference condition is
a critical step in the
process.
¦ how habitat will be assessed to ensure comparability between the
reference condition and the habitat at the biosurvey site before
human impacts;
¦ parameters and methods of measurement; and
¦ how data from the biosurvey are to be evaluated.
Data management, analysis, and reporting requirements should also
be determined before any fieldwork is begun. Specific information dealing
with the designation of reference condition and biosurvey sites is pro-
vided in Chapter 3.
Because knowledge of biological communities and habitats surround-
ing the surface waters of the study region is essential to effective biological
monitoring, definition of the reference condition is a critical step in the
process. Careful designation of the reference condition can reduce the like-
lihood of problems and minimize the costs associated with fieldwork.
Knowledge of the reference condition may derive from historical data
or from pilot studies of local or regional sites that are relatively undis-
turbed. Macroinvertebrate and fish assemblage data have often been rou-
tinely collected by state fish and wildlife agencies, water quality agencies,
universities, and others responsible for stream management. Although
these historical databases are often overlooked in environmental evalu-
ations, they can be valuable sources of information. An estimation of bio-
logical integrity at a minimally impaired site may be accomplished by
reviewing existing data and publications for specific streams and rivers.
Fausch et al. (1984) developed fish species richness expectations for sev-
eral midwestern streams based on historical data sets. Obviously, the use-
fulness of historical data for establishing reference condition is dependent
on the original objective of the data collection effort, the collection meth-
ods, and the quality of the data. Even if historical data are inadequate for
direct use in designating the reference condition, they may provide sub-
stantial insight about preexisting conditions at the test or study sites.
¦ Performing Biosurveys. Performance of the actual biosurvey to charac-
terize the reference condition entails several activities. Often, a presurvey
(pilot study) is necessary to finalize the study plan and the actual logistics
of the fieldwork. Upon cojmpletion of the study plan, technical staff must
be fully briefed regarding the study's objectives, quality assurance and
quality control operations, and methods of data collection and summariza-
tion. At this point, the actual biosurvey may be performed. Biosurveys
may include routine local monitoring, sampling over wide geographic ar-
eas, or special case evaluations at one or a few sites.
¦ Establishing Biocriteria. After the biosurveys have been completed or
the historical data evaluated, the biological status of the reference condi-
tion is used to help define the biocriteria. Based on the results of the sur-
veys, some refinement of aquatic life use designations may be needed for
particular streams or rivers. After writing the biocriteria, they must un-
dergo final review and approval by each state and the EPA.
Certain attributes should be considered when drafting formal biocrite-
ria. Ideally, biocriteria should be readily understandable and scientifically
24
-------
CHAPTER 2:
Components of Biocriteria
and legally defensible. Further, they should be protective of the most sen-
sitive element of the biota included in the designated aquatic life use of
the stream or river and yet express an attainable condition.
Thus, biocriteria should be used in decision making, not only for rou-
tine management procedures but also for guiding resource policy determi-
nations. For those decisions to be robust, quality assurance programs must
ensure long-term database management, including data entry, manipula-
tion, and analysis.
Biocriteria provide an initial determination of impairment or attain-
ment. Their use may also help to determine sources and causes of degra-
dation when combined with survey information and knowledge of how
organisms react to different stresses (e.g., sight-feeding fish decline when
turbidity increases; tolerant species increase with nutrient enrichment-
anomalies of 40 to 60 percent occur only in the presence of complex toxic
effluents and impacts). These response signatures are vital to the success-
ful use of biocriteria to atta'in water resource protection.
The endpoint of water resource protection using biocriteria is broader
than clean water. The endpoint of biocriteria and water resource legisla-
tion is "to restore and maintain the physical, chemical, and biological in-
tegrity of the nation's waters."
Suggested Readings
Davies, S.P., L. Tsomides, D.L. Courtemanch, and F. Drummond. 1991. Biological Moni-
toring and Biocriteria Development. Prog. Sum. Maine Dep. Environ. Prot.,
Augusta, ME.
Gallant, A.L, et al. 1989. Regionalization as a Tool for Managing Environmental Re-
sources, EPA/600/3-89-060. U.S. Environ. Prot. Agency, Environ. Res. Lab., Corval-
Iis, OR.
Karr, J.R. 1991. Biological integrity: A long-neglected aspect of water resource manage-
ment. Ecol. Appl. 1:66-84.
North Carolina Department of Environmental Health and Natural Resources. 1990.
Standard Operating Procedures, Biological Monitoring. Environ. Sci. Branch, Eco-
systems Analysis Unit, Biol. Assess. Group, Div. Environ. Manage., Water Qual.
Sec., Raleigh, N.C.
Ohio Environmental Protection Agency. 1987. Biological Criteria for the Protection of
Aquatic Life. In The Role of Biological Data in Water Quality Assessment. Vol. 1.
Div. Water Qual. Monitor. Assess., Surface Water Sec., Columbus, OH.
. 1990. The Use of Biocriteria in the Ohio EPA Surface Water Monitoring and As-
sessment Program. Columbus, OH.
Plafkin, J.L. 1989.- Water quality-based controls and ecosystem recovery. Pages 87-96 in J.
Cairns Jr., ed. Rehabilitating Damaged Ecosystems. Vol. 2, CRC Press, Boca Raton, FL.
U.S. Environmental Protection Agency. 1990. Biological Criteria: National Program
Guidance for Surface Waters. EPA-440/5-90-004. Off. Water, Washington, DC.
Biocriteria should
be readily
understandable and*
scientifically and
legally defensible.
Further, they should
be protective of the
most sensitive
designated aquatic
life use of streams
and rive'rs and yet
express an attainable
condition.
The endpoint of
biocriteria and water
resource legislation is
"to restore and
maintain the physical,
chemical, and
biological integrity of
the nation's waters."
25
-------
Intentionally Blank Page
-------
CHAPTER 3.
The Reference Condition
The term biocriteria implies the notion of comparison to the highest at-
tainable condition. The reference condition establishes the basis for
making comparisons and for detecting use impairment; it should be appli-
cable to an individual waterbody, such as a stream segment, but also to
similar waterbodies on a regional scale. The reference condition is a criti-
cal element in the development of a biocriteria program.
Establishing the Reference Condition
Recognizing that absolutely pristine habitats do not exist (even the most re-
mote lakes and streams are subject to atmospheric deposition), "resource
managers must agree to accept sites at which minimal impacts exist or have
been achieved as the reference condition for a given region. Acceptable ref-
erence conditions will differ among geographic regions and states because
soil conditions, stream morphology, vegetation, and dominant land use dif-
fer between regions. In heavily agricultural, industrial-commercial, or
urbanized regions, undisturbed streams or reaches may not exist, and refer-
ence conditions may need to be determined based on that which is likely •
attainable, the historical record, or other methods of estimation.
Reference conditions can be established using a combination of meth-
ods — reference sites, historical data, simulation models, and expert con-
sensus.
¦ Historical Data. In some cases, data are available that describe past bio-
logical conditions in the region. Careful scrutiny and evaluation of these
data can be an important initial phase in the biocriteria development proc-
ess because they provide insight about the communities that have been or
can be achieved in various waterbody types. These records are usually
available in natural history museums, university collections, and some
agencies, such as state water resource and fish and wildlife departments;
however, some historical biological surveys were conducted at impaired
sites, used different sampling methods, were insufficiently documented,
or had objectives markedly different from biocriteria determination. Such
data would be of questionable value for establishing precise reference con-
ditions and should be used advisedly.
¦ Reference Sites. Reference sites refer to locations in similar waterbodies
and habitat types at which data can be collected for comparison with test
sites. Typical reference sites include sites that are upstream of point
Purpose:
To provide guidance
for defining biological
expectations based
on a reference
condition, and for
making comparisons
to test sites.
The reference
condition establishes
the basis for making '
comparisons and for
detecting use
impairment; it should
be applicable to an
individual waterbody,
such as a stream
segment, but also to
similar waterbodies
on a regional scale.
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Reference conditions
can be established
using a combination
of methods —
reference sites,
historical data,
simulation models,
and expert consensus.
sources; sites in nearby watersheds; sites that occur along gradients of im-
pact (near field/far field); and regional reference sites that may be applied
to a variety of test sites in a given area. Sites upstream of point sources
may or may not exhibit the quality of the overall reference condition.
However, their proximity to the site in question makes them a useful
qualifier for regional references, specifically in controversial situations.
Achieving biological conditions may be described through a statistical
evaluation that integrates biological attributes from a group of sites that
have the same characteristics and expectations. This approach can be used
to establish biological criteria for aquatic life uses and to test the prob-
ability that a particular test site has a biological community comparable to
that established group (Maine Dep. Environ. Prot. Agency, 1993).
¦ Simulation Models. Simulation models include mathematical models
(logical constructs following from first principles and assumptions), statis-
tical models (built from observed relationships between variables), or a
combination of the two. The complexity of mathematical models that can
predict reference conditions is potentially unlimited, but as complexity in-
creases, the costs will be higher and some of the model's predictive ability
will be lost (Peters, 1991). Thus, models that predict biological reference
conditions' should only be used as a last resort and with great caution be-
cause they may involve complex and untestable hypotheses (Peters, 1991;
Oreskes et al. 1994). Nevertheless, several models that predict water qual-
ity in rivers and reservoirs from first principles of physics and chemistry
have been quite successful (e.g., Kennedy and Walker, 1990). Mathematical
models to predict biological conditions have been less successful and, so
far, not very useful in an assessment or management context.
Statistical models can be fairly simple in formulation, such as the
Vollenweider model and the morphoedaptic index to predict trophic
status (Vollenweider, 1975; Vighi and Chiaudani, 1985). These models re-
quire a sufficiently large database to develop predictive relationships and,
in their current state of development, predict only nutrient conditions, not
the structure of biological communities
Hybrid models use both first principles and statistical relationships be-
tween variables. Hybrids are typically large simulation models intended to
predict the behavior of a stream over time; they are commonly used to pre-
dict water quality for management (Kennedy and Walker, 1990). Most exist-
ing models predict water quality variables such as chlorophyll a, nutrient
concentrations, Secchi depth, and oxygen demand. Inferring the composi-
tion of biological assemblages from predicted water quality would require
another model relating assemblages to stream water quality.
Model development for biological criteria is still rudimentary. How-
ever, as state databases expand, this tool will become more important and
will likely assume a growing role in establishing reference conditions.
¦ Expert Opinion/Consensus. When no candidate reference sites are ac-
ceptable, and models are deemed unreliable, then expert consensus is a
necessary alternative to establish reference expectations. Under such cir-
cumstances, the reference condition may be defined using expert opinion
based on sound ecological principles applicable to the region of interest.
Several skilled biologists and natural resource managers should be con-
vened for the assessment. Each of these experts should be familiar with
-------
CHAPTER 3:
The Reference Condition
the streams and aquatic biota of the region as they will be asked to de-
velop a description of the assemblage in relatively unimpacted streams based
on their collective experience. The description developed by consensus may
therefore be more qualitative than quantitative. Even when reference sites are
available and models may also be useful, this panel of specialists should be
convened to evaluate all the data and help develop the biocriteria.
In sum, investigators will incorporate any or all of these usually inter-
dependent techniques in the effort to establish reference conditions. That
is, historical data, reference sites, simulation models, and expert opin-
ion/consensus can and should be used mutually to support reference con-
dition decisions; however, the use of actual reference sites to establish
reference conditions is always important. Such sites represent achievable
goals, and they can be regularly monitored. Historical data and expert
opinion should also be used to make decisions regarding the selection of
these reference sites. Such a panel of experts can be reconvened to help es-
tablish the subsequent, and related, biological criteria. Simulation models
that incorporate historical data combined with expert opinion are the pri-
mary alternative to reference sites and may be most useful in the assess-
ment of significantly altered sites or waterbodies unique to the region
under study.
The most appropriate approach to establishing reference conditions is
to conduct a preliminary resource assessment to determine the feasibility
of using reference sites (Fig. 3-1). If reference sites are not acceptable, then
even greater reliance must be placed on the other elements, and some
form of simulation modeling may be the next best alternative. This situ-
ation would occur if no "natural" sites exist and if "minimally impaired
sites" are unacceptable. Biological attributes can be modeled from neigh-
boring regional site classes, expert consensus, and/or a composite of
"best" ecological (historical) data. Such models may be the only viable
means of examining significantly altered systems. The expectations de-
rived from these models may be regarded as hypothetical or temporary
until more realistic attainment goals can be developed.
Thus, the use of reference sites remains the best data source to estimate
present-day attainment conditions and is the basis for the emphasis on ref-
erence sites that follows. The selection of minimally disturbed sites from a
site class provides the most realistic basis for expecting that biological in-
tegrity can be attained. In this situation, the central tendency of the bio-
logical measure is a conservative estimate of the expected biological
condition. Some states, for example, Ohio and Florida, use a lower percen-
tile (25th percentile) as their threshold for attainment. When relatively few
sites are unimpaired and the sites are more than minimally disturbed, an
upper percentile from the range of biological values from all sites may
have to be used instead. An interim expected biological condition can be
developed from this approach that can be revisited after restoration efforts
have been initiated and evaluated by the specialists.
The Use of Reference Sites
The determination of the reference condition primarily from reference
sites is based on the premise that streams minimally affected by human ac-
tivity will exhibit biological conditions most natural and attainable for
The most .
appropriate approach
to establishing
reference conditions
is to conduct a
preliminary resource
assessment to
determine the
feasibility of using
reference sites.
The determination of
the reference
condition from
reference sites is
based, on the premise
that streams minimally'
affected by human
activity will exhibit
biological conditions
most natural and
attainable for streams
in the region.
29
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Two primary
considerations guide
the selection of
reference sites:
minimal impairment
and
representativeness.
Sites that are
undisturbed by
human activities are
ideal reference sites.
However, land use
practices and
atmospheric pollution
have so altered the
landscape and quality
of water resources
nationally that truly
undisturbed sites are
rarely available.
PRELIMINARY RESOURCE ASSESSMENT
Reference Sites
More Than
Minimally
Disturbed
Minimally
Disturbed
Reference
Sites
Acceptable
No Reference Sites
Where
"natural"
sites exist,
establish
expectations.
Central
Tendency
Biological
Integrity
Expectation
No "natural"
sites exist,
select best
available
(may require
sampling all
sites).
Upper Tail
Tendency
Interim
Expectation
Ecological
Modeling
No "natural"
sites exist,
select best
available
(may require
sampling all
sites).
Use (1) neighboring
site classes, (2) expert
consensus, or (3)
composite of "best"
pcnlnglnal information
Hypothetical
Expectation
Figure 3-1.—Approach to establishing reference conditions.
streams in the region. Anthropogenic effects include all possible human
influences, for example, watershed disturbances, habitat alteration, non-
point source runoff, point source discharges, atmospheric deposition, and
angling pressure. The premise does not consider any human activities as
improvements; for example, planting non-native riparian vegetation or
stocking with artificially high abundances of game or non-native fish are
not improvements relative to biological integrity. In practice, most refer-
ence sites will have some of these impacts; however, the selection of refer-
ence sites is made from those with the least anthropogenic influences.
Reference sites must be carefully selected because they will be used as
sources for the biocriteria benchmarks against which test sites will be com-
pared. The conditions at reference sites should represent the best range of
conditions that can be achieved by similar streams within a particular eco-
logical region. The key to making such biocriteria benchmarks protective
is to organize sites into classes so that the minimum acceptable perform-
ance is commensurate with the capability of the resource. Therefore, two
primary considerations guide the selection of reference sites within each
class: minimal impairment and representativeness.
30
-------
CHAPTER 3:
The Reference Condition
¦ Minimal Impairment. Sites that are undisturbed by human activities
are ideal reference sites. However, land use practices and atmospheric pol-
lution have so altered the landscape and quality of water resources nation-
ally, that truly undisturbed sites are rarely available. In fact, it can be
argued that no unimpaired sites exist. Therefore, a criterion of "minimally
impaired" must be used to determine the selection of reference sites. In re-
gions where even such minimally impaired sites are significantly de-
graded, the search for suitable sites should be extended over a wider area,
and multistate cooperation may be essential. The purpose of selecting
minimally impaired sites to represent reference conditions is primarily
goal-setting. Once attainment of these conditions is achieved on a large
scale, a higher criterion is possible. In no instance should any notably de-
graded condition be accepted as the reference for criteria development.
¦ Representativeness. Reference sites must be representative of the wa-
terbodies under investigation; that is, they must exhibit conditions similar
to those of other sites in the same region. Sites that contain locally unusual
environmental factors will result in uncharacteristic biological conditions
and should be avoided.
The overall goal in the establishment of the reiference condition from
carefully selected reference sites is to describe the biota that investigators
can expect to find at sites of interest. These "test or assessment sites" will
be compared to the reference sites to determine whether impairment ex-
ists. The characteristics of appropriate reference sites vary among regions
of the country and for different waterbody and habitat types. In general,
the following characteristics (modified from Hughes et al. 1986) are typical
of ideal reference sites:
• Extensive, natural, riparian vegetation representative of the region.
• Representative diversity of substrate materials (fines, gravel, cobbles,
boulders) appropriate to the region.
• Natural channel structures typical of the region (e.g., pools, riffles, runs,
backwaters, and glides).
• Natural hydrograph — in some cases, the flow patterns display large sea-
sonal differences in response to rainfall and snowmelt; in other cases, sta-
ble discharges are typical of water that originates from underground
sources. Biota evolve in the face of natural discharge patterns.
• Banks representative of undisturbed streams in the region (generally cov-
ered by riparian vegetation with little evidence of bank erosion, or under-
cut banks stabilized by root wads). Banks should provide cover for
aquatic biota.
• Natural color and odor — in some regions, clear, cold water is typical of
the waterbody types in the region; in others, the water is turbid or
stained.
• Presence of animals, such as piscivorous birds, mammals, amphibians,
and reptiles, that are representative of the region and derive some sup-
port from aquatic ecosystems.
A single minimally impaired site cannot be truly representative of an entire
region or population of sites, and a frequent difficulty is matching upstream
A criterion of
"minimally impaired"
must be used to
determine the
selection of reference
sites.
The overall goal in ,
the establishment of
the reference
condition from
carefully- selected
reference sites is to
describe the biota
that investigators may
expect to find at sites
of interest.
31
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
In developing and
adjusting the
biocriteria, managers
must strike a balance
between the ideal
restoration of the
water resource and
the fact that human
activity affects the
environment
and downstream habitats for valid comparison. For example, if habitat is
degraded upstream but not downstream, the effects of a discharge may be
masked. Reference conditions based on multiple sites are more repre-
sentative and form a valid basis for establishing quantitative biocriteria.
One problem in the use of minimally impaired sites as references is
what to do if an area is extensively degraded so that even these sites indi-
cate significant deterioration. Many systems are altered through channeli-
zation, urbanization, construction of dams and highways, or management
for certain sport fisheries or reservoirs (Karr and Dionne, 1991). The condi-
tion of these systems is a result of societal decisions that have to be taken
into account in the development of biocriteria, but these decisions should
not compromise the objective of defining the natural state. Biocriteria can
be qualified by the assignment of designated uses, but the reference condi-
tion should describe the site as one would expect to find it under natural
or minimally impaired conditions.
Although the biocriteria established for altered systems serve as a
baseline for judging further degradation, their ultimate goal is to achieve
the sites' recovery to the best attainable condition — as represented by
conditions at "minimally impaired" sites. Consensus of expert opinion
and historical data play an important role in characterizing the reference
condition for these systems, as does the application of innovative manage-
ment practices to obtain improvement.
• In developing and adjusting the biocriteria, managers must strike a
balance between the ideal restoration of the water resource and the fact
that human activity affects the environment. The most appropriate course
of action will use minimally impaired sites as the maximum amount of
degradation that will be tolerated, thereby ensuring adherence to the anti-
degradation policy of the Clean Water Act. Continual monitoring should
provide the feedback necessary to make reference site and criteria adjust-
ments as warranted during the restoration process.
Characterizing Reference Conditions
Characterization of regional reference conditions for biocriteria develop-
ment consists of the following steps:
1. Classification of the resource. All streams are not alike; therefore,
reference conditions (expectations) will differ among geographic
regions and stream types.
2. Selection of the best available sites in each resource class as candi-
date references.
3. Characterization — including confirmation and refinement of the
reference conditions — based on a biological survey of reference
sites.
Classification
The purpose of classification is to group similar things together, that is, to
prevent the comparison of apples and oranges. Meaningful classification is
not arbitrary (an apple is not an orange); professional judgment is usually
necessary to arrive at a workable system that recognizes different conditions,
32
-------
CHAPTER 3:
The Reference Condition
without considering each waterbody or watershed a special case. By class-
ifying, we reduce the complexity of biological information. Classification
improves the resolution or sensitivity of biological surveys to detect im-
pairment by partitioning or accounting for variation among sites.
There are two fundamental approaches to classification: a priori and a
posteriori (Conquest et al. 1994). A priori classification is a system based
on preconceived information and theories, for example, using physiog-
raphic provinces to classify streams. The a posteriori approach bases the
classification solely on the data collected and finds classes (e.g., using clus-
ter analysis) within these data.
In operational assessment and management of streams, an assessment
site is assigned to a class (e.g., mountain headwater streams) before it is ac-
tually surveyed and biological data are collected. Ideally, sites should be as-
signed to a class from mapped information before any sampling is done.
Therefore, an a priori classification based on maps or other easily obtainable
secondary information is often developed for characterizing reference con-
ditions. The biosurvey data are subsequently used to test that classification.
Stream characteristics that are readily affected by human activities or
occur as a biological response to physical or chemical conditions should
not be used as classification variables. Such responses may include land
use, habitat condition, or nutrient concentrations. For example, in the
southern Rockies ecoregion, riparian zones are heavily forested; and in the
neighboring Arizona/New Mexico Plateau ecoregion, riparian zones are
relatively unvegetated. The classification variable in this case is ecoregion,
and riparian vegetation is a response to ecoregion. If dense riparian vege-
tation were used as a classification variable, we would ran the risk of mis-
classifying an unimpaired, unvegetated stream in the Arizona/New
Mexico Plateau as impaired by comparison to natural streams in the
southern Rockies. This example shows that the best classification variables
are those that are readily obtained from maps or regional water charac-
teristics such as ecoregion, gradient, alkalinity, and hardness.
Framework for Preliminary Classification
The intent of this protocol is not to develop a classification scheme appli-
cable to the entire United States. Classification must be regional in scope
and use regional expertise to determine which variables to use in a given
region. Further, classification should be parsimonious to avoid prolifera-
tion of classes that do not contribute to assessment.
Ecoregions
Biologists have long noted that assemblages and communities can be
classified according to distinct geographical patterns (e.g., Wallace, 1869;
MacArthur, 1972). We observe areas of the country within which types of
resources and their attributes are ecologically consistent and similar when
compared to those of other areas. The recognition of such patterns occurs
at various levels: global, continental, regional, and local.
Regionalization identifies these natural spatial patterns. It accounts for
spatial variation by partitioning the landscape into smaller areas of greater
homogeneity. Ecological regionalization (as one type of regionalization) re-
sults in a map of ecological regions, or ecoregions. Such maps bring spatial
organization to ecological variability. They are useful in a variety of ways,
The intent of this
protocol is not to
develop a s
classification scheme
applicable to the'
entire United States.
Classification must be
regional in scope and
use regional expertise
to determine which
variables to use in a
given region.
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
The basic goal of
regionalization is to
depict areas of
ecological
homogeneity relative
to other areas.
One advantage of
having a consistent
framework is that
states that share the
same ecoregion can
cooperate across
political boundaries.
In times of limited
resources, such
cooperation makes
financial as well as
scientific sense.
for example, to summarize the condition of resources in a particular area,
to identify potential or achievable ecological conditions (e.g., regionally
achievable biocriteria), to characterize typical impact types and impair-
ments, to develop protective and remedial procedures that are tailored to
unique regional characteristics, and to present scenarios of realistically
achievable ecological conditions in particular regions (Gallant et al. 1989;
Hughes et al. 1990; Omernik and Gallant, 1990).
The basic goal of regionalization is to depict areas of ecological homoge-
neity relative to other areas. Fenneman (1946) defined physiographic prov-
inces within which the physical characteristics of the landscape, for
example, surface relief and slope, were homogenous relative to other areas.
Kuchler (1964) identified regions of similar potential natural vegetation.
Ecological regionalization should take into account all pertinent avail-
able information in the depiction of regions, at whatever scale the regions
are to be defined (Omernik, 1987). Primary categories of information used
in the process are (1) factors that control spatial patterns, such as climate,
topography, and mineral availability (soils, geology); and (2) factors that
respond to or integrate these controlling factors, such as vegetation and
land use. Both sets of categories and each factor within them must be
judged for their usefulness in depicting regions. In some areas, one combi-
nation of factors may be more useful than another for detecting regional
patterns, and care must be taken to select the right combination each time.
The complex interplay among the various factors must also be considered.
Omernik's approach to defining ecoregions grew out of an effort to
classify streams for more effective water quality management. Thus, it is
one of the few ecological frameworks expressly intended for water quality
assessment. In examining spatial patterns of stream quality data, it became
clear that neither major land resource areas nor Bailey's ecoregions were
adequate (Hughes and Omernik, 1981; Omernik, 1987; Omernik et al.
1982). Hydrologic unit classifications have also been used as a framework
for water quality assessments, and drainage basins influence fish distribu-
tions, but the spatial differences in the quantity and quality of aquatic re-
sources usually correspond more to ecoregions than to topographic
divides (Omernik and Griffith, 1991).
Ecoregions have been used successfully to stratify the biotic charac-
teristics of streams in Arkansas (Rohm et al. 1987), Nebraska (Bazata, 1991),
Ohio (Larsen et al. 1986); Oregon (Hughes et al. 1987; Whittier et al. 1988),
Wisconsin (Lyons, 1989), and the region of the Appalachians (Gerritsen et al.
1993). Arkansas, Minnesota, and Ohio use the ecoregion/biocriteria ap-
proach in their standards program; and several other states, such as Florida,
Mississippi, Alabama, Idaho, Montana, Oregon, Washington, and Iowa, are
evaluating the advantages of using ecoregions for biological assessments.
One advantage of having a consistent framework is that states that
share the same ecoregion can cooperate across political boundaries. In
times of limited resources, such cooperation makes financial as well as sci-
entific sense. Where ecoregional biological criteria and use designations
have been tested, they have proven to be cost-effective and protective tools
(Hughes, 1989). EPA's Science Advisory Board (SAB) has concluded that
the ecoregion concept "is superior to the classification methods that are
currently used by most environmental managers" (U.S. Environ. Prot.
Agency, 1991e).
34
-------
CHAPTER 3: -
The Reference Condition
Careful review of the purposes of regionalization and selection of the
appropriate regional framework is an important part of the development
of biocriteria. It may also be necessary to increase the resolution of existing
regional frameworks by defining separate regions or subregions. Tech-
niques for this, process are described in the references listed in this docu- •
ment, particularly in Omernik's studies and Iffrig and Bowles's
compendium of regional frameworks (1993)..
Watersheds
Watersheds are a spatial organizing unit that, can be used to develop
biocriteria; however, watershed boundaries are not inconsistent with
ecoregions. Increasing attention has been focused on reorienting water
quality management programs to operate basinwide on a more compre-
hensive, coordinated basis than is possible within strict programmatic
boundaries. EPA's Watershed Protection Approach (U.S. Environ. Prot.
Agency, 1991f; 1993) encourages states to move in the direction of basin-
wide water quality management. The basinwide approach provides a
framework within which to design an optimal mix of water quality man-
agement strategies. By integrating and coordinating across program and
agency boundaries, basinwide management teams can achieve integrated
solutions using limited resources. Thus, they can address the most signifi-
cant water quality problems without losing sight of other factors contrib-
uting to the degradation of the resource. The basinwide approach helps
managers achieve their short- and long-term goals for the basin by allow-
ing the application of resources in a timely and geographically targeted
manner.
Basinwide management as designed and implemented by states and
EPA contains certain features that make it a fitting element of the biocrite-
ria process:
¦ River Basin Management Units. The state is divided into large-scale
basins that provide unique units for. management. All program activities
that can be facilitated by or that affect basinwide management are coordi-
nated. For instance, data requirements are aggregated and incorporated
within monitoring plans, interpretations are pooled to arrive at overall as-
sessments, and management recommendations are the result of collabora-
tion (e.g., teams of modelers, permit writers, biologists, hydrologists,
planners, engineers).
¦ Geographic Risk-based Targeting. Because all states have limited re-
sources and are not able to assess and solve every problem in a watershed,
basin management frameworks establish a set of criteria for giving priority to
the most important problems in a given area. These problems may include
risks to water quality, aquatic life, or human health. While every basin in a
state is visited during a basin management cycle, some waters within and
across basins receive a great deal more attention than others.;
¦ Direct Link to Regionalization. An important feature of the basin man-
agement approach is its ability to incorporate a nested hierarchy of hydro-
logic units. Minshall (1993) discusses the need to assess ecological
conditions in streams and rivers within a hierarchical landscape-scale ap-
proach. Frissell et al. (1986) present a hierarchical framework for class-
ifying stream habitat within an overall watershed perspective. Their
Careful review of the
purposes for
regionalization and
selection of the
appropriate regional
framework is an
important part of the
development of
biocriteria. It may also
be necessary to
increase the
resolution of existing
regional frameworks
by defining separate
regions or subregions.
35
-------
Table 3-1.— Hierarchical classification of stream riparian habitats (from Mlnshall, 1993; after Frissell et al. 1986),
BOUNDARIES I
STREAM HABrTAT
(LINEAR SPATIAL
SCALE)
DEFINING MEASURES
LONGITUDINAL
LATERAL
APPLICATION
SOURCE OF INFORMATION
PROCEDURE/GUIDELINES
REFERENCES
BiogeocUmatic
region
(10s m)
Regional climate
Regional geology
Regional topography
Regional terrestrial
vegetation
Flow regime
Region; State;
Forest; District
Topographic maps (15")
Geologic maps (15")
Landset photos
Annual discharge records
Omernik, 1987
Poff and Ward, 1989
Stream system
(104-10 m)
Local climate
Local geology
Local topography
Local terrestrial
vegetation
Thermal regime
Drainage divides, and
seacoast, or
catchment, area
Drainage divides,
bedrock faults,
joints controlling
ridge valley
development
Basinwkle surveys;
Cumulative impacts;
Integration of sites
within watersheds
Topographic maps (7.5*)
Geologic maps
Vegetation maps
Aerial photos
Annual temperature
records
Omernik and Gallant,
1986
Vannote and Sweeney
1980 '
Chorley et al. 1984
Gregory and Walling, 1973
Segment system
(10s-102 m)
Tributary junctions
Major geologic
discontinuities
Tributary junctions;
major falls; bedrock
lithologic or structural
discontinuities
Valley sideslopes
or bedrock
outcrops
controlling lateral
migration
Parted watersheds;
Segment classes
(e.g., uplands vs
lowlands)
Topographic maps (7.5')
Ground reconnaissance
Low level aerial photos
Reach system,
(101-102m)
Channel slope
Valley form
Bed material
Riparian vegetation
Slope breaks;
structures capable of
withstanding
< 50-year flood
Local sideslopes
or
erosion-resistant
banks;
50-year flood plain
margins
Local effects;
Grazing allotments;
Dredging
Group survey/mapping
Frissell et al. 1986
Rosgen, 1985; 1933
Minshail et al. 1989
Minshall, 1984
McDonald et al. 1991
Plafkin et al. 1989
Platts et al. 1983, 1987
Petersen, 1992
Pool/riffle system
(10°-10'm)
Bedform and
material origin
Persistence
Mean depth and
velocity
Water surface and
bed profile slope
breaks; location of
genetic structures
Mean annual flood
channel;
midchannel bars;
other flow-splitting
obstructions
Aquatic habitat
inventories;
Fisheries
Censuses
Group survey/mapping
Frissell etal. 1986
Bissonet al. 1981
McCain etal. 1990
Microhabitat system
(10' -10°m)
Surface particle size;
underlying particle
size; water depth;
velocity; overhead
cover (type)
Zones; differing
substrate type; size
arrangement
Same as
longitudinal
Characterization of
local spatial
heterogeneity and
effects (e.g., wading
by fishers)
Direct measurement
-
-------
CHAPTER 3:
The Reference Condition
framework is designed so that the class of any particular system is par-
tially determined by the class of the higher-level system to which it be-
longs.
At the broadest scale of organization, Frissell et al. (1986) recognized
stream systems (i.e., watersheds), followed in order of increasing spatial
resolution (and decreasing spatial extent) by segment, reach, pool or riffle,
and microhabitat systems. Minshall (1993) extends the upper end of this
classification scheme to include biogeoclimatic regions, thus providing a
direct connection to ecoregions; and Gregory et al. (1991) similarly discuss
the ecosystem attributes of riparian zones.
Table 3-1 summarizes the Frissell et al. (1986) classification framework
as modified by Minshall (1993). Initial stratification of sites by biogeocli-
matic regions can be performed using ecoregion delineation (Omernik,
1987). Incorporation of flow information using procedures of Poff and
Ward (1989) provides further refinement of this scale of stratification and
includes explicit recognition of flow as a major environmental determi-
nant of stream and river ecosystems (Minshall, 1993; Rabeni and Jacobsen,
1993).
Ecoregions are the preferred classification for establishing reference
expectations in watersheds because biota and biotic metrics respond to
ecoregional differences. Ecoregional stream systems are defined primarily
by local conditions of climate, geology, topography, and terrestrial vegeta-
tion. Three examples of ecoregions are sufficient to illustrate biological
variability:
1. The Calapooia River watershed (Fig. 3-2) in western Oregon
crosses three ecoregions: the Willamette Valley plains; the transit
tional foothills region; and the Western Cascades (Omernik and
Griffith, 1991). Fish, benthic macroinvertebrates, and chemical and
physical habitat from 17 sites along the length of the watershed
were sampled to assess changes in the river as it passed through
these ecoregions. The presumption was that similar biological com-
RillemetU Ytiltj Plflim
viliMtU* ¥«11*f
Vdtiri CtKila
FISH ASSEMBLAGES
• VillamtUt Vstlij
• VilltmtUt UWty FssUili!)
• Brush Crttfc
O Vtstirn Coiesiei
O Vcitirn Coiccio Huitjtm
Ecoregions are the
preferred
classification for
establishing reference
expectations in
watersheds because
biota and biotic
metrics respond to
ecoregional
differences.
Figure 3-2.—Reciprocal averaging ordination of sites by fish species in the Calapooia
River watershed, Oregon. The Inset shows the correspondence between fish assem-
blages In the rivers and ecoregions.
37
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Acceptable
reference sites will
differ among
geographic regions
and states because
soil conditions,
stream morphology,
physiography,
vegetation, and
dominant land uses
differ among regions.
munities would be found in areas of similar habitat, and that vari-
ation would correspond to observable patterns of change in the ter-
restrial features of the watershed.
The study results indicate that imposing an ecoregions frame-
work on the watershed delineation is a useful predictor of stream
reaches having similar biological communities. Although there was
change in the communities along the watershed, distinct assem-
blages could be identified corresponding to the separate ecoregions
within the Calapooia River watershed.
2. Ohio consists of two hydrographic basins, a Lake Erie drainage
and an Ohio River drainage. Hydrographic boundaries restrict fish
dispersal, and there are minor faunal differences between the two
basins (Ohio Environ. Prot. Agency, 1987; Yoder, 1991). Ohio also
includes parts of five ecoregions, and ecoregional differences ac-
count for a substantial amount of the variance in fish metrics and
in the index of biotic integrity (IBI). Two ecoregions straddle the di-
vide between the basins, one is entirely in the Lake Erie drainage,
and two are entirely in the Ohio River drainage. If there are major
differences between drainage basins, then the ecoregions that
straddle the basins should be more variable. However, variability
of IBI scores in all five ecoregions is similar, showing that drainage
basins are negligible compared to ecoregions for explaining bio-
logical variability.
3. Florida comprises two major drainages, the Gulf of Mexico and the
Atlantic Ocean. Examination of invertebrate metrics at reference
sites in Florida1 reveal three ecoregional classes: northwest Florida
(the Florida panhandle); peninsular Florida, and northeast Florida
(EA, Inc., and Tetra Tech, Inc., 1994). Peninsular and northeast Flor-
ida both straddle the divide between the Atlantic and Gulf drain-
ages; yet there are no major differences in metric values between
Atlantic and Gulf basin sites on the Florida peninsula, and the pen-
insula differs markedly from the panhandle region, which is in the
Gulf drainage.
Biogeographic differences between watersheds can be important when
the watersheds are separated by a major, largely impenetrable barrier,
such as the Continental Divide. Drainage dividers in more level terrain
apparently do not cause significant differences in reference expectations.
Thus, implementation of biocriteria, as noted earlier, is best accom-
plished through an ecoregionalization-approach. The implications of this
with jespect to states that are developing basinwide management ap-
proaches is that there may be a set of reference conditions and biocriteria
established for each of the separate ecoregion areas within a given basin;
Ecoregional reference conditions and biocriteria will likely be transferable _
across basins in a given state and — to the extent that ecoregions cross
state boundaries — across states. This transferability enhances the ability
of adjacent states to develop coordinated basinwide management plans
for interstate basins by providing a common set of reference conditions
and data to be applied in the corresponding ecoregions.
38
-------
. CHAPTER 3:
The Reference Condition
Site Selection
Because absolutely pristine habitats do not exist, resource managers must,
as previously noted, decide what level of disturbance is acceptable in the
area that represents the reference condition. That is, a critical element in
establishing reference conditions is deciding how to determine that a site
is only "minimally impaired." How much degradation can be allowed?
Acceptable reference sites will differ among geographic regions and states
because soil conditions, stream morphology, physiography, vegetation,
and dominant land uses differ among regions.
The selection of representative and minimally impaired reference sites
involves qualitative and quantitative information based on past experi-
ence and potential disturbances in regional streams. Factors that should be
considered in a preliminary selection, in approximate order of importance,
include the following:
1. All drainage within the ecoregion of interest.
2. No upstream impoundments.
3. No known discharges (NPDES) or contaminants in place.
4. No known spills or other pollution incidents.
5. Low human population density.
\
6. Low agricultural activity..
7. Low road and highway density.
-
8. Drainage on public lands.
9. Minimal nonpoint source problems (agriculture, urban, logging,
mining, feedlots, acidic deposition).
10. No known intensive fish stocking (e.g., put-and-take stocking) or
other management activities that would substantially shift the
community composition.
In most settled regions of the country, reference sites will be selected
by searching topographic maps for streams with the least human impacts.
If candidate reference sites are more numerous than can be sampled, they
should be selected randomly. Random selection will be especially impor-
tant in regions with large undeveloped or undisturbed areas (e.g., moun-
tainous regions, federal lands). Agricultural and heavily populated
regions — including most of the East, Midwest, and California — will re-
quire subjective (nonrandom) reference site selection.
In most settled
regions of the country,
reference sites will be
selected by searching
topographic maps for
streams with the least
human impacts.
Montana Reference Conditions
The Montana Department of Health and Environmental Sciences (1990)
has compiled data that describe reference conditions. Thirty-eight streams
were proportionally allocated among six ecoregions in Montana, and the
following criteria were used to determine a set of candidate reference
streajris.
1. Most or all of the drainage basin of candidate streams is in the
"most typical" are.a of the ecoregion.
39
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
2. Each ecoregion includes at least two second-order streams, two
third-order .streams, and two fourth- or fifth-order streams.
3. Reference streams are not water quality limited.
4. The same streams serve as references for proposed Montana
nonpoint source demonstration projects.
5. Reference streams adequately represent the major water use
classifications in each ecoregion.
6. Information is available on the kinds and abundances of fish
species present in the streams.
7. Sampling sites have comparable habitat from stream to stream
and are located to minimize human impacts and access problems.
Site selection in the Appalachian Ridge and Valley
Because of differences in dominant land use and amounts of degradation,
neighboring ecoregions may have widely different reference sites and con-
ditions. For example, in the Central Appalachian Ridge and Valley ecore-
gion, criteria for selecting reference sites differ between the region's
agricultural valley subecoregions and its forested ridge subecoregions
(Gerritsen et al. 1993; Omernik et al. 1992).
The Ridge and Valley region of the Appalachians consists of sharply
folded sedimentary strata that have eroded, resulting in a washboard-like
relief of resistant ridges alternating with valleys of less-resistant rocks. The
region has been divided into four subecoregions corresponding to ridges
and valleys of different parent material (Omernik et al. 1992):
• Limestone valleys are characterized by calcareous bedrock and predomi-
nantly agricultural land use.
• Shale valleys are characterized by noncalcareous bedrock, primarily
shale; and lower intensity agricultural land use.
• Sandstone ridges are characterized by highly resistant sandstones and
forested land use.
• Shale ridges are characterized by shale bedrock and forested land use.
Each subecoregion imparts characteristic topography, hydrology, and
water chemistry to streams and thus influences biota. The subecoregions
are not continuous but interdigitate throughout the Ridge and Valley.
The least impacted sites occur on the ridges, where land use is pre-
dominantly forested, and where protected lands (e.g., national forests, rec-
reation areas) are common. In contrast, nearly all streams in the valleys,
and especially in the limestone valleys, are impacted by agriculture, habi-
tat modification, and other nonpoint sources. "Minimally impaired" is,
therefore, interpreted on a relative, sliding scale in each subecoregion. Ref-
erence sites for the ridges are strictly defined: they are unimpacted except
by atmospheric sources. They have no discharges, nearly complete forest
cover in the drainage, and no recent construction or clearcutting in the
drainage. Reference sites in the valley subecoregions are less strictly de-
fined; that is, the interpretation of minimally impaired is flexible enough
to allow a sufficient number of reference sites to be selected.
40
-------
CHAPTER 3:
The Reference Condition
Confirming Reference Conditions — Successful
Classifications
Following site selection, reference sites are surveyed (see Chapter 4) to col-
lect biological and physical data. The data are used to confirm and refine
v the a priori classification, to characterize reference conditions, and to es-
tablish biocriteria (see Chapter 6). Classification is a general guide for con-
firming reference conditions; its effectiveness is its ability to partition
variation. If a classification does not account for variability, it is of little
use; the greater the amount of variance accounted for by classification, the
more effective the classification.
A key analysis method for evaluating the strength of metrics to detect
impairment is a graphic display using box-and-whisker plots (Fig. 3-3). In
Max
Min
maximum
75th percentile
median
25th percentile
minimum
interquartile
range
scope for
detecting
impairment
Reference Impaired
a. Metrics that have high values under reference (unimpaired) conditions.
Max -
Min
scope for
detecting
impairment
interquartile
range
Reference Impaired
b. Metrics that have low values under reference conditions.
Figure 3-3.—Generalized box-and-whisker plots Illustrating percentiles and the detec-
tion coefficient of metrics.
Classification is a
general guide for
confirming reference
conditions; its
effectiveness is its
ability to partition
variation.
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
The fundamental
problem of biological
assessment is not
whether two
populations (or
samples) have a
different mean, but
whether an individual
site is a member of
the least-impaired
reference population.
Since assessment is
based on multiple
metrics or species
composition,
multivariate tests may
be more convenient
than a succession of
individual tests.
the display shown here, the central point is the median value of the vari-
able; the box shows the 25th and 75th percentiles (interquartile range); and
the whiskers show the minimum to the maximum values (range). Box-
and-whisker plots are simple, straightforward, and powerful; the inter-
quartile ranges are used to evaluate real differences between two areas
and to determine whether a particular metric is a good candidate for use
in the assessment.
Statistical methods used by biologists to determine whether two or
more populations have different means using t tests include the analysis
of variance and various nonparametric methods. However, the fundamen-
tal problem of biological assessment is not whether two populations (or
samples) have a different mean, but whether an individual site is a mem-
ber of the least-impaired reference population. If it is not, then the second
question is, how far has it deviated from that reference? Such biological
assessment requires the entire distribution of a metric, which is easily
shown with a box-and-whisker plot.
In operational bioassessment, metric values below the lower quartile
of reference conditions are typically judged impaired to some degree (e.g.,
Ohio Environ. Prot. Agency, 1990). The actual percentile chosen (25,10, or
5) is arbitrary and reflects the amount of uncertainly a monitoring pro-
gram can tolerate. The distance from the lower quartile can be termed a
"scope for detection" (Fig. 3-3). The larger this distance is compared to the
interquartile range, the easier it is to detect deviations from the reference
condition. Thus, we define a "detection coefficient" as the ratio of the in-
terquartile .range to the scope for detection. This coefficient is analogous to
the coefficient of variation (CV); the smaller the value, the easier it is to de-
tect impairment.
Univariate tests of classifications include all the standard statistical
tests for comparing two or more groups: f test, analysis of variance,1 sign
test, Wilcoxon rank test, Mann-Whitney U-test (Ludwig and Reynolds,
1988). These methods are used to test for significant differences between
groups (or classes) and to confirm or reject the classes. However, failure to
confirm the classification for any single response variable does not mean
that it will fail for other response variables.
Since assessment is based on multiple metrics or species composition,
multivariate tests may be more convenient than a succession of individual
tests. Discriminant analysis is a multivariate test included in many statisti-
cal software packages. It is a one-way analysis of variance that tests differ-
ences between a set of groups based on several response variables; and it
can be used as a test of classifications (Conquest et al. 1994), provided that
the assumptions of linearity and normality are met.
A satisfactory analysis is to develop quantitative, predictive models of
biological response to habitat variables. Using a defined population of refer-
ence sites that are relatively undisturbed, investigators can develop an em-
pirical (statistical) model that predicts biological communities based on the
habitat variables (e.g., Wright et al. 1984; Moss et al. 1987). Univariate mod-
els, such as multiple regression or analysis of covariance, are linear and re-
quire appropriately transformed linear variables. Community metrics tend
to respond linearly, or can be readily transformed to linearly responding
variables. Species abundances are typically nonlinear (usually unimodal) in
response to environmental variables and require nonlinear models.
42
-------
CHAPTER 3:
The Reference Condition
Median
HELP IP EOLP WAP
ECOREGIONS
Figure 3-4.—Index of Biotlc Integrity at Ohio reference sites.
• <10.0
ECBP
in
UJ
o
LLi
Q.
(/>
30,0
20.0
10.0
0.0
¦k
LOG_WA
Figure 3-5,—Fish species richness as a function of the log of watershed area. Bars to
right indicate range of observations before regression and range of residuals after re-
gression. Residuals have smaller variance than the original observations.
The role classification plays in partitioning variation can be illustrated
using an example drawn from an extensive biosurvey database developed
by the Ohio EPA. A national map of ecoregions (Omernik, 1987) indicates
that parts of five ecoregions fall within Ohio. Comparison of the range of
IBI, a measure of fish assemblage condition, illustrates that one ecoregion,
the Huron/Erie Lake Plain, is characterized by substantially lower values
than that observed in the other ecoregions (Fig. 3-4). The IBI was highest
in the Western Allegheny Plateau ecoregion.
Careful classification
contributes
significantly to the
refinement and use of
reference conditions
for establishing
biocriteria.
43
-------
BIOLOGICAL CRITERIA.
Technical Guidance for Streams $nd Small Rivers
In this example, classification is used iteratively, that is, decisions for
successive classifications are based on their ability to partition variation
from that which would be present on a statewide basis.
One way to partition variance is by examining possible gradients to
which the indicators of biotic condition may be related. Some possible gra-
dients are stream size, physical habitat condition, and stream gradient. In
Figure 3-5, species richness is plotted against a log of watershed area; the
watershed area is used as a surrogate measure of stream size. The relation-
ship is clear: increasing species richness in the reference site occurs as
stream size (watershed area) increases. In this case, watershed size is used
as a covariate to provide adjustments in the expected number of species
associated with the drainage area within each class size.
In summary, careful classification contributes significantly to the re-
finement and use of reference conditions for establishing biocriteria. An it-
erative process is envisioned by which various classifications of regions
and subregions are proposed and evaluated against partitioning of vari-
ance: successful classifications partition variance effectively; ineffective
classifications provide little improvement beyond no classification. This
evaluation process should generally involve multiple metrics to judge the
success of multiple purpose ecoregion classifications.
Suggested Readings
Gallant, A.L. et al. 1989. Regionalization as a tool for managing environmental re-
sources. EPA 600/3-89/060. Environ. Res. Lab., U.S. Environ. Prot. Agency, Corval-
lis, OR.
Hughes, R.M., D.P. Larsen, and J.M. Omernik. 1986. Regional reference sites: A method
for assessing stream potentials. Environ. Manage. 10:629-35.
Iffrig, G.E and M. Bowles. 1983. A compendium of ecological and natural subdivisions
of the U.S. Nat. Areas J. 3:3-11.
Omernik, J.M. 1987. Ecoregions of the conterminous United States. Annu. Ass. Am.
Geogr. 77(l):118-25.
Omernik, J.M. and G.E. Griffith. 1991. Ecological regions versus hydrologic units:
Frameworks for managing water quality. J. Soil Water Conserv. 46(5):334-40.
U.S. Environmental Protection Agency. 1991d. Biological Criteria: Research and Regula-
tion Proceedings of the Symposium. EPA-440/5-91-005. Off. Water, Washington,
DC.
44
-------
CHAPTER 4.
Conducting the
Biosurvey
The primary goals of a bioassessment-biocriteria program are to evalu-
ate water resource integrity, to provide information on the attainabil-
ity and appropriateness of existing uses, and to determine the extent and
degree of water resource impairment.
State bioassessment-biocriteria programs are usually designed to ad-
dress one or more of four water resource management objectives:
1. Aquatic Life Use Designation. Determine and assess aquatic life
uses that should be attained in streams and rivers. Helping to des-
ignate and assess aquatic life uses is a major function of biological
criteria.
2. Sensitive Waters Identification. Characterize high quality waters
for protection. High quality waters may become part of the refer-
ence database or be classified separately as unique waters.
3. Diagnostics. Determine sources of impairment and potential stres-
sors. Biological response signatures are used in conjunction with
chemical, toxicological, and physical data to identify causes of im-
pairment.
4. Program Evaluation. Monitor effectiveness of pollution abatement
programs, including wastewater treatment, watershed restoration,
and other water resource quality improvement programs. Biosur-
veys and the biocriteria benchmarks are used to assess the recovery
of the aquatic community.
Detailed multidisciplinary ecological studies are often designed to ex-
amine aquatic systems by measuring the elements and processes of bio-
logical communities and by describing the physical and chemical
characteristics of the waterbody. Biological attributes that may be included
in such studies are individual health, trophic organization, measures of
primary, secondary, and tertiary production (bodily growth and reproduc-
tion), recruitment of key species, predator-prey relationships, population
dynamics, and taxonomic structure of assemblages.
While seasonal accommodation is preferable for most bioassessment
programs, a single annual sample at a carefully selected time is sufficient
Purpose:
To provide guidance
to technical staff for
developing
cost-effective
biosurvey methods
with appropriate .
. resources, expertise,
and technical
considerations.
45
-------
BIOLOGICAL CRITERIA
Technical Guidance for Streams and Small Rivers
Quality assurance
and control should be
a continuous process
throughout the
development and
operation of the
biocriteria program,
including all aspects
of the study.
Quality assurance
and control pervade
all aspects of an
ecological study:
a Study design
m Field operations
m Laboratory activities
m Data analysis
¦ Reporting
to characterize biological conditions accurately. Selection of the sampling
period should be based on efforts to minimize variability and maximize
the efficiency of the equipment and the accessibility of the biota being
sampled. Minimal between-year variability is partially addressed by sam-
pling at the same time each year to correct for the natural variability in
seasonal cycles..
Water quantity, quality, and climatic conditions should help rather
than hinder the efficiency of the sampling gear. For example, if certain
flow conditions are necessary for the equipment's performance, sampling
schedules should coincide with those conditions. Above all, sampling
should occur when the targeted assemblage or assemblages are accessible.
For fish, the optimal sampling period in most parts of the country is likely
to be from June through September; in general, these months avoid high
and low flows, spawning periods, and migration activity. Sampling should
be timed to avoid extremes in environmental and biological conditions.
Quality Assurance Planning
A major consideration when designing bioassessment studies is not
whether a particular biosurvey approach is more refined than another, but
whether the selected approach will achieve the objectives defined in the
management plan. A clear definition of management responsibilities and
effective quality assurance and quality control procedures (see Chapter 2)
are essential to ensure the usefulness of monitoring data (Plafkin et al.
1989).
Quality assurance plans have two primary functions (Klemm et al.
1990). The first function is to ensure that the survey process reliably meets
program objectives; the second is to monitor the reliability of the survey
data to determine their accuracy, precision, completeness, comparability,
and representativeness.
A quality assurance plan should be developed at the onset of an eco-
logical study to delineate responsibility, establish accountability, and en-
sure the reliability of the data (Stribling and Barbour, 1991). The quality
assurance plan should answer three questions:
• What kind of data or information is needed?
• Why is the information or data needed?
• What level of quality is needed to ensure the reliability of decisions
based on these data?
Quality assurance for a biocriteria program is concerned with the in-
tegrity of the data used to establish biocriteria limits and thresholds along
with the documentation that supports the derivation and maintenance of
the biocriteria. Quality assurance for specific studies pertains to the data
acquisition, their application to established biocriteria, and the validity of
associated judgments.
Quality assurance and control should be a continuous process
throughout the development and operation of the program, including all
aspects of the study: design, field collection, habitat assessment, labora-
tory processing of samples, database management, analysis, and report-
-------
CHAPTER 4:
Conducting the Biosurvey
ing. The appropriateness of the investigator's methods and procedures
and the quality of the data to be obtained must be assured before the re-
sults can be accepted and used in decision making. Quality assurance is
accomplished through data quality objectives, investigator training, stand-
ardized data gathering and processing procedures, verification of data re-
producibility, and instrument calibration and maintenance.
The use of data quality objectives in field studies (Klemm et al. 1990;
Plafidn et al. 1989; U.S. Environ. Prot. Agency, 1984b, 1986) has much to of-
fer the biocriteria development and implementation process. Data quality
objectives are qualitative and quantitative statements within the quality
assurance plan that address specific decisions or regulatory actions. Gen-
erally, data quality objectives consist of a priori statements about the level
of uncertainty a decision maker will accept in environmental data. Once
the objectives are stated, the quality of particular data can be measured us-
ing predetermined types and amounts of error associated with their col-
lection and interpretation.
Quality Management
The implementation of a biocriteria program requires quality management
or the proper combination of resources and expertise. State agencies will
differ in levels of biological expertise, facilities, and quality of equipment.
States already having well-developed bioassessment programs generally
have experienced and well-trained biologists, appropriately equipped fa-
cilities, and properly maintained sampling gear. A successful biocriteria
program depends on (1) a clear definition of goals, (2) the active use of
biomonitoring data in decision making, and (3) the allocation of adequate
resources to ensure a high quality program.
Biocriteria Program Structure, Personnel, and Resources
Monitoring agencies can and should enhance their program by coopera-
tion with others. For example, they should seek coordination with, and
staff assistance from, state fishery, land management, geology, agriculture,
and water quality agencies. If federally employed aquatic biologists are
stationed in a state or if the state has substantial federal lands, cooperative
bioassessments and biocriteria development programs should be initiated.
Scientists at state universities should also be included in the planning and
monitoring phases of the program; their students make excellent field as-
sistants and future state ecologists.
¦ Personnel. Several trained and experienced biologists should be avail-
able .to provide more thorough evaluations, support for various activities,
and serve as quality control checks. They should have training and experi-
ence commensurate with the needs of the program. At least one staff
member should be familiar with establishing a quality assurance frame-
work. The program should have at least one biologist for every 4,000 miles
of stream in the state (C. Yoder and R. Thoma, personal communication).
¦ Resources. Laboratory and field facilities and services should be in
place and operationally consistent with the designed purposes of the pro-
gram so that high quality environmental data may be generated and proc-
essed in an efficient and cost-effective manner (Klemm et al. 1990).
Monitoring agencies
can and should
enhance their
program by
coordination with, and
staff assistance from,
state fishery, land
management,
geology, agriculture,
and water quality
agencies.
47
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Adequate taxonomic references and scientific literature should support
data processing and interpretation. The following program and technical
considerations should guide the design and implementation of the biocrit-
eria program.
¦ Program Elements
1. Quality assurance and quality control (e.g., standard operating
procedures, training)
2. Delineated reference conditions with annual updates
corresponding to seasons of sampling
3. Multiple assemblage biosurvey
4. Habitat assessment
5. Documentation of program and study plans
¦ Technical Considerations
1. Assign taxonomy to the lowest possible level based on published
keys and descriptions; maintain voucher collections
2. Schedule multiple season sampling if warranted by type of
impact and life strategy of assemblage
3. Use multiple metrics to refine the assessment
4. Initiate detailed quality assurance and quality control procedures
in the field, laboratory, and taxonomy
5. Provide computer hardware and software (database management,
data analysis) with computer training of staff
Different levels of training and experience are necessary for the per-
sonnel involved in biocriteria programs. The qualifications and general job
descriptions of four levels of professional staff are presented here. Also de-
scribed are suitable substitutions for these prerequisites and experience.
¦ Professional Staff
1. Level 4 — Plans, conducts, and supervises projects of major signifi-
• cance, necessitating advanced knowledge and the ability to origi-
nate and apply new and unique methods and procedures. Supplies
technical advice and counsel to other professionals. Generally op-
erates with wide latitude for unreviewed action.
Typical Title: Project Manager, Chief Biologist.
Normal Qualifications: Ph.D. or M.S. and equivalent experience.
Experience: Ten or more years, at least three years in a leadership
or managerial position-
2. Level 3 — Under general supervision of project manager, plans,
conducts and supervises bioassessment tasks such as trend moni-
toring or special studies. Estimates and schedules work to meet
completion dates. Directs support assistance, reviews progress, and
evaluates results; makes changes in methods, design, or equipment
as necessary. Operates with some latitude for unreviewed action or
decision.
4
-------
CHAPTER 4:
' Conducting the Biosurvey
Typical Title: Project Biologist, Group Leader, Crew Leader.
Normal Qualifications: M.S., B.S., or equivalent experience.
Experience: Six or more years in or related to bioassessment, two
to three years in a supervisory capacity.
3. Level 2 — Under supervision of a chief biologist or project man-
ager, carries out assignments associated with projects. Translates
technical guidance received from supervisor into usable data appli-
cable to the particular assignment; coordinates the activities of jun-
iors or technicians. Work assignments are varied and require some
originality and ingenuity. . .
Typical Title: Associate Biologist, Environmental Scientist.
Normal Qualifications: B.S. or equivalent experience
Experience: Three to eight years in or related to freshwater biol-
ogy- .
4. Level 1 — Lowest or entering classification. Works under close su-
pervision of a group or crew leader. Gathers and correlates basic
data and performs routine analyses. Works on less complicated as-
signments that require little evaluation.
Typical Title: Field Technician.
Normal Qualifications: B.S. or Associate Degree and equivalent
experience.
Experience: zero to three years.
¦ Experience/Qualifications Substitutions
1. Any combination of additional years of experience in the proposed
field of expertise and full-time college-level study in the particular
field totaling four years of structured, directed education may be
substituted for a B.S.
* '
2. A B.S. and any combination of additional years of experience and
graduate-level study in the proposed field of expertise totaling two
years may be substituted for the M.S.
3. A B.S. and any combination of additional years of experience and
graduate study in the proposed field of expertise totaling four
years; or an M.S. and two years of either additional experience or
graduate-level study in the proposed field may be an acceptable
substitute for the Ph.D.
4. Additional years of graduate-level study in an appropriate field
will be considered equal to years of experience on a one-for-one ba- -
sis.
The quality manager will identify project responsibilities and account-
abilities for the bioassessment program. In states with limited resources,
the basic responsibilities for all levels will rest with relatively few indi-
viduals; however, the accountability of each position will be quite distinct.
'9
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Effective quality
control procedures
are essential to insure
the usefulness of the
data for biocriteria
development and
environmental
decision making, and
to maintain the
hioassessment
program.
Project Manager / Principal Investigator
OA Officer
ECOLOGICAL PROJECT ACTIVITY CLASSES
SAMPLING
DESIGN
FIELD
ACTIVITIES
LABORATORY
ACTIVITIES
DATA
ANALYSIS
REPORTING
Sampling
Design
Coordinator
Design
QC
Field
Leader
Field
QC
—j Statistician]
—ISeniorPersonnel]
—| User/Contacts |
-Biota
Laboratory
Manager/Leader
Laboratory
QC
I Taxonomy i
—[Wfate:
Data Processing
Leader
Data
QC
Document Production
Coordinator
J Reporting
1 QC
| Sample Processing
I Data Presentation Data Interpretation
[ Sample Handling
Data Entry
Technical Editor
Figure 4-1.—Organizational chart Illustrating project organization and lines of respon-
sibility.
Quality management is an important planning aspect of the biocriteria
development process that focuses attention on establishing and improving
quality in all aspects of a project. Quality management requires that all
personnel involved in a biocriteria project (from senior management to
field and laboratory technicians) be aware of and responsive to data needs
and expectations. The surest way to achieve total quality management
(TQM) in an environmental program is to implement an achievable qual-
ity assurance program.
Quality Control Elements in an Ecological Study
Effective quality control procedures are essential to insure the usefulness
of the data for biocriteria development and environmental decision mak-
ing, and to maintain the bioassessment program. The organizational chart
in Figure 4-1 identifies the major activity classes in an ecological project;
Table 4-1 outlines" the quality control elements that are integral to those ac-
tivities.
All activity classes or phases of field ecological studies have potential
error sources associated with them (Barbour and Thornley, 1990). Some
general quality control elements for reducing error are discussed here; for
more specific approaches, the investigator should refer to Klemm et al.
(1990) for benthic macroinvertebrates; and to Karr et al. (1986), Lyons
(1992), and Ohio Environ. Prot. Agency (1987) for fish.
¦ Study Design. Considerations relating to potential error in the study
design range from limited resources to insufficient sample replication to
selection of inappropriate variables-. Two important considerations for de-
veloping a study design are interrelated: the availability of baseline data in
historical information or pilot studies and the capacity to identify poten-
50
-------
CHAPTER 4:
Conducting the Biosarvey
Table 4-1.—Quality control elements integral to the activities in an ecological
study.
A. Quality Management
1, Delineate responsibilities
2. List accountabilities
3. Identify quality assurance officer
4. Develop quality assurance plan
5. Use bioassessment information in decision making
• B. Study Design
1, Pilot study or site reconnaissance
. 2. Account for environmental strata
3. Incorporate historical data
a. Attempt to duplicate regimes
b. Attempt to use similar equipment (if appropriate to current objectives)
4. Termination of control point
5. Areas of potential error
a. Available resources ,
b. Logistics ' '
c. Response variables
d. Weather
e. Seasonality
f. Site selection
g. Habitat variability '
h. Population variability
i. Equipment
6. Additional performance effect criteria
C. Sample Collection
1. Instrument calibration and maintenance
2. Field crew
a. Training •
b. Evaluation
3. Field equipment
4. Sample handling
5. Effort checks
6. Field crew efficiency
7. Areas of potential error
a. Climate
b. Site selection
c. Sampling efficiency of equipment
d. Equipment operation; human error
e. Field notes
f. Samples
i. Processing
ii. Transportation
iii. Tracking
8. Additional performance effect criteria
D. Sample Processing
1. Sorting and verification
2, Taxonomy
3. Duplicate processing
4, Archival procedures
5. Training
6. Data handling
7. Interlaboratory training and collaboration
'
8. Areas of potential concern
a. Sample tracking • •
b. Improper storage
c. Sample preparation
d. Reference error (taxonomy) '
e. Taxonomic error (human), . \
(continued on next page)
51
-------
BIOLOGICAL CRITERIA.
Technical Guidance for Streams and Small Rivers
Two of the most
important
considerations in
developing a study
design are the
availability of baseline
data in historical
information or pilot
studies and the
identification of
potential sources of
error.
Table 4-1.— Continued.
f. Counting error
g. Sorting efficiency
9. Additional performance effect criteria
E. Data Analysis
1. Training
2. Data
a. Handling
b. Reporting
3. Standardized database
4. Standardized analyses
5. Peer Review
6. Range control
7. Statistical power analysis
8. Areas of potential error
a. Inappropriate statistics
b. Errors in database
c. Database management
d. Programming errors
e. Analytical misinterpretation
9. Additional performance effect criteria
F. Report Preparation
1. Training
2. Peer review
3. Technical editor
4. Standard format
5. Areas of potential error
a. Transcription
b. Poor presentation
c. Obscure language
d. Addressing performance effect criteria
6. Additional performance effect criteria
tial sources of error. In fact, having adequate baseline information may be
the only way to identify sources of error. As more than one quality control
element may be used to reduce potential error, the interaction among
quality control elements must be considered to ensure the overall quality
of the plan.
Six qualitative and quantitative characteristics are usually employed
to describe data quality:
¦ Precision. The level of agreement among repeated measurements of
the same characteristic.
¦ Accuracy. The level of agreement between the true and the meas-
ured value; the divergence between the two is referred to as bias.
¦ Representativeness. The degree to which the collected data accu-
rately and precisely reflect the frequency distribution of a specific
variable in the population.
¦ Completeness. The amount of data collected compared to the
planned amount.
¦ Comparability. The degree to which data from one source can be
compared to other sources.
52
-------
CHAPTER 4:
Conducting the Biosurvey
¦ Measurability. The degree to which measured data remain within
the detection limits of the analysis — often a function of the sensitiv-
ity of instrumentation.
These characteristics should be considered and defined before the data collec-
tion begins. Taken collectively, they provide a summary characterization of
the data quality needed for a particular environmental decision.
¦ Field Operations. The major quality control elements in field opera-
tions are instrument calibration and maintenance, crew training and
evaluation, field equipment, sample handling, and additional effort
checks. The potential errors in field operations range from personnel defi-
ciencies to equipment problems. Training is the most important quality
control element for field operations. Establishing and maintaining a
voucher specimen collection is also important. Vouchers are a mechanism
for achieving the source of the data, particularly for benthos. Use of a pro-
tocol for double data entry and comparison can also increase the quality of
a database.
¦ Laboratory Operations. The quality control elements in laboratory op-
erations are classified as sorting and verification, taxonomy, duplicate proc-
essing, archival procedures, training, and data handling. Potential error
sources associated with sample processing are best controlled by staff train-
ing. Controlling taxonomic error requires well-trained staff with expertise
to verify identifications. Counting error and sorting efficiency are usually
the most prominent error considerations; they may be controlled by dupli-
cate processing, sorting, and verification procedures. Errors associated with
transcription during the data entry process can be significant. In the labora-
tory, as in the field, the use of a protocol for double data entry and compari-
son can increase the quality of a. database, and the establishment and
maintenance of a voucher specimen collection should be considered.
¦ Data Analysis. Peer review and range of values are the important qual-
ity control elements for data analysis. Peer review helps control operator
variability, and measurement values must be kept within the. range of
natural or normal variability. Further, if inappropriate statistics are used to
analyze the data, erroneous conclusions may be drawn regarding trends.
Undetected errors in the database or programming can be disastrous, and
unless steps are taken to oversee data handling and analysis, problems re-
lated to database management will arise. The use of standardized com-
puter software for database management and analysis can minimize
errors associated with tabulation and statistics. A final consideration is the
possible misinterpretation of the findings. These potential errors are best
controlled by qualified staff and adequate training.
¦ Reporting. The quality control elements in the reporting activity in-
clude training, peer review, and the use of a technical editor and standard
formats. The use of obscure, language can often mislead the reader. Peer
review and review by a technical editor are essential to the development
of a scientific document. If the primary objective or central question of the
study is not specifically addressed in the report or the report is ambiva-
lent, then an error in the reporting process has occurred.
53
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Metric
Metric
QC Elements
Specific Questions
Statement of Problem
Acceptable
Uncertainty
Acceptable
Uncertainly
Acceptability
of Study
Sources of
Potential Error
Select Variables
to be Measured
Identify all Variables
Affecting Problem
Develop Judgment
Criteria
- Site Description Characterization
of Problem
- Historical Data
- Data Gaps
Figure 4-2.—Summary of Data Quality Objective (DQO) process for ecological studies
(taken from Barbour and Thornley, 1990).
Data Quality Objectives
The data quality objectives process occurs during the final creation of the
research design. Although its aspects are inherently interrelated, the devel-
opment of data quality objectiyes is not directly linear. Rather, this develop-
ment is an iterative or circular process, as shown in Figure 4-2. The initial
statement of the problem evolves from specific questions about existing
data; then comes the identification and selection of the variables to be
measured, which influence the further refinement of the questions; and, fi-
nally, judgment criteria axe developed for each variable, acceptable uncer-
tainty levels are established, and sources of potential error are identified.
The result of the data quality objectives process 'is a formal document
that can be separate from or part of a formal quality assurance plan. It
may also be included in narrative form in a project workplan. The data
quality objectives document should state the study's primary objectives,
specific questions, and rationale; it should also justify the selection of vari-
ables, establish judgment criteria (by developing a logic statement for each
54
-------
¦ CHAPTER 4
Conducting the Biosurvey
variable), and specify acceptable levels of uncertainty. This information
does not have to be presented in a stepwise fashion, but it should be read-
ily available.
All staff involved in the biocriteria development process — senior
management, program staff, and all technical staff — should be included
in formulating data quality objectives. In fact, quality management in eco-
logical studies requires that all personnel involved in a project be aware of
and responsive to detailed needs and expectations. If appropriately exe-
cuted, data quality objectives will formalize and document all manage-
ment decision points, the necessary data collection and analysis
procedures, the data interpretation steps, and the potential consequences
of making an incorrect decision.
Further details of quality assurance and control programs specific to
fish and macroinvertebrate field surveys, and methods for determining
biological condition, are provided in Klemm et al. (1990) and Plafkin et al.
(1989). General guidance for developing comprehensive quality assurance
and control plans are discussed in the Code of Federal Regulations (40
CFR Part 30), and U.S. Environ. Prot. Agency (1980a,b; 1984a,c). For infor-
mation and guidance specific to data quality objectives, see Klemm et al.
(1990), Plafkin et al. (1989), and U.S. Environ. Prot. Agency (1984b, 1986).
Study Design
The primary focus of the study design is to establish objectives, and the
statement of the problem to be resolved is the central theme of the objec-
tives. For instance, the central problem or question may be, "Is the biologi-
cal integrity of a specified area of a particular watershed impaired by the
operation of a wastewater facility?" This question has several features
that, in turn, provide a foundation for more specific questions. The first
feature is the concept of biological integrity, which implies that a measur-
able reference condition exists for the aquatic assemblages being studied.
The second feature delineates the spatial area to be evaluated in the water-
shed; the third determines whether or not a problem is attributable to the
operation of the facility. Still, more specific questions, or testable hypothe-
ses, related to the central problem may be constructed.
1. Is impairment of the biological condition detectable in the algae,
fish, or macroinvertebrate assemblages?
2. Is degradation altering the energy base, water quality, flow
regime, habitat structure, or other aspect of the environment?
3. Is there a history of problems in this area of the watershed?
4. What was the historical condition of the aquatic community?
Based on these questions, it is possible to select the biotic and abiotic
variables to be measured. For each variable, an acceptable level of degra-
dation should be identified before conducting the biosurvey. Thus, the
study design, includes selecting the aquatic assemblages, resolving the
technical issues associated with their ecology and proper sampling, estab-
lishing standard operating procedures, and beginning the biosurvey pro-
gram.
The primary focus of
the study design is to
establish objectives,
and the statement of
the problem to be
resolved is the central
theme of the
objectives.
55
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
A critical decision in
the design of
biocriteria programs
is how to select
appropriate indicators
of biotic condition.
The importance of
the periphyton
assemblage within
most stream
ecosystems makes it
a prime candidate for
consideration as a
bioassessment-
biosurvey target.
Biosurveys of Targeted Assemblages
A critical decision in the design of biocriteria programs is how to select ap-
propriate indicators of biotic condition. Biosurvey of the targeted assem-
blages is the most widely employed approach to biocriteria development.
This approach, which has been used by Ohio, Illinois, North Carolina,
Maine, Arkansas, New York, and Vermont, focuses on a selected compo-
nent of the biological community; it samples one or several specific
aquatic community segments to measure biological condition. Monitoring
the specific characteristics of these assemblages helps assess the effects of
a variety of environmental conditions (Ohio Environ. Prot. Agency, 1987).
A number of different organisms associated with lotic systems (i.e.,
streams and rivers) lend themselves to bioassessment procedures. Com-
monly measured assemblages include, but are not restricted to, macro-
phytes, algae, macroinvertebrates, and fish. The targeted assemblage
approach to bioassessment can also focus on a single assemblage (e.g.,
periphyton) or several assemblages (e.g., periphyton, macroinvertebrates,
and fish). The attributes measured may be functional parameters, such as
photosynthesis or respiration, or other attributes, such as individual
health. Examples of widely used methods and'techniques for targeted as-
semblages are found in Karr (1981), Karr et al. (1986), Ohio Environ. Prot.
Agency (1987), Plafkin et al. (1989), Standard Methods (1989), U.S. Envi-
ron. Prot. Agency (1990), and Weber (1973). The primary advantages of
this approach are its flexibility, practicality, cost-effectiveness, and relative
scientific rigor.
Attributes of Selected Assemblages
¦ Periphyton. The periphyton assemblage is composed of benthic algae,
bacteria, their secretions, associated detritus, and various species of mi-
croinvertebrates (Lamberti and Moore, 1984). Periphyton are an important
energy base in many lotic situations (Dudley et al. 1986; Minshall, 1978; Ste-
inman and Parker, 1990) and serve as the primary nutrient source for many
stream organisms (Lamberti and Moore, 1984). The capacity of benthic as-
semblages to colonize and increase in biomass is influenced by variability
in stream channel geomorphology, flow rates, herbivore grazing pressure,
light intensity, seasonality, and random processes (Coleman and Dahm,
1990; Grimm and Fisher, 1989; Hamilton and Duthie, 1984; Korte and Blinn,
1983; Lamberti et al. 1987; Patrick, 1949; Poff et al. 1990; Steinman and
Mclntire, 1986,1987; Steinman et al. 1987; and Stevenson, 1990).
The importance of the periphyton assemblage within most stream eco-
systems makes it a prime candidate for consideration as a bioassessment-
biosurvey target. More specific advantages are outlined by Plafkin et al.
(1989):
• The rapid algal reproduction rates and short life cycles of periphyton
make them valuable indicators of short-term impacts.
• Physical and chemical factors have direct effects on the structure and
functions of periphyton and on their production.
• Periphyton sampling methods are straightforward, and the samples are
easily quantified and standardized.
56
-------
. CHAPTER 4:
Conducting the Biosurvey
• Methods have also been standardized for recording functional and
nontaxonomic characteristics of periphyton communities, such as
biomass and chlorophyll measurements.
• Algal components of periphyton are sensitive to some pollutants to
which other organisms may be relatively tolerant.
¦ Macrophytes. The macrophyte assemblage consists of large aquatic
plants that may be rooted, unrooted, vascular, or algiforms. Both emergent
and submergent macrophytes provide numerous benefits to streams and
small rivers thus helping them to support healthy, dynamic, biological
communities (Campbell and Clark, 1983; Hurley, 1990; and Miller et al.
1989). Some understanding of the distributional characteristics and envi-
ronmental conditions affecting macrophytes (Hynes, 1970) enhance their
use in bioassessment strategies. Hynes (1970) and Westlake (1975) discuss
differences in lotic macrophyte assemblages based on habitat factors such
as water hardness, pH, gradient, and propensity for siltation.
Some investigators have emphasized the influence of macrophytes on
habitat structure (Carpenter and. Lodge, 1986; Gregg and Rose, 1982,1985;
McDermid and Naiman, 1983; Miller et al. 1989; Pandit, 1984); others have
studied water chemistry, nutrient cycling, and macroinvertebrate coloniza-
tion (McDermid and Naiman, 1983; Miller et al. 1989). Pandit (1984), Sed-
don (1972), and Westlake (1975) pointed to the use of macrophytes as an
indicator assemblage in lotic situations.
Aquatic macrophytes are an important food source for birds and mam-
mals. Fassett (1957) lists 36 species of waterfowl, nine marshbirds, four
shorebirds, and nine upland game birds that feed on these plants. He also
lists beaver, deer, moose, muskrat, and porcupines as aquatic macrophyte
herbivores. The use of macrophytes in bioassessment programs has nu-
merous advantages:
• Macrophyte taxonomy to the generic level is relatively straightforward.
• Because the establishment of macrophyte populations in a specific
habitat depends partly on local environmental conditions, they are
potentially very useful as site-specific indicators.
• Because their specific microhabitat structure does not limit germination,
macrophytes are potentially found in high population densities.
• The growth patterns of individual macrophytes are directly influenced
by herbivore activity.
• The longevity, distribution, and rate of their population growth may
directly reflect prevailing conditions.
¦ Macroinvertebrates. Macroinvertebrates are the visibly distinguishable
crustaceans, molluscs, insects, and other fairly large aquatic invertebrates.
Benthic macroinvertebrate assemblages are important indicators of local-
ized environmental conditions because they inhabit the degraded or con-
taminated resources and can be exposed to degradation directly
throughout their life history. Their characteristics can be regarded as a re-
flection of the integration of short-term environmental variability (Plafkin
et al. 1989). At sensitive life stages, they respond quickly to stress; how-
Benthic
macroinvertebrate
assemblages are '
important indicators
of localized
environmental
conditions.
57
-------
BIOLOGICAL CRITERIA;
Technical Guidance for Streams and Small Rivers
Fish assemblages
are well suited to help
define environmental
conditions because
fish inhabit the
receiving waters
continuously, and with
lifespans up to 10
years, they can easily
represent the
integrated historical
effects of chemical,
physical, and
biological habitat
factors.
ever, the overall assemblage responds more slowly. Other advantages of
using macroinvertebrates include the following;
• Sampling methods are well developed and require minimal personnel
and inexpensive gear.
t
• Macroinvertebrates play a major role in the nutritional ecology of
commercial and sport fisheries.
• Most streams support sufficient abundance levels for assessment.
• Molluscs, many species of Crustacea, and some insects are largely
immobile. As residential organisms, they are particularly valuable
indicators of site conditions over time.
• Many states have already performed background benthic surveys, have
personnel trained in benthic biology, and can often get assistance in
sampling from lay groups.
¦ Fish. Fish assemblages are well suited to help define environmental
conditions — either natural or impaired. Fish are long-lived and inhabit
the receiving waters continuously. With lifespans up to 10 years, they can
easily represent the integrated historical effects of chemical, physical, and
biological habitat factors (Ohio Environ. Prot, Agency, 1987), Power (1990)
found that fish exert significant influence on the food chain in lotic sys-
tems. More specific advantages of using the fish assemblage for bioassess-
ment (Karr et al. 1986; Plafkin et al. 1989) include the following:
• Fish are usually present in lotic systems except for some headwaters.
• Their populations generally include species that feed at a variety of
trophic levels.
• Species composition and dominants are relatively stable in most areas.
• The migration patterns and wide-ranging foraging behavior of some
fish allow investigators to accumulate effects from relatively large-scale
habitats.
• In comparison to other potential bioassessment groups, fish are
relatively easy to identify. •
• Autecological studies for many freshwater species are extensive, so
their life histories are relatively well known.
• Public, and therefore, legislative appreciation for fish is apparent in the
fishable goal of the Clean Water Act, the Endangered Species Act (50
percent of "endangered" vertebrate species are fish), and in more
specific commercial and sport fisheries legislation.
• Historical survey data are probably best documented for fish.
• Investigators can often get assistance from lay groups.
¦ Wildlife. Mammals, birds, reptiles, and amphibians can also provide
valuable information for bioassessment decisions. Croonquist and Brooks
(1991), applying the concept of response guilds, found that bird species
with high habitat specificity decrease with increasing habitat alteration.
5t
-------
' CHAPTER 4:
Conducting the Biosurvey
This approach has considerable potential for development of an avian in-
dex of biotic integrity. Birds have been shown to reflect the condition of ri-
parian systems.
Because amphibians live part "of their life cycle in an. aqueous or damp
environment, they are a link between the aquatic and terrestrial environ-
ments. They are also sensitive to littoral zone and riparian disturbances
and to changes in their food resources (macroinvertebrates and periphy-
ton). The latter may affect their fitness or force them to emigrate from the
home range to another foraging zone. Other advantages of including a
biosurvey of mammals, birds, and amphibians in biomonitoring programs
are the following:
• Their longer life spans make them well suited for evaluation of
cumulative effects.
• The relatively large body size of birds and their behaviors (e.g., singing)
allow visual and auditory observation to supply most of the necessary
information.
• Birds are sensitive to riparian alteration.
• Wildlife taxonomy is well understood.
• Many biomarkers — physical and chemical alterations in the species in
response to contamination — appear in these organisms, and an
increased likelihood for sublethal effects in non-emigrating individuals.
• Trapping techniques for small mammals are relatively straightforward,
and their tracks and droppings also provide easily attainable survey
data.
• The public is usually able to assist in conducting wildlife assessments.
Synthesis
Many bioassessment programs focus on a single assemblage for reasons of
regulatory iocus or mandate, available expertise, resource limitations, or
public awareness and interest. However, state agencies are encouraged to
incorporate more than one assemblage (e.g., fish and benthic macroinver-
tebrates) into their assessment programs. Biological programs that use two
or three assemblages and include different trophic levels within each
group (e.g., primary, secondary, and tertiary consumers) will provide a
more rigorous and ecologically meaningful evaluation of a system's bio-
logical integrity (U.S. Environ. Prot. Agency, 1990) and a greater range of
temporal responsiveness.
Impairments that are difficult to detect because of the temporal or spa-
tial habits or the pollution tolerances of one group may be revealed
through impairments in different species or assemblages (Ohio Environ.
Prot. Agency, 1987). Mount et al. (1984) found that benthic and fish assem-
blages responded differently to the same inputs in the Ottawa River in
Ohio. Benthic diversity and abundance responded negatively to organic
loading from a sewage treatment plant and exhibited no observable re-
sponse to chemical input from industrial effluent. Fish exhibited no re-
sponse to the organic inputs and a negative response to metals. In a more
Biological programs
that use two or three
assemblages and
. include different
trophic levels within
each group will
provide a more
rigorous and
ecologically
meaningful evaluation
of a system's
biological integrity
and a greater range
of temporal
responsiveness.
5.
-------
BlCLOCICAL criteria
Technic$1 Guidance for Streams and Small Rivers
recent assessment, the Ohio EPA found that distinct response signatures
(Yoder, 1991) in both fish and macroinvertebrate assemblages indicated an
adverse effect from the sewage treatment plant. Selection of aquatic com-
munity components that show different sensitivities and responses to the
same disturbance will help identify the nature of a problem (U.S. Environ.
Prot. Agency, 1990).
Selecting a single assemblage for assessment may provide inadequate
resolution for certain impacts that are highly seasonal in occurrence. Or-
ganisms having short life cycles may not reflect direct exposure to highly
variable impacts at critical times or when complex cumulative impacts are
present. Depending on the collection period, those organisms may provide
a false sense of ecosystem health if other assemblages of longer-lived
populations are under stress. In cases in which periodic pulses of contami-
nants may occur, long-lived populations may be slow to exhibit response,
whereas short-lived organisms may be severely affected.
The occurrence of multiple stressors and seasonal variation in the in-
tensity of stressors require that more than one assemblage be incorporated
into biocriteria programs whenever practical. Not all assemblages dis-
cussed here are in constant contact with the aquatic habitat component.
Those that are — the macroinvertebrates, macrophytes, fish, and periphy-
ton — will exhibit direct, and potentially more rapid, responses to water
resource degradation. The assemblage comprising mammals, birds, and
amphibians indicates the quality of the riparian corridor and can reflect lo-
cal land use impacts on the water resource.
Aquatic organisms respond to stress in a variety of ways ranging from
alterations in community composition and structure to increases or de-
creases in the biomass of a single or multiple species, or mortality. Fish
and drifting macroinvertebrates also exhibit avoidance behavior by seek-
ing refugia from short- and long-term disturbances.
Careful selection of taxonomic groups can provide a balanced assess-
ment that is sufficiently broad to describe the composition and condition
of an aquatic ecosystem, yet practical enough for use on a routine basis
(Karr et al. 1986; Lenat, 1988; Plafkin et al. 1989). When selecting commu-
nity components to include in a biological assessment, primary emphasis
should be given to including species or taxa that (1) serve as effective indi-
cators of high biological integrity, that is, those likely to live in unimpaired
waters, (2) represent a range of pollution tolerances, (3) provide predict-
able, repeatable results from consistent sampling, (4) can be readily identi-
fied by trained state personnel (U.S. Environ. Prot. Agency, 1990), (5) show
a consistent response to pollution stress, and (6) closely represent local, in-
digenous biota.
Technical Issues
The methods and procedures used in bioassessment programs should be
based on the study objectives and associated technical issues, including
the selection of the proper sampling period, sites, and sampling regime;
and the determination of the appropriate habitats to be sampled.
Aquatic organisms
respond to stress in a
variety of ways
ranging from
alterations in
community
composition and
structure to increases
or decreases in the
biomass of a single or
multiple species, or
mortality.
-------
. CHAPTER 4:
Conducting the Biosurvey
Selection of the Proper Sampling Periods
The ideal sampling procedure is to survey the biological community with
each change of season, then select the appropriate sampling periods that
accommodate seasonal variation. Such indexing makes the best use of the
biological data. It ensures that the sources of ecological disturbance will be
monitored and trends documented, and that additional information will
be available in the event of spills or other unanticipated events.
In this way, the response of the community to episodic events (e.g.,
chemical spills) can be assessed throughout the year. Seasonal impacts,
which may be highly variable, can be more effectively characterized
through more frequent sampling. Impacts from certain stresses may occur
or be "worst-case" at specific times of the year, and it may be important to
provide adequate documentation of the biological condition during these
times. EPA's Science Advisory Board (SAB) suggests that sampling should
— at a minimum — include the major components of the fall-winter and
spring-summer (or wet season-dry season) community structure. The
Florida Department of Environmental Protection has instituted a program
that encompasses sampling during two index periods that correspond to
this approach.
If some fish and invertebrate life cycles (e.g., spawning, growth, mi-
gration, and emergence) cause marked seasonal changes in stream assem-
blages, then each sampling season, will require a separate reference
database, metrics, and biocriteria. When such multiple index periods are
used, the operational costs, at least initially, may be considerably higher
than if surveys were conducted only once a year. Therefore, states must
weigh their needs and the long-term value of this information against
these costs. Seasonality must always be considered, and where possible,
year-round data should be developed even if it has to be phased in slowly
over time and as budgets allow.
The alternative, a single index period, will be deficient; it will not docu-
ment spills or other single episode or transitory events including stresses
that take place in other seasons. It should be selected only if seasonality is
not a factor in the program objectives. Still, the major or initial applications
of state biocriteria are likely to be assessment and management planning re-
lated to chronic habitat alteration and point and nonpoint sources. Such
chronic stress impacts are more efficiently assessed with a single index pe-
riod approach. Resident fish and benthic invertebrate assemblages integrate
stress effects over the course of a year, and their seasonal cycles of abun-
dance and taxa composition are fairly predictable within the limits of inter-
annual variability. Single season indexing also represents a cost savings
compared'to seasonal or more frequent sampling.
Given these considerations, state managers must choose the approach
most appropriate to their needs and budgets. They must avoid the tempta-
tion to spread multiseason sampling so thin that neither seasonal measure-
ments nor indexing are properly achieved. It is better to do a single index
period well than to do two poorly. Presuming, therefore, that most states
will initially design their biological criteria programs around single season
surveys, the following discussion emphasizes index period designs,
The optimal biological sampling period will be consistent with recruit-
ment cycles of the organisms from reproduction to emergence and migra-
The ideal sampling
procedure is to
survey the biological
community with each
change of season,
then select the
appropriate sampling
periods that
accommodate
seasonal variation.
State managers must
choose the approach
most appropriate to
their needs and
budgets.
61
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
The optimal
biological sampling
period will be
consistent with
recruitment cycles of
the organisms from
reproduction to
emergence and
migration, such that
the maximum amount
of information can be
derived from the data.
tion, such that the maximum amount of information can be derived from
the data. Optimal conditions for biological sampling can be defined as that
period of time during which the target assemblages have stabilized after
larval recruitment and subsequent mortality and the use of their niche
space is at its fullest. Where necessary, a compromise between biologically
optimal conditions and water and flow conditions appropriate for the
sampling gear must be made. Therefore, selection of the sampling period
should be based on efforts to
• minimize between-year variability resulting from natural events,
• maximize gear efficiency, and
• maximize target assemblage accessibility.
Field collections scheduled to correspond to the optimal biological
sampling period provide the most accurate assessment of community re-
sponse to adverse conditions over an annual cycle. Sampling during these
periods may not be logistically feasible, however, as a result of adverse
weather conditions, staff availability, scheduling constraints, or other fac-
tors. The nature of the suspected stressor is an especially important con-
sideration. An agency may be required to perform biological sampling
during periods of greatest environmental stress, such as low flow and
high temperature periods for point source discharges or high flow and
runoff periods for nonpoint source discharges.
Although an estimate of aquatic community structure during optimal
biological conditions should reflect the effect of, or recovery from, envi-
ronmental stress periods (Ohio Environ. Prot. Agency, 1987), assessment of
worst-case conditions may be needed under certain permitting regulations
or as a follow-up to sampling during biologically optimal periods in
which impairment was detected.
Ecological conditions and, thus, optimal sampling periods, vary sea-
sonally as a result of regional climate patterns and the life cycles of the bi-
ota. Seven major climatological regions are represented within the
contiguous United States (Fig. 4-3). The primary influence of seasonal
changes in temperature and rainfall on stream biota is on biological proc-
esses (e.g., production, growth, reproduction, distribution, and locomo-
tion). The level of biodiversity may also change seasonally. Even within an
ecological region, some scaling of the optimal collection period may be
necessary, depending on the elevation of the site, the habitat type, and
other broad environmental variables.
Temperature and rainfall are the principal weather factors influencing
the selection of sampling protocols and timing. Sampling will be impossi-
ble in frozen streams or during extreme high flows. Even subtle changes in
temperature and flow may preclude certain kinds of sampling by affecting
the equipment or the distribution of target assemblages.
The purpose of the biological sampling program (trend monitoring,
special studies) also influences the sampling protocol. Special studies may
be conducted at any time depending on need; but trend monitoring stud-
ies will focus on annual sampling events with varying sampling frequen-
cies. The most appropriate season for such collections is determined by
considering all technical and nontechnical factors. Technical factors in-
clude the selected assemblage, recruitment cycles, and severity of degra-
62
-------
CHAPTER 4:
Conducting the Biosurvey
A
¥
•.Boston.
New York. N.Y
Udelphu. Pj
w.
Dtnxef
Kjnsat
Atlanta
Q • HgWifl C0"«K»fOW» Q • o»s»»t
H * Hu-fiiO SuDUOO C*! jJ-^J •MViN
0 " M^hnndi
Figure 4-3.—Classification of U.S. climatological regions.
dation or contamination; nontechnical factors include such matters as lo-
gistics and personnel. From a practical standpoint, many states may select
a sampling period that includes the summer and early fall months.
The investigator must carefully define the objectives of a monitoring
program before these design issues can be resolved. Will specific questions
be answered by sampling during periods of optimal biological condition
or during periods of maximum impact? (These two periods may coincide.)
Seasonal considerations are important because community taxonomic
structure and the functional composition of some assemblages undergo
natural changes in each season and annual cycle.
Natural cycles may also be influenced by chemical or physical altera-
tions. From the traditional perspective of evaluating pollution impacts,
summertime low flow conditions are often chosen to assess effects from
point source discharges. Low flow conditions capture the effects of minimal
effluent dilution in combination with the natural stressors of low water ve-
locity and high temperature. Minimal effluent dilution occurs in summer
because the lower quantity of water decreases the ability of the receiving
waters to reduce the concentration levels of discharged compounds.
The effects of nonpoint source pollution on the aquatic community are
evaluated during the recovery period following high flow because these
effects are largely driven by runoff in the watershed. Nonpoint source
loadings are estimated using samples collected during periods of high
flow. Their actual effects, however, should be based on sampling outside
the flow extremes. The effect of regulated and minimum flows are a par-
ticular problem during the winter season in the western United States.
Regulated flows are a function of anthropogenic activity, usually associ-
ated with dams and reservoirs. Sampling activities should be avoided dur-
ing high and low extremes.
Special studies conducted by state agencies in response to specific
regulatory requirements or catastrophic events (e.g., oil spills) may not oc-
cur in an optimal season. In these situations, the data should be inter-
Special studies may
be conducted at any
time depending on
need; but trend
monitoring studies will
focus on annual
sampling events with
varying sampling
frequencies,
>3
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
The selection of an
appropriate sampling
season depends on
the seasonal
attributes of the
aquatic community,
but the administrative
issues of sampling
efficiency, safety,
regulatory
requirements, and
appropriate metrics
for data analysis are
equally significant.
preted through concurrent reference data or through a seasonal adjust-
ment to established reference data. If base biocriteria are established for a
reference database for a single season, then data collected from the test
sites during this season are directly comparable.
Two options are available for collections at test sites during seasons
other than that used for base criteria. First, selected reference stations can
be sampled concurrently with the test sites to provide baseline compari-
sons for data interpretation. Criteria established during the optimal season
represent a range of values that can be extrapolated to other seasons. In
this manner, a percentage of the reference may be acceptable as an alter-
nate criterion.
The second option may be to develop adjustments for an annual cycle.
This can be done through seasonal collections of the reference database to
document natural seasonal variation. Alternatively, a knowledge of sea-
sonal appearance and disappearance of particular forms can be used to
develop adjustments.
This discussion has focused on the seasonal attributes of the aquatic
community. The administrative issues of sampling efficiency, safety, regu-
latory requirements, and appropriate metrics for data analysis are equally
significant and must also be considered in light of the sampling objectives.
The following paragraphs consider the sampling protocol in relation to the
seasonal attributes of benthic, periphyton, and fish assemblages.
Benthos
Maximum information for a benthic community is obtained when most of
its populations are within a size range (later instars) that can be retained
during standard sieving and sorting and be identified with the most confi-
dence. Reproductive periods and different life stages of aquatic insects are
related to the abundance of particular food supplies (Cummins and Klug,
1979). Peak emergence and reproduction typically occur in the spring and
fall, although onset and duration vary somewhat across the United States.
During peak recruitment of the young, approximately 80 percent are too
small to be captured in sufficient numbers to characterize the community
accurately, and the food source requirements for early instars may be dif-
ferent from those for later inst.ars. Therefore, the biologically optimal sam-
pling season occurs following the period of initial recruitment and high
mortality of young, and when the food resource has stabilized to support
a balanced indigenous community.
The comparative time frames for sampling the benthic community are
illustrated in Figure 4-4. The seasonal timetable shows annual high and
low flow periods, emergence peaks for aquatic insect communities, and
biologically optimal sampling periods (BOSP) for a stream in the New
England region. High and low flow correspond to periods of high and low
rainfall and associated runoff. Emergence is triggered by average daily
temperature and photoperiod and usually occurs at peak intervals in
spring and fall. The biologically optimal sampling period falls between the
peaks in late winter and late summer and occurs after the population has
been exposed to two-thirds of the aquatic phase of the organism's life cy-
cle measured in degree days (that is, in units calculated as the product of
time and temperature over a specified interval).
64
-------
CHAPTER 4:
Conducting the Biosurvey
High
Flow
Recruitment
Emergence
Low Flow / Low Temp. (Ice)
High
Flow
Low Flow
High
Temp.
Figure 4-4.—Biological and hydrological factors for sampling period selection In the
Northeast (macrolnvertebrates). The gray area is the overlap between emergence and
recuitment.
In this example (Fig.4-4), sampling in July and early August satisfies
most of the criteria for collecting a representative sample at a time of sig-
nificant chemical contaminant stress. It should be noted that chronic non-
point source impacts such as sedimentation will be reflected in the quality
of the benthic community after flow has returned to near normal follow-
ing high flow conditions.
In the context of a single population, seasonality may be a significant
factor. The early instars are small and difficult to identify, and the young
nymphs have a generalized feeding strategy of collecting and scavenging.
Only in later instars does feeding specialization occur and the quality of
the food source become reflected in the condition of the population. In the
case of Stenonema, the middle and late instars specialize as scrapers. Scrap-
ers are often considered a pollution sensitive functional feeding group be-
cause their food source — diatom algae — responds to the early effects of
pollution within the stream.
Periphyton
Periphyton assemblages are associations of algae, bacteria, and fungi that
colonize the substrates in a stream. For purposes of bioassessment, most
periphyton evaluations focus on diatom algae. The periphyton assem-
blage exhibits different seasonal abundance patterns than fish or benthos.
The key difference is that periphyton assemblages are sufficiently abun-
dant to be collected year-round from streams in temperate zones. Their
biologically optimal sampling period may be based on relatively stable
conditions but must also account for the comparison of diatom assem-
blages within similar stages of seasonal succession.
The limiting factors for diatoms are light, temperature, nutrients,
water velocity, grazing, and interactions among algae via metabolites. Ob-
viously, the abiotic factors go through an annual cycle of change and, like
benthos, the assemblage composition shifts as the changing conditions fa-
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
vor new species. This process of seasonal succession creates significant
seasonal differences in periphyton assemblages that must be considered in
developing a study design. Besides changes in periphyton species compo-
sition, additional seasonal issues must be controlled to compare collections
among sites and annual trends. Two major considerations are (1) the dif-
ferences in biomass related to light and temperature regimes and (2) the
comparisons of periphyton assemblages that have been subjected to heavy
rains and scour with those that have matured under more stable hydro-
logical conditions. Differences in light and temperature regimes may re-
flect human influences, for example, alterations of the stream channel and
removal of riparian vegetation.
Fish,
Like periphyton and benthic invertebrates, the fish fauna at a site is likely
to vary seasonally. In the Northwest, for example, annual spawning mi-
grations of anadromous salmonids set in motion a seasonal cycle of major
importance to the biota. Seasonal migrations of fish are less striking but
common in other areas as well. Most frequently, fish movements involve
upstream movements in search of spawning areas to serve as* nesting and
nursery areas for young fish. Upstream areas often provide richer food
supplies and lower predation rates than downstream areas.
Because of geographic variation in flows and temperatures, no general
pattern occurs across all regions. A seasonal timetable representative of
physical conditions and fish assemblage activities in the New England re-
gion is illustrated in Figure 4-5. Unless the sampling objective includes the
study of unusual flow conditions and concurrent biotic responses, field
sampling protocols should avoid extreme flow conditions (low or high)
that may represent unusual stress, assemblage instability, or result in dan-
ger to field crews.
Sampling in several regions of the country has demonstrated that opti-
mal fish sampling periods can be defined with relative ease. Generally,
sampling periods should follow the spring spawning migrations that coin-
Low Flow / Low Temp. (Ice)
JAN
DEC
FEB
High
Flow
NOV,
IAR
High
Flow
OCT
SEP
MAY
JUN
Coldwater
Fish Spawning ¦¦¦
Anadromous
Migration I I
Warmwater
Fish Spawning l
Low Flow
High
Temp.
JUL
Figure 4-5.—Biological and hydrologlcal factors for sampling period selection In the
Northeast (fish).
66
-------
CHAPTER 4:
Conducting the Biosurvey
cide with periods of high flow. Most states in eastern North America select
the summer period for sampling (June through August) to coincide with
periods of low to moderate stream flow and avoid the variable flow condi-
tions of early spring and autumn (Karr et al. 1986). Fish assemblages dur-
ing summer are relatively stable and contain the full range of resident
species, including all major components of age-structured populations.
Angermeier and Karr (1986) have outlined sampling rationale, including
the merit of excluding young-of-the-year (YOY) from spring and late sum-
mer samples. This exclusion reduces variability and the problem of identi-
fying and sampling very small fry. Excluding YOY from most analyses
improves reliability and does not weaken the interpretation of the sys-
tem's condition.
The scenario presented in Figure 4-5 identifies high and low flow peri-
ods in early spring and late summer for streams in the northeastern
United States. The number of species is likely to peak in the spring with
the spawning migration; the number of individuals .will peak in the early
autumn with the addition of YOY. The biologically optimal sampling pe-
riod (BOSP) corresponds to seasonal effects within the fish assemblage
and to the flow dynamics that influence sampling efficiency. Because the
physical condition of the streams affects the efficiency of fish sampling
gear, it also affects the nature or quality of the resulting data. For example,
the effectiveness of passive equipment (e.g., trap nets) can be substantially
reduced during periods of high or low flow, and the efficiency of active
equipment (e.g., electrofishing gear) is reduced by turbidity, water tem-
perature, and conductivity.
Sampling can typically begin in May or June in most areas and pro-
ceed into September unless unusually low flow periods occur during late
summer drought. The probability that low flow periods will, occur in late
summer increases in watersheds that have been severely modified by ur-
banization or agricultural land use, in which case low flow sampling
should be avoided.
Decisions about
which habitats to
sample are critical to
the success of a
biocriteria program.
Selection of Habitat for Aquatic Assemblage Evaluations
Stream environments contain a number of macro-and microhabitat types,
including pools, riffles, and raceways, or surface and hyporheic zones. The
latter refers to regions of saturated sediment beneath or beside the stream
(Lincoln et al. 1982). Larger rivers have even more complex habitat con-
figurations. Because no single sampling protocol can provide accurate
samples of the resident biota in all habitats, decisions about habitats are
critical to the success of a biocriteria program. These decisions are usually
made in concert with the decision about the assemblages to be sampled,
the sampling methods to be used, and the seasonal pattern of sampling.
Selection of habitats for sampling may be influenced by institutional
requirements, such as sampling and analysis protocols that are part of an
existing monitoring program, or the need to develop data that are consis-
tent with a historical database; however, historical approaches should not
be retained without careful evaluation of their ability to provide the data
necessary to make informed resource decisions in future years.
Periphyton, invertebrates, and fish species in a stream vary in their
distribution among major habitats. Depending on the data quality objec-
tives established for the specific project or program, one or more assem-
67
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
A major
consideration in the
development of
bioassessment '
procedures is
whether sampling all
habitats is necessary
to evaluate biological
integrity or whether
selected habitats can
provide sufficient
information.
blages may be targeted for inclusion in biosurvey activities. Attributes of
several potential assemblages and their several advantages were de-
scribed earlier in this chapter.
A major consideration in the development of bioassessment proce-
dures is whether sampling all habitats is necessary to evaluate biological
integrity or whether selected habitats can provide sufficient information.
The selection of single habitat over multiple habitat, or vice versa, influ-
ences study design and may influence selection of the biotic assemblage to
be sampled. Some taxa include individuals whose mobility or natural spa-
tial distribution requires multiple habitat sampling.
Generally, fish sampling reduces the need to make more detailed habi-
tat decisions because most fish in small to medium rivers can be sampled
using seines or electrofishing methods that efficiently sample all major
surface water habitats except hyporheic zones and bank burrows. By sam-
pling the full diversity of stream habitats for fish, the importance of fish
movements among microhabitats for resting and foraging is reduced. Effi-
cient sampling of all local habitats limits the problem of correcting evalu-
ations of taxa in case the intensity of sampling varies among the range of
available habitats.
Habitats to be sampled for periphyton require different analytical ap-
proaches. For example, periphyton assemblages may develop more easily
on rigid or hard substrates. Though periphyton can grow on the leaves
and stems of macrophytes, more prolific growths are generally seen on the
hard surfaces of large substrate particles (e.g., cobble or small boulders).
Steinman and Mclntire (1986) found-that substrate type is one of several
characteristics that affect the taxonomic structure .of lotic periphyton as-
semblages. Other factors are the dispersal and colonization rates of taxa in
the species pool, competitive interactions, herbivory, chemical composi-
tion of the environment, and the character of ecological disturbances. Be-
cause it is difficult to remove or collect periphyton from natural substrates
(Austin et al. 1981), hard surfaces (either natural or artificial) are usually
the focus of sampling efforts. Most strategies for sampling periphyton as-
semblages are single habitat though other variables introduce additional
complexity.
Benthic macroinvertebrates inhabit various habitats in lotic situations,
for example, riffles, pools, snags, or macrophyte beds. Complete charac-
terization of the assemblage requires a multihabitat and multisampling
protocol such as that advocated by Lenat (1988). The benthic macroinver-
tebrate protocols for rapid bioassessment advocated by Plafkin et al.
(1989) were developed for sampling the most productive and dominant
benthic habitat in wadable streams. Consequently, riffles and cobble sub-
strate were the primary focus of the rapid bioassessment protocols be-
cause that habitat is predominant across the country.
This approach works for small streams and streams that are domi-
nated by riffles; however, it requires additional evaluation and technical
development for use in other habitats. Plafkin et al. (1989) argue that the
habitat where riffles predominate, will often be the most productive and
stable habitat for the benthic community. The production of the habitat is
related to provision of refugia, food resources, and necessary community
interactions. It may be necessary to document the extent and character of
the habitat because streams differ in these qualities, which differences may
68
-------
CHAPTER 4:
Conducting the Biosurvey
be related to natural and anthropogenic- causes. In some streams, riffles are
not a dominant feature, and the emphasis on them may be misleading.
Since the issuance of the Rapid Assessment Protocols (RBPs) in 1989,
rapid assessment techniques have evolved to focus on sampling of more
than one habitat type, usually in the proportion of their representation at
the sites of interest. These techniques have been primarily designed for
low gradient streams (Mid-Atlantic Coastal Streams Workgroup, 1993;
Florida Dep. Environ. Prot. 1994) and encompass the sampling of four or
five habitat categories.
The sampling of a single habitat type (e.g., riffles or runs) is intended
to limit the variability inherent in sampling natural substrates and to en-
hance the evaluation of attributes in an assemblage that will vary substan-
tially in various habitats. Double, composited square meter kick net
samples (2 m2) are used in RBPs to collect large representative samples
from riffle or run areas. Other gear can also be used to collect such com-
posite samples.
Multihabitat sampling allows the evaluation of a broad range of effects
on the benthic assemblage. However, it may also introduce variability into
comparisons of the benthic assemblage among sites. Multihabitat investi-
gations of water resource integrity are potentially confounded by (1) the
absence of a particular habitat at a station, and (2) the potential differences
in the quality and quantity of a habitat. As more habitats are sampled, the
more difficult it is to control for comparable habitat among sites; and the
absence of a habitat type at one or more stations exacerbates the problem.
However, some states, such as North Carolina, have been successful in us-
ing a multihabitat sampling approach and advocate this technique as be-
ing more appropriate than simply sampling the riffle or run (Lenat, 1988).
A case study in association with the North Carolina Department of En-
vironmental Management addressed the issue of sampling strategy and
indicated that the riffle assemblage and the multihabitat assemblage re-
' sponded similarly to differences among stations (Plafkin et al. 1989). For
example, under stress, taxa richness was reduced by the same proportion
in both the riffle and the multihabitat assemblage samples at a given sta-
tion. These responses suggest that either the riffle assemblage or the multi-
habitat assemblage can be used to assess biotic integrity in streams in
which riffles are prevalent.
Kerans et al. (1992) examined patterns of variability and the contribu-
tion of pool versus riffle invertebrate samples to the evaluation of biotic
integrity and the detection of different kinds of degradation. They evalu-
ated over a dozen attributes of the invertebrate assemblages including
numbers of species (total and for a number of taxa) as well as several eco-
logical classifications. At least eight attributes exhibited spatial or tempo-
ral trends, or both, depending on whether the habitat was pools or riffles.
Attributes that were temporally and spatially unpredictable included
some that are most commonly used in stream bioassessment. Kerans et al.
conclude that measures of human impact on biotic integrity may be biased
if sampling is restricted to only one habitat.
The choice of sampling habitats also entails a choice of sampling meth-
ods because conventional sampling methods for invertebrates vary in
their efficiency among habitats. Surber and Hess samplers are used for rif-
fles, while grab samplers are used most efficiently in the soft substrate of
The choice of
sampling habitats
also entails a choice
of sampling methods.
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
In either the single
habitat or muitihabitat
approach, the most
prevalent and
physically stable
habitat that is likely to
reflect anthropogenic
disturbance in the
watershed should be
chosen.
The habitat with the
most diverse fauna is
preferred — riffles
followed by hard,
coarse substrates,
snags, aquatic
vegetation, and soft
substrates.
pool habitats. Several forms of net samplers have been developed for vari-
ous stream habitats: kick nets or seines (Plafkin et at. 1989; Lenat, 1988), D-
frame nets (Montana Dep. Health Environ. Sci,, 1990), and slack
(rectangular frame) samplers (Cuffney et al. 1993). Passive colonization-
dependent samplers (e.g., Hester-Dendy samplers) may also be used for
evaluation of invertebrate assemblages (Ohio Environ. Prot. Agency, 1987).
Substrate Choices
In either the single habitat or muitihabitat approach, the most prevalent
and physically stable habitat that is likely to reflect anthropogenic distur-
bance in the watershed should be chosen. These habitats will vary region-
ally because of differences in topography, geology, and climate. The
biological community in a particular stream may also change in response
to increasing stream size (Vannote et al. 1980). The key to sampling, perti-
nent to benthic invertebrate surveys, is to select the habitats that support a
similar assemblage of benthos within a range of stream sizes. Habitats that
have been used for benthos are riffles, snags, downed trees, submerged
aquatic vegetation, shorezone vegetation, and sediments, such as sand,
silt, or clay (Table 4-2).
The habitat with the most diverse fauna is emphasized by most inves-
tigators because it offers the highest probability of sampling the most sen-
sitive taxa. Riffles usually fit this criterion, and when present, are
preferred. This habitat type is followed by hard, coarse substrates, snags,
aquatic vegetation, and soft substrates. If multiple habitats are selected,
similarity in habitat quality and comparable levels of effort among sam-
pling sites must be considered.
Natural and Artificial Substrates
Most benthic surveys employ direct sampling of natural substrates. This
method is particularly important if habitat alteration is suspected as the
cause of impairment. A major assumption is that every habitat has a bio-
logical potential, which is reflected in the resident biotic community. Be-
Table 4-2.—Common benthic habitats.
SNAGS/DOWNED TREES
SHOREZONE VEGETATION
• Productive in blackwater streams
• Present in most streams
(Benke st al. 1984)
• Diversity of epifauna
» Measures riparian impacts
• Community dependent on
* Dominated by shredders and collectors
well-prepared substrate
« May be seasonal
SUBMERGED AQUATIC VEGETATION
SILT/MUD
• Productive in coastal zones
• Pool communities
• High standing crop
* Dominated by fauna
• Seasonal habitat
• Sediment quality and water quality effects
• Snails usually abundant
• Fauna usually tolerant to iow oxygen
SHIFTING SAND
LEAF LITTER/DEBRIS
* Prevalent In erosionai areas
• Prevalent in forested streams
• Dominated by opportunistic infauna
« Measures riparian impacts
~ Sediment quality and water quality effects
*' Dominated by shredders
~ High dominance by monotypic fauna
• Microbial preparation of substrate
70
-------
CHAPTER 4;
Conducting the Biosurvey
cause interpretation depends on the level of assemblage development
within the existing habitat, sampling natural substrates is recommended.
If, however, an artificial substrate can be matched to the natural substrate
(e.g., using a rock basket sampler in a cobble substrate stream), then such
artificial substrates may also be used (Sci. Advis. Board, 1993). Maine uses
this rock basket approach. The Ohio EPA biocriteria program (Ohio Envi-
ron. Prot. Agency, 1987) has successfully used Hester-Dendy multiplate ar-
tificial substrate samplers supplemented by qualitative, natural substrate
samples to assess biological integrity using benthic assemblages.
The advantages and disadvantages of artificial substrates (Cairns,
1982) relative to natural substrates are the following:
¦ Advantages of Sampling with Artificial Substrates
1. Enhances sampling opportunities in locations that are difficult to
sample effectively.
2. Permits standardized sampling by eliminating subjectivity in
sample collection technique.
3. Minimizes confounding effects of habitat differences by providing
a standardized microhabitat.
4. Directs the interpretation to specific water quality questions
without interference of habitat variability.
5. Increases the ease of placing samplers in discrete areas to discrimi-
nate impacts associated with multiple dischargers.
¦ Disadvantages of Sampling with Artificial Substrates
1. Requires the investigator to make two trips for each artificial
substrate sample (one to set and one to retrieve).
2. Measures colonization potential rather than resident community
structure.
3. Allows problems such as sampler disturbance and loss to occur.
4. Complicates interpretation of the effects of habitat structure.
If artificial substrates are selected, the surface area of the materials
should be standardized among units. Introduced substrates, in the context
of biological monitoring, are artificial substrates that are constructed to
match natural bottom materials at the site of the survey. An example of in-
troduced substrates are rock baskets, such as those used by Maine (Davies
et al. 1991), in which baskets that contain rocks native to the region of
known surface area are partially buried in the bottom sediment. Where
possible, the use of introduced substrate is preferable to other types of ar-
tificial substrate as recommended by the SAB (1993). Rock baskets or other
substrates should be placed in waters of similar depths, velocities, and
daily sun and shade regimes.
71
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Standard operating
procedures should be
adhered to in all
phases of fieldwork,
data analysis, and
evaluation. Such
standards are
essential for
maintaining
consistency and
comparability among
data sets and for
appropriate quality
assurance and control.
Standardization of Techniques
Standard operating procedures should be adhered to in all phases of field-
work, data analysis, and evaluation. Such standards are essential for main-
taining consistency and comparability among data sets and for
appropriate quality assurance and control (Kent and Payne, 1988; Klemm
et al. 1990; Smith et al. 1988). Without standard operating procedures to
mimic previous studies, the difficulties encountered in comparing tempo-
ral and spatial data or analytic results may be substantial. The inherent
variability of the sampling process (Cairns and Pratt, 1986) can be reduced
through standardization of sampling gear, gear efficiency, level of effort,
subsampling methods, handling and processing procedures, and com-
puter software. Standardization of project activities provides considerable
strength in reducing, controlling, and understanding variability.
Sample Collection
A major influence on the comparability of field ecological projects is the
type and intensity of appropriate training and professional experience for
all personnel (Barbour and Thornley, 1990). Similar exposure to sampling
methods and standard operating procedures can reduce the amount of
variation from one sampling event or project to the next. Standardizing
the equipment relative to operator efficiency, sampling effort, and the area
to be sampled greatly affects data quality. Operator efficiency depends on
the operator's experience, dexterity, stamina, and adherence to specified
survey requirements. Physical habitat conditions at the time of sampling
(e.g., flow levels, current velocity, and temperature) also influence effi-
ciency. Active sampling efforts (e.g., using net samples or electrofishing)
may be standardized as a function of person-hours spent at each sampling
station and by tracking the physical area or volume sampled. Passive
methods (e.g., artificial substrates, trap nets) may be standardized by
tracking the person-hours and the exposure time. This choice is often dic-
tated by the earlier selection of the assemblage to be sampled; for some, a
relatively small selection of sampling techniques may be available. A cer-
tain sampling area or volume may be required to obtain an appropriate
sample size from a particular community and to estimate the natural vari-
ability of that community at the sampling station.
Once the assemblage, sampling equipment, and method have been cho-
sen, standard operating procedures can be written for field operations, in-
cluding a clear description of the sampling effort to be applied during each
sampling event. All employees should have this documentation, and new
employees should be accompanied in the field by experienced staff until they
are thoroughly familiar with all procedures (Ohio Environ. Prot. Agency,
1987).
Processing samples in the field requires several critical steps. Sample
containers for benthic invertebrates and voucher fish should be marked
with appropriate and complete information on internal and external la-
bels. Other identifying information and descriptions of visual observa-
tions should be recorded in a field notebook.
Data on birds and mammals, which consist primarily of visual obser-
vations and for which accurate field taxonomy is possible, will not require
subsequent processing in the laboratory. However, the details of each ob-
72
-------
CHAPTER 4:
Conducting the Biosarvey
servation should be carefully recorded so that they may be checked later.
Most fish sampling requires sorting, recording, and releasing the fish at
the site of capture. Fish sampling crews should have a reference collection
available in the field, and specimens should be collected and accurately la-
beled so that identifications can be confirmed.
Sample containers with preserved specimens should be assigned
unique serial or identification numbers. These numbers should be re-
corded in a logbook along with the appropriate labeling information. All
sample containers or specimens should be appropriately packaged for
transportation and continued processing in the laboratory.
For assemblages in which extremely large numbers of individuals or
associated substrate are obtained in each sample as is often the case with
small fish, benthic macroinvertebrates, periphyton, or planktonic organ-
isms, it may be impractical and costly to process an entire sample. In such
cases, standardized random subsampling, similar to that recommended by
Plafkin et al. (1989), is a valid and cost-effective alternative.
As a subsampling method is developed, every attempt must be made
to reduce bias. Therefore, guidelines are needed to standardize the effort
and to eliminate investigator subjectivity. Rapid bioassessment protocols,
for example, maintain subsampling consistency by defining the mode (a
gridded pan), by placing limitations on the mechanics of subsampling and
the subsample size, and by assuring that the subsampling technique is
consistently random.
Standardized
random subsampling
is a valid and
cost-effective
alternative to
processing an entire
sample. As a
subsampling method
is developed, every
attempt must be
made to reduce bias.
Sample Processing
The need for specialized training and expertise is most necessary during
the identification of organisms. Unless the project objectives direct other-
wise, each specimen should be identified to the most specific taxonomic
level possible using current literature. Some techniques may require iden-
tification only to the ordinal, familial, or generic level (Ohio Environ. Prot.
Agency, 1987; Plafkin et al. 1989), but the most accurate information on tol-
erances and sensitivities is found at the species level.
Nevertheless, taxonomic resolution should be set at a level achievable
by appropriately trained state personnel. State water resource agencies
should find it beneficial to establish collaborative working arrangements
with local and regional experts who can provide training, technical sup-
port, and quality assurance and control. Stream ecology research over the
last decade indicates that a specific minimal level of resolution should be
set (i.e., the "lowest achievable taxonomic level" is not a helpful criterion)
and that additional refinement should be left to individual state groups as
their capabilities permit (Sci. Advis. Board, 1993).
The SAB further states that proposed levels of intensity and taxonomic
resolution must receive a thorough evaluation by the scientific research
community. For example, adult and juvenile fish should usually be identi-
fiable by species (Sci. Advis. Board, 1993). The identification of larval fish
may provide useful information; however, it may only be feasible to iden-
tify them to the generic or familial levels. Reasonable candidate levels for
stream macroinvertebrates are given in Table 4-3.
Once the samples have been analyzed (identified, enumerated, and
measured), reference (voucher) material should be placed in the well-estab-
73
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Table 4-3.— Proposed minimal levels of taxonomlc resolution for stream
macroinvertebrates (taken from Sci. Advis. Board, 1993).
TAXONOMIC LEVEL
GROUPS
Genus
Plecoptera (in part), Ephemeroptera, Odonata, Trichoptera,
Megaloptera, Neuroptera, Lepidoptera, Coleoptera (in part,
larvae and adults), Hemiptera, Diptera (Tipulidae and
Simulidae), Crustacea, Mollusoa
Tribe
Chironominae
Subfamily
Chironomidae .
Family
Diptera {other than Chironomidae, Tipulidae and Simulidae),
Oligochaeta, Plecoptera (in part), Coleoptera (In part)
Order
Other nontnsect groups
lished network of federal, state, and university museums for regionally cen-
tralized curation (Sci. Advis. Board, 1993). This action ensures a second level
of quality control for specimen identification. Preferably, collection and
identification of voucher specimens will be coordinated with taxonomic ex-
perts in regional museums. These repositories, which have always been the
centers for systematics, should continue to be used for this function (Sci.
Advis. Board, 1993). The SAB recommends that once the information on the
samples has been entered into a database and verified, the repository insti-
tutions 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.
AE identifications should be made using the most up-to-date and ap-
propriate taxonomic keys. Verification should be done in one of two ways:
(1) by comparison with a preestablished reference or research specimen
collection, or (2) by having specimens confirmed by taxonomic experts fa-
miliar with the group in question (Borror et al. 1989). A regional consensus
of taxonomic certainty is critical to ensure that the results are comparable
both spatially and temporally. The taxonomists should always be con-
tacted by telephone or mail before any specimens are sent to their atten-
tion. It is also important to follow their advice on the proper methods for
packing and shipping samples. Damaged specimens may be useless and
impossible to identify.
Suggested Readings
Hart, D. (editor). 1990. Proc. Third Annual Ecological Quality Assurance Workshop. U.S.
Environ. Prot. Agency, Can. Min. Environ,, Burlington, Ontario.
Karr, J.R. et al. 1986. Assessing Biological Integrity in Running Waters: A Method and Its
Rationale. Spec. Publ. 5. Illinois Nat. History Surv., Urbana, IL.
Klemm, D.J., P.A. Lewis, F. Fulk, and J.M. Lazorchak. 1990. Macroinvertebrate Field and
Laboratory Methods for Evaluating the Biological Integrity of Surface Waters.
EPA/600/4-90-030. Off. Res. Develop., U.S. Environ. Prot. Agency, Washington,
DC.
74
-------
CHAPTER 4:
Conducting the Biosurvey
Mid-Atlantic Coastal Streams Workgroup. 1993. Standard Operating Procedures and
Technical Basis: Macroinvertebrate Collection and Habitat Assessment for Low-gra-
dient Nontidal Streams. Draft Rep. Delaware Dep. Nat. Res. Environ. Conserv.,
Dover.
Ohio Environmental Protection Agency. 1987. Biological Criteria for the Protection of
Aquaitic Life. Volume 3: Standardized Biological Field Sampling and Laboratory
Methods for Assessing Fish and Macroinvertebrate Communities. Monitor. Assess.
Prog., Surface Water Sec., Div. Water Qual., Columbus, OH.
. 1990. The Use of Biocriteria in the Ohio EPA Surface Water Monitoring and As-
sessment Program. Columbus, OH.
U.S. Environmental Protection Agency. 1980b. Interim Guidelines and Specifications for
Preparing Quality Assurance Project Plans. QAMS-005/80. Qual. Assur. Manage.
Staff, Off. Res. Dev., Washington, DC.
. 1984c. Guidance for Preparation of Combined Work/Quality Assurance Project
Plans for Environmental Monitoring. Rep. OWRS QA-1. Washington, DC.
. 1989. Preparing Perfect Project Plans. A Pocket Guide for the Preparation of
Quality Assurance Project Plans. EPA/600/9-89/087. Risk Reduction Eng. Lab.,
Off. Res. Dev., Cincinnati, OH.
— . 1990. Biological Criteria: National Program Guidance for Surface Waters. EPA-
440/5-90-004. Off. Water, Washington, DC.
75
-------
Intentionally Blank Page
-------
CHAPTER 5.
Evaluating
Environmental Effects
^^hould a biological survey reveal a significant departure from reference
conditions or criteria, the next step is to seek diagnostic information
leading to remedial action. This action entails the investigation of an array
of physical, chemical, and biological factors to determine the likely source
of degradation in the water resource.
Five major environmental factors affect and determine water resource '
integrity (Karr and Dudley, 1981; Karr et al. 1986). These factors are water
quality, habitat structure, flow regime, energy source, and biotic interac-
tions. Monitoring programs must integrate, measure, and evaluate the in-
.Purpose:
To provide managers
with an understanding
of the factors that
affect and determine
water resource
integrity.
fluences of these factors (Fig. 5-1). A comprehensive discussion of all five
and the enormous variety of human actions that alter them is beyond the
scope of this document. We can, however, present a conceptual sketch of
each one and how it influences the integrity of the water resource. Several
considerations are involved in evaluating these complex factors.
Human actions often alter one or more of those factors and thus alter
the resident biota. Alterations may be obvious, such as the extinction of
species or the introduction of exotics, or they may be more subtle, such as
altered survival rates, reproductive success, or predation intensity. Protec-
tion or restoration of biotic integrity requires identification of the proc-
esses that have been altered by human actions. Careful evaluation of the
conditions in a watershed can play a critical role in identifying the poten-
tial causes of degradation. That identification process is essential to de-
velop the most cost-effective approaches to improving the quality of water
resources.
¦
Water Quality
The physical and chemical, attributes of water are critical components of
the quality of a water resource. Because the earliest water resource legisla-
tion (e.g., the Refuse Act of 1899) dealt with disease and oil pollution in
navigable waters, emphasis has traditionally been on the physical and
chemical properties of water. Physical and chemical attributes of special
concern include but are not limited to temperature, dissolved oxygen, pH,
hardness, turbidity, concentrations of soluble and insoluble organics and
inorganics, alkalinity, nutrients, heavy metals, and an array of toxic sub-
stances. These substances may have simple chemical properties, or their
77
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
ECOLOGICAL
IMPACT OF
HUMAN-INDUCED
ALTERATIONS
1. Energy Source
TVpa, amount, and partieJa
*iz» of organic material
antarinc a straam frwn „
tha riparian zone varaua
primary production In tha
strum
Seasonal pattern of available
energy
2. Water Quality
Temperature
rvmatf
Dissolved oxygen
Nutrients (primarily nitrogen
and phosphor"
Organic and inorganic
chemicals, natural and synthetic
Heavy metals and toxic
substances
pH
3. Habitat Structure and Quality
Stintrate type and quantity
Water dapth and cunrt
velocity
Spawning, nursery, and
hieing places
Diversity (pools, Mas,
woody debris)
4. Flow Regime
Watar volume
Temporal distribution of
foods and low flow*
Flow regiiation
S. Biotic Interactions
Competition
Preda&on
Disease
Parasitism
Decreased coarse particulate organic matter
' Increased fine particulate organic matter
Increased algal production
Expanded temperature extremes
Increased turbidity
Altered diurnal cycle ofdssolved oxygen
Increased nutrients (especially soluble
nitrogen and phosphorus)
Increased impended solids
Decreased stability of substrate and banks
due to erosion and sodmentation
Mora uniform water depth
Reduced habitat heterogeneity
_ Decreased channel sinuosity
Reduced habitat area due to shortened channel
Decreased instream cover and riparian vegetation
Altered flout extremes (both magnitude and
frequency at high and Sew fkws)
Increased maximum fkM wkxity
.Decreased minimum ficw velocity
Reduced diversity of microhabltat velocities
Fewer protected sites
Increased frequency of diseased fish
Altered primary and secondary production
Altered bophlc structure
- Mend decomposition rates ami String
Disruption of seasonal rhythms
Shifts In species composition and relative
abundance
Shifts In imertebrate functional groups
(Increased scrapers and decreased shredders)
Shifts In trophic gtilds (increased omnivons
and decreased piscivores)
Increased frequency of fish hybridization
Figure 5-1.—Five major classes of environmental factors that affect aquatic biota in lotlc systems. Right column lists
selected expected results of anthropogenic perturbation (Karr et al. 1986).
-------
CHAPTER 5:
Evaluating Environmental Effects
dynamics may be complex and changing, depending on other constituents
in a particular situation including the geological strata, soils, and land use
in the region. The number of elements and compounds that influence
water quality is very large without human influences; with them, the com-
plexity of the problem is even greater. The human effects on biological
processes may be direct (i.e., they may cause mortality), or they may shift
the balance among species as a result of subtle effects, such as reduced re-
productive rates or changing competitive ability. Aquatic life use designa-
tions provide protection at various levels from the multitude of
anthropogenic effects.
The EPA encourages states to fully integrate biological surveys, whole-
effluent and ambient toxicity testing, and chemical-specific analyses to as-
sess attainment or nonattainment of designated aquatic life uses in state
water quality standards (U.S. Environ. Prot, Agency, 1991c). Ohio EPA
used numeric biological criteria within an existing framework of tiered
aquatic life uses to establish attainable, baseline expectations on a regional
basis (Yoder, 1991). Use attainment status in the Ohio water quality stand-
ards results in a classification of "full attainment," if all applicable numeric
biocriteria are met; "partial attainment," if at least one aquatic assemblage
exhibits nonattainment but no lower than a "fair" narrative rating; and
"nonattainment," if none of the applicable biocriteria are met, or if one as-
semblage reflects a "poor" or "very poor" narrative rating.
North Carolina's Department of Environment, Health, and Natural Re-
sources has used in-stream biota to assess water quality since the mid-
1970s (Overton, 1991), and the water quality regulations in the North
Carolina code have been revised to take biological impairment into ac-
count. In addition, when fiscal realities in North Carolina required a more
efficient water quality program, all NPDES permits within a given river ba-
sin were scheduled to be issued within the same year (Overton, 1991). The
same strategy makes biological assessment more efficient because the de-
partment can focus the assessment on specific river basins coincident with
the renewal permits. Other states may have to consider similar strategies to
conserve resources. *
The Maryland Department of the Environment, Water Quality Moni-
toring Division, uses biological assessment as part of a. statewide. water
quality monitoring network (Primrose et al. 1991). Using biological assess-
ment, Maryland has been able to differentiate among various degrees of
impairment* and unimpairment, and to distinguish particular water qual-
ity impacts.
The Arkansas Department of Pollution Control and Ecology devel-
oped a bioassessment technique in the mid-1980s to assess the impact on
receiving waters of discharges exceeding water quality-based limits
(Shackleford, 1988). Using its bioassessment approach as a screening tool,
Arkansas follows a formal decision tree for assessing compliance with es-
tablished water quality limits (Fig. 5-2). The initial bioassessment screen
may result in the application of other biological, toxicological, or chemical
methods. After completion of screening, an on-site decision can be made
for subsequent action. In situations where "no impairment" or "minimal
impairment" classifications are obtained, field efforts are reduced in fre-
quency or intensity until further information indicates a problem. Streams
classified as "substantially" or "excessively" impaired trigger additional
The EPA¦ encourages
states to fully-
integrate biological
surveys,
whole-effluent and
ambient toxicity
testing, and
chemical-specific
analyses to assess
attainment or
nonattainment of
designated aquatic
life uses in state water
quality standards.
79
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
SCREENING LEVEL
No or Minimal Impolrmert
Substantia! or Bcessive impairment
Oiermnalion of
Potential Geneffc Cause
No Further investigation
INTEGRATION LEVEL
Generic Cause - Organic
or Rysical AJteraftan
Generic Cause - Toxic
Further Investigative Action ~
may include cherrfcal analysis of water, sediments
or fish flesh or Microtek aqueous boa&says or
sediment Woassays
I
ComMeration for ToXetty Reduction
EraMion
Development of Bsrrrit Writ® and Corrpllanse
MonHof ins ftoyam; Application el Numeric andtar
ftetral'rve Slte-Specffic Criteria
I
COMPLIANCE MONITORING LEVEL
Deieirrfnalion of Coinplanoe Status Vis
Bsrmttee Supported Monitoring
I
COMPUANCE INSPECTION LEVEL
Wrification oI Compliance Status; Trend
Monitoring
Figure 5-2.—Decision matrix for application of rapid bloassessments In Arkansas for permitted point source dis-
charges (Shackleford, 1988).
'€
-------
GtiAPTEH 5:
Evaluating Environmental Effects
investigative steps that employ an integration of methods (Shackleford,
1988).
The definitive evaluation of water quality impacts often requires ex-
pensive laboratory analyses. However, careful review of conditions in the
watershed can provide early warning signals about the potential for water
resource degradation. For example, the presence of industrial, domestic,
and agricultural sources of chemical contaminants may be indicated by
odors, froth, or colors in the water. These conditions should be noted dur-
ing field surveys for their potential diagnostic value.
Habitat Structure
The physical structure of stream environments is critical to the ecological
health or integrity of lotic water resources. Attributes of significance to or-
ganisms in streams are channel morphology including width, depth, and
sinuosity; floodplain" shape and size; channel gradient; in-stream cover
such as presence of boulders and woody debris; substrate type and the di-
versity of substrates within a stream reach; riparian vegetation and the
canopy cover that it provides; and bank stability.
Channel morphology in natural watersheds is typically meandering with
substrate diversity created by varying velocities along and across the chan-
nel. As a result, substrates are sorted to form pools and riffles that create hori-
zontal variation in the physical environment, If a channel has been artificially
straightened and dredged (channelized), temporal recovery will recreate sub-
strate diversity through vertical and lateral meandering processes (Hupp,
1992; Hupp and Simon, 1986). Because no stream channel is stable, a tempo-
ral dimension of diversity also exists. These physical attributes are closely
tied to other environmental conditions and impairments (Table 5-1).
The influence of habitat structure spans the range from regional geog-
raphy to the pattern of interstitial spaces between rocks in the river sub-
strate. Habitat structure on all scales is critical to the biology of most
stream organisms, and subtle or massive habitat alteration on any scale
may influence the quality of the water resource.
The influence of habitat structure on the aquatic community causes
natural variability even in undisturbed communities. Understanding the
relationship of expected trends in biological condition as a result of
changes in habitat structure is an important feature of biological assess-
ments. Ohio EPA found that their measurement of habitat quality, the
Qualitative Habitat Evaluation Index (QHEI), was significantly correlated
with the Index of Biotic Integrity (IBI) in Ohio streams (Fig. 5-3) with r =
0.47 (Rankin, 1991) on a broad scale over the state. Rankin also found that
stream habitat quality and land use at various geographic scales are im-
portant influences on fish assemblages and that relatively intact stream
habitat throughout the drainage can compensate for short stretches of
poor habitat. In contrast, however, habitat-sensitive species may be re-
duced or destroyed in stream basins with extensive degraded conditions,
even if short stretches of good habitat exist. The Maryland Department of
the Environment, using the relationship between habitat structure and
biological condition, demonstrated effects from various influences (Fig. 5-
4) including agricultural runoff, treatment plant effluent, channelization,
and landfill operations (Primrose et al. 1991).
Careful review of
conditions in the
watershed can
provide early warning
signals about the
potential for water
resource degradation.
81
-------
BIOLOGICAL CRITERIA.
Technical Guidance for Streams and Small Rivers
An assessment of
habitat structure is
critical to any
evaluation of
ecological integrity.
Habitat assessment
provides information
on habitat quality; it
also identifies obvious
constraints on the
site's potential to
achieve attainment,
assists in the
selection of
appropriate sampling
stations, and provides
basic information for
interpreting biosurvey
results.
Table 5-1.— Parameters that may be useful in evaluating environmental
conditions and their relationship to geographic scales and the environmental
factors influenced by human actions.
CATEGORY BY
GEOGRAPHIC SCALE
PARAMETER
ENVIRONMENTAL
FACTORS4
1. Watershed
Land use'
Flow stability'
Flow regime
Physical habitat
2. Riparian and
bank structure
Upper bank stabilitya,,,h
Bank vegetative stabilitya,f,h
Woody riparian vegetation*1
— species identity
— number of species
Grazing'or other disruptive pressures8''
Streamside cover (% vegetation)®''
Riparian vegetative zone width8''
Stream bank erosion'
Flow regime
Energy base
Physical habitat
3. Channel
morphology
Channel alteration®1"''
Bottom scouring3
Deposition8
Pool/riffle, run/bend ratio"
Lower bank channel capacity®
Channel sinuositya,l'h
Channel gradient''11
Bank form/bend morphology"
Flow regime
Energy base
Biotic interactions
Water quality
Physical habitat
4. In-stream
Substrate composition/size; % rubble,
gravel, submerged logs, undercut
banks, or other stable habitatEI,Cid'0i,
% pools'
Pool substrate characterization3
Pool variability8
% embeddedness of gravel, cobble,
and boulder particles by fine sediment;
sedimentation®"0*'
Rate of sedimentation
Flow rate8d
Velocity/depth8,40
Canopy cover (shading)8''
Stream surface shading (vegetation,
cliffs, mountains, undercut banks,
!ogs)b'a,f
Stream width®'"
Water temperature0
Flow regime
Energy base
Biotic interactions
Water quality
Physical habitat
REFERENCES:
"Piafkin et al. 1989 . 'Osborne et al, 1991
"platts et al. 1987 "Barton et a!. 1985
cPlatts et al. 1983; Armour et al. 1983 ^Hupp and Simon, 1986; 1991
"Rankin, 1991 Karr and Dionne, 1991
'Gorman, 1988 'Karr, 1991
Habitat Quality and Biological Condition
The variability of environmental conditions directly affects patterns of life,
population, and the micro- and macrogeographic distribution of organ-
isms (Cooper, 1984; Price, 1975; Smith, 1974). An assessment of habitat
structure is therefore critical to any evaluation of ecological integrity (Karr
et al. 1986; Piafkin et al. 1989). Habitat assessment provides information on
habitat quality; it also identifies obvious constraints on the site's potential
to achieve attainment, assists in the selection of appropriate sampling sta-
8,
-------
CHAPTERS:
Evaluating En vironmehta) Effects
IBI
60
55
50
45
40
35
30
25
20
15
10<-
20
N = 465
r2=0.47
• • •
• •
• • •
• •••'
• • • • •
• •
• •
• • •
\
Point Size is Related
to Number of Data
Points Overlapping
30
40
50
60 •
QHEI
70
80
90
100
Figure 5-3.—Qualitative Habitat Evaluation Index (QHEI) versus the Index of Blotlc In-
tegrity (IBI) for 465 relatively unimpacted and habitat modified Ohio stream sites
(Rankin, 1991).
Unimpaired
Heavy Woodland Buffer
Moderately
Impaired
Landfi
Severely
impaired'
Agricultural
STP Efii.
Ditched
Moderately Supporting
Nonsupporting
Supporting
Figure 5-4.-
10 20 30 ' 40 50 60 70 80
Habitat - % of Reference
-Choptank and Chester rivers tributaries (Primrose et al. 1991)..
100
tions, and provides basic information for interpreting biosurvey results
(Atkinson, 1985; Osborne et ai. 1991). A carefully conducted habitat evalu-
ation is essential for distinguishing cause and effect elements from among
the five environmental factors influenced by human activity.
Development of a Habitat Assessment Approach
The development of a stream habitat assessment approach follows a logi-
cal sequence beginning with the, characterization of the waterbody. Only
similar aquatic systems may be compared; habitat structural parameters
applicable to one part of the country may not be applicable in another. For
instance, the extent of canopy cover differs between forested mountain
streams and open prairie streams found in the southwest. Thus, the ab-
sence of canopy cover is a more important habitat influence in a forested
Only similar aquatic
systems may be
compared; habitat
structural ¦parameters
applicable to one part
of the country may
not be applicable in
another.
The development of
a stream habitat
assessment follows a
logical sequence.
Waterbody Characteristics
¦ I '
Selection of the taxa
(Benthic Macro-
invertebrates, Fish)
I
Influential Habitat
Variables
(Flow, Shade, Substrate,
Buffer Zone)
I
Judgment Criteria
(Optimal, Suboptimal,
Marginal, Poor)
13
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Gradient is perhaps
the most influential
factor for segregating
a lotic waterbody
because it is related
to topography and
landform, geological
formations, and
elevation, which in
turn influence
vegetation patterns.
stream than in open streams (Barbour and Stribling, 1991). Another con-
sideration would be broad physiographic characteristics, for example, ele-
vation, general topography and gradient, and predominant soil types.
Finally, the biogeographic distribution of species and assemblages of or-
ganisms varies regionally.
Selection of the taxa, that is, the biological community to be studied, is
the important next step. Ideally, this selection is based on the best approach
to a comprehensive water resource assessment. However, the availability of
resources and the training of available staff will have significant influence.
The selection of one or more assemblages is important for determining
which habitat variables are most influential for community development.
For each parameter, the range of conditions to be expected is determined
and divided into scoring categories. These scoring categories (optimal,
suboptimal, marginal, and poor) form the basis of criteria that allow habi-
tats to be judged during on-site evaluation. An important call must then
be made. If habitat structure is degraded relative to the expectations pro-
vided by the appropriate reference condition, some inference must be
drawn about the nature and cause of the difference. If the study site is de-
graded relative to the reference, then habitat structure has been identified
as a potential cause of reduced biotic condition. If habitat structural differ-
ences result from the natural landscape rather than human interference,
then the possibility that an inappropriate reference condition was used
must be considered.
The habitat assessment approach outlined here (following Barbour and
Stribling, 1991; Plafkin et al. 1989) is applicable to wadable streams and riv-
ers. Because fish and benthic macroinvertebrates are the focal points of these
recommended bioassessment procedures, habitat structural parameters were
chosen that influence the development of these communities. Although
streams across the country exhibit a wide range of variability, some generali-
zations can be made. Gradient is perhaps the most influential factor for dis-
tinguishing lotic waterbodies because it is related to topography and
landform, geological formations, and elevation, which in turn influence
vegetation patterns. Four generic stream categories related to gradient can be
identified: mountain, piedmont, valley plains, and coastal plains. Several
habitat attributes serve as a framework for assessing Habitat quality:
• Substrate variety/in-stream cover
• Bottom substrate characterization/embeddedness
• Flow or velocity/depth
• Canopy cover (shading)
• Channel alteration
• Bottom scouring and deposition
• Pool to riffle and run to bend ratios, channel sinuosity
• Lower bank channel capacity
• Upper bank stability
• Bank vegetative stability (grazing or other disruptive pressure)
• Streamside cover
• Riparian vegetative zone width
•4
-------
CHAPTERS:
Evaluating Environmental Effects
While the investigator is on-site, the quality of each parameter can be
assessed; First, numeric value from a scale based on a gradient of condi-
tions is assigned to assess the quality of each parameter. Then, a composite
of information from each parameter is compared to a reference condition.
Such a quantified assessment of habitat structure provides a more mean-
ingful interpretation of biological condition. Habitat assessment incorpo-
rates information on stream segments or reaches. However, a linear
relationship between site-specific quality of habitat and community per-
formance may not exist to the point that habitat structural condition can
be used to "predict" biological performance with accuracy.
If habitat degradation has occurred, mitigation or improvement of the
habitat through stream restoration activities should be evaluated. Imple-
mentation of water quality improvements can be independent of habitat
quality, but judgment of the improvement in biological integrity cannot.
Flow Regime
Fluctuating water levels are an integral part of the stream ecosystem, and
the biota are dependent on seasonal flow variation. High flow events are
especially important in maintaining the habitat complexity of pools, rif-
fles, clean substrates, and bars (Hill et al. 1991). Aquatic organisms have
evolved to compensate for changing flow regimes, even periodic cata-
strophic flow conditions. High water periods are determined by the fre-
quency, occurrence, and type of precipitation event as well as antecedent
conditions such as soil moisture, time since last rain, and amount and type
of soil cover. Dewatering the channel for major periods as a result of hu-
man actions is clearly a degradation of the water resource, but more subtle
changes in the volume and periods of flow may have equally devastating
effects on the resident biota.
Jones and Clark (1987) discuss the effects of urbanization on the fun-
damental hydrology of watersheds and the natural flow regime. Increases
in impervious surface area (e.g., roads, parking lots) result in a substantial
increase in the proportion- of rainfall that is rapidly discharged from the
watershed as direct runoff and streamflow. Such runoff increases the vol-
ume of flood flows and instances of channel .instability. Leonard and Orth
(1986) developed a cultural pollution index to evaluate the health of the
fish community subject to the effects of road density, population encroach-
ment, mining, and organic pollution. These effects have substantial influ-
ence on flow regime. Steedman (1988) also evaluated the condition of fish
communities in heavily urbanized areas of Ontario. He found that certain
attributes that are relatively sensitive to urbanization effects can serve as
pertinent response signatures.
Ohio EPA found that the presence or absence of channelization influ-
enced the relationship between the quality of habitat structure and the
condition of the fish community (Ohio Environ. Prot. Agency, 1990). In the
absence of channelization, for example, Twin Creek and Kokosing River
(Fig. 5-5) had high IBI values, even in the presence of sporadic degraded
habitat. In these instances, the relatively good habitat quality throughout
the watershed supported the fish community in short reaches of de-
graded, habitat (Rankin, 1991). In channelized lotic systems, for example,
Tiffin River and Little Auglaize River (Fig. 5-5), the best habitats were de-
Implementation of
water quality
improvements can be
independent of
habitat quality, but
judgment of the
improvement in
biological integrity
cannot.
Fluctuating water
levels are an integral
part of the stream
ecosystem, and the
biota are dependent
on seasonal flow
variation.
85
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
60
o*
•• _
TWINCR.
TIFFIN R,
L.AUGLAIZE R.
KOKOSINGR.
mmm 40
^ 30
20
10 20 30 40 50 60 70 80 90 100
0
QHEl
Figure 5-5.—Relationship of the Index of Blotic Integrity (IBI) to changes In the quality
of habitat structure through the Qualitative Habitat Evaluation Index (QHEl) in chan-
nelized (triangles) and unchannellzed (circles) (Ohio Environ. Prot. Agency, 1990).
graded and IBI scores remained essentially unchanged as the habitat was
degraded further. The quality of habitat structure and the flow regime are
intricately associated. In areas of extensive channelization, communities
may consist only of generalists and opportunists able to withstand harsh
flow conditions directly, or the secondary effects of those flow conditions
(e.g., reduced abundance of food or presence of habitat refuges).
¦ Effects of Channelization. Unchannelized or otherwise unmodified
streams have normal, low-level, and mostly consistent rates of sediment
deposition on the bed and low, convex banks. The channel usually has
some degree of meandering, and the banks lose very little mass during
either low or high flows.
Efforts to control flooding and to drain wetlands often involve chan-
nelization of streams to provide more rapid removal of water. Unfortu-
nately, these activities create unstable channels with higher gradients and
without meanders. Hydrogeomorphic processes tend to restore the dy-
namic stability of these systems over time (Hupp and Simon, 1991). The
stream continuum hypothesis (Vannote et al. 1980) depicts the stream as
an upstream-downstream gradient of gradually changing physical condi-
tions and associated adjustments in functional attributes of the biota.
Biological processes in downstream areas are linked to those in up-
stream areas by the flow of water, nutrients, and organic materials. Be-
cause channelization produces an increase in flow velocity or scour, active
bed degradation occurs, causing the movement of substrate particles
downstream. As bed degradation continues, degradation of lower stream-
banks begins, eventually producing bank failure and concave upward
banks. During this period of severe instability, the channel is rapidly (in a
geologic sense) becoming wider and the water level shallower, sometimes
producing a braided flow pattern. Channel widening causes persistent
86
-------
CHAPTER 5:
Evaluating Environmental Effects
bank failure in the downstream areas and results in losses of canopy cover
and detrital input. These degradation processes move upstream, reducing
the rate of channel widening and providing depositional sediment in
downstream areas.
Hydrological processes in channelized streams have direct effects on
the substrate (embeddedness, scour, and particle size distribution). Trans-
ported sediment causes aggradation to occur downstream with deposition
on the bed and at the bases of banks. Accretion occurs on the banks with
the beginning of the stabilization processes, and seed supplies from ripar-
ian Vegetation or windblown from other areas settle on these deposits. As
vegetation, particularly woody species, becomes established on bank de-
positional surfaces, stability increases. During this phase of the channel re-
covery process, meandering features develop through deposition and
vegetative, stabilization of point bars (inside bend). The return of dis-
turbed stream channels to a dynamically stable, meandering morphology
results primarily from the aggradation of banks and beds and the estab-
lishment of riparian stands of woody vegetation (Hupp, 1992; Hupp and
Simon, 1986, 1991; Simon and Hupp, 1987). Hupp (1992) has estimated
that an average of 65 years is needed for this recovery process in non-
bedrock controlled, channelized streams in west Tennessee.
A complete concrete lining of natural waterways in western states has
long been used to control wet weather flooding. Low flows of reclaimed
water are the only source of water for most of the year in these "streams."
Wet weather flows are commonly enormous and rapid. Though techni-
cally listed as streams and rivers, these engineered channels do not clearly
fit definitions commonly understood for either "aquatic habitat" or
"streams."
'¦ Effects of Flow Regulation. Many streams are characterized by highly
variable and unpredictable flow regimes (Bain et al. 1988). Aquatic macro-
phyte stands have been shown to be affected by current velocity, but the
degree and manner varies with the size of the channel (Chambers et al.
1991). In regulated streams, the importance of a bank-to-midstream habi-
tat orientation becomes magnified. Flow changes displace the shallow
shoreline zones, forcing fish restricted to these areas (small fish that use
shallow, slow microhabitats) to relocate to maintain their specific set of
habitat conditions (Bain et al. 1988). Therefore, if shallow-water habitats
are unstable and unable to sustain a well-balanced assemblage, then the
functional value of the assemblage is lost and a reduction in organismal
population density may follow.
Gislason (1985) illustrates a similar pattern for aquatic insect distribu-
tion in fluctuating flows. Bain et al. (1988) also suggest that without the
functional availability of shallow, slow, shoreline areas, the stream envi-
ronment becomes one general type of unstable habitat, dominated by a
few habitat generalists and those species using mostly mid-stream habi-
tats. In these cases, the dominance of generalists confounds the assess-
ment of contiguous impact types such as nonpoint source runoff and point
source discharges. Comparison of historical and current flow conditions
can provide valuable information about the extent to which flow altera-
tion is responsible for degradation in biological integrity.
Comparison of
historical and current
flow conditions can
provide valuable
information about the
extent to which flow
alteration is
responsible for
degradation in
biological integrity.
87
-------
BIOLOGICAL CRITERIA
Technical Guidance for Streams and Small Rivers
Alterations to the
energy base are not
independent of
alterations to habitat
structure. In many
instances,
assessment of habitat
quality is an
assessment of
impacts to the energy
base.
Energy Source
Stream organisms have evolved to accept and use the energy available to
them in natural watersheds. For most small or headwater streams in for-
ested areas of North America, a period of major leaf fall occurs in the
autumn. Leaves, in a form referred to as coarse particulate organic matter
(CPOM), reach the water and are quickly colonized by bacteria and fungi.
The organisms then provide food for invertebrates, which are in turn
eaten by fish and other vertebrates. The relative balance of production and
respiration varies as a function of stream size, according to the stream con-
tinuum hypothesis (Vannote et al. 1980).
Human alteration of the source, type, and quantity of organic material
entering streams can affect biological integrity in many ways. Natural
shifts in the energy base occur along stream and river gradients, thus pro-
viding a major dimension of resource partitioning for the aquatic commu-
nity. The stream continuum concept (Vannote et al. 1980) outlines different
attributes of communities as the energy base shifts from heterotrophic (ex-
ternal) to autotrophic (internal) inputs. These shifts are generally related
to increases in drainage area catchments, but exceptions do occur that are
related to localized conditions.
Along the stream/river gradient (Fig. 5-6), Cummins (1983) describes
the measurement of this shift as a photosynthesis/respiration (P/R) ratio.
This P/R ratio is less than 1 in the headwater areas of streams and large
rivers. Therefore, these reaches are heterotrophic because in-stream photo-
synthesis is not a primary energy source. The P/R ratio is greater than 1 in
the mid-sized rivers where in-stream photosynthesis is a major contribu-
tor to the energy base; the latter are autotrophic. The removal of riparian
vegetation for agriculture, channelization, or strip mining, or the shift
from natural riparian flora to introduced species for urbanization projects
alters the energy base of the aquatic system. Although the stream contin-
uum is thought to no longer hold true for the majority of watersheds, it
does exemplify the important considerations in energy base and aquatic
ecosystem interaction.
Alterations to the energy base are not independent of alterations to
habitat structure. In many instances, assessment of habitat quality is an as-
sessment of impacts to the energy base. However, the evaluation of
changes in the energy base can be strengthened by a systematic riparian
assessment based on a delineation of natural flora. Alterations in the spe-
cies of riparian plants influence the functional representation of the
aquatic trophic structure biota.
Wilhelm and Ladd (1988) developed a basic tool for conducting natu-
ral area assessments in the Chicago region. They presented a checklist of
vascular plants of the Chicago region and assigned each species a coeffi-
cient of conservatism. This measure expresses the value of the species rela-
tive to all other elements in the flora and its particular tie with ancestral
vegetation. Low scores are given to native species that are relatively ubiq-
uitous under a broad set of disturbance conditions; high scores are given
to species that are sensitive to disturbance; and no scores are assigned to
non-native species. In this manner, vegetation can be assessed as repre-
senting natural or disturbance conditions.
-------
CHAPTERS
Evaluating Environmental Effects
(0,5 METERS)
MICROBES
01
a:
<1-2 METERS)
lEDATORS
COLLECTORS
P/r<1
iLl
oc
IfRODUCERS
(VASCULAR
M HYDROPHYTES)
(4-6 METERS)
CD
MICROBES)
C P OM
UJ 4-
Q
(10 METERS)
PRODUCERS
GRAZERS
UJ
(50-75 METERS)
•REDATORS
en
o
COLLECTORS
MICROBES
ce
LlI
CO
10-
12-1(700 METERS)
Figure S-6.—Diagrammatic representation of the stream continuum to illustrate vari-
ation In trophic structure of benthic Invertebrates (adapted from Cummins, 1983).
Applying this method to riparian corridors would require a similar
classification of vegetation. However, much literature is available to aid in
classifying riparian flora. The U.S. Forest Service has compiled an exten-
sive database on riparian systems that has been published in several re-
ports (e.g., Platts et al. 1983). Hupp and Simon (1991) recognize early
successional species of woody vegetation in riparian zones of disturbed
and recovering stream channels in western Tennessee. Padgett et al. (1989)
provide a substantial list of references documenting vegetation classifica-
tion in many of the western states.
89
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Biotic Interactions
Predation, competition, disease, and mutualistic interactions influence
where and when species occur within streams. Larval stages of mussels,
for example, must attach to the gills of specific fish species to complete
their life cycles. Stream communities are often dominated by a few
"strongly interacting" species that may have disproportionate effects on
the other members of the community (Hart, 1992; Power, 1990). The addi-
tion of human influences may alter the integrity of these interactions in
ways that alter the abundances of local species and may even cause their
demise. Additional human influences are harvests for sport and commer-
cial purposes and the introduction of exotic species, sometimes intention-
ally but often inadvertently. The practice of stocking fish can be an
ecological or genetic disturbance, especially if naturally occurring popula-
tions are replaced or infiltrated by stocked individuals. However, the ac-
ceptance of this practice is an important societal decision; its advantages
and disadvantages must be carefully weighed.
Cumulative Impacts
Even when human actions have an influence on only one of these factors,
the effect may cascade through several others. For example, clearing land
for agriculture alters the erosion rate and thus the extent to which sedi-
mentation may alter the regional biota. Removal of natural vegetation re-
duces shading, water infiltration, and groundwater recharge, thereby
increasing water temperatures, insolation, and the frequency of flood and
drought flows. The resultant agricultural activities may change the stream
through channelization, and thus further influence habitat structure. Al-
terations in the land cover and the channel often have major impacts on
water quality (e.g., increased amounts of nitrogen and phosphorus in the
runoff from agricultural fields or pesticides in the water). Excess nutrients
in modified channels exposed to ample sunlight will enhance the growth
of nuisance algae, especially during summer's low flow periods.
Unfortunately, human influences on stream ecosystems cannot be eas-
ily categorized (Karr, 1991). The close association between alteration of
habitat structure and other impact types complicates the determination of
"cause and effect." However, this dimension becomes paramount when
mitigative measures are-crucial to the attainment of designated uses or
biocriteria. In many cases, deductive reasoning, thorough review of the
biological data, and use of biological response signatures supported by
other environmental data (i.e., physical characterization, toxicity testing,
and chemical analyses) aid the assessment of impairment.
The implications of significantly altered systems, for example, chan-
nelized streams in urban areas or stream flows regulated by hydroelectric
dams, are that reference conditions different from the natural system may
have to be established to represent these systems and to evaluate other im-
pact types (Karr and Dionne, 1991). When major impacts (i.e., significant
habitat alterations) are present, it is 'difficult to adequately evaluate
changes in community elements and processes that may be attributable to
other impacts.
90
-------
CHAPTERS:
Evaluating Environmental Effects
60
50
IMPACT TYPE
'GRADIENT"
1
BIOLOGICAL RESPONSE
STREAM/
IMPACTS
1 40 b
30
20
10
least impacted,
"Reference"
Concisions
Minor sewage and
most agricultural
NPS impacts
Moderate anrich-
ment.siltation.iow
QO.habitat impacts
CSO/urban impacts,
. chronic toxicity
Complex toxic
(acuta), acid mine,
toxic sediments
FLOW
VERY POOR
G
BIG DARBY CR j
(Municipal. Agr. !
NPS) J
WALNUT CR.
(Industrial/
Conventional,
Municipal)
HOCKING H.
(Municipal w/Pre-
traatmant.CSO)
RUSH ca
(Acid Mine
Drainage)
RIVER MILE
Figure 5-7.—Biological community response as portrayed by the Index of Blotlc In-
tegrity (IB!) In four similarly sized Ohio rivers with different types of point and non-
point source Impacts (Yoder, 1991).
The diversity of influences on the quality of water resources requires the
kind of multiple attribute approach common to recent biocriteria program ef-
forts. The use of a multiple attribute approach enables the development of
biological response signatures to assess probable "causes and effects."
Using biological response signatures, Ohio EPA (Yoder, 1991) was able
to assign each of their more severely degraded situations to one of six
groups:
• complex municipal and industrial wastes,
• conventional municipal and industrial wastes,
• combined sewer overflow and urbanization,
• channelization,
• agricultural nonpoint source, or
• other, often complex, impacts.
The Ohio EPA also found that various impact types may have one or
two biological response characteristics in common. In rare cases, they have
three in common. Therefore, only a multiple assemblage, multimetric ap-
proach enables a differentiation, among impact types. In certain cases, the
severity of the impact is related to the type of impact. The IBI has been
used by Ohio EPA to characterize these impact types (Fig. 5-7).
Suggested Readings
Atkinson, S.F. 1985. Habitat-based methods for biological impact assessment. Environ.
Prof. 7:265-82.
The diversity of
influences on the
quality of water
resources requires
the kind of multiple
attribute approach
common to recent
biocriteria program
efforts. The use of a
multiple attribute
approach enables the
development of
biological response
signatures to assess
probable "causes and
effects."
91
-------
I
I
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Bain, M.B., j.T, Finn, and H.E. Booke, 1988, Streamflow regulation and fish community
structure. Ecology 69(2):382-92.
Ball, J. 1982. Stream classification guidelines for Wisconsin. In 1983 Water Quality Standards
Handbook. Off. Water Reg. Standards, U.S. Environ. Prot. Agency, Washington, DC.
Barbour, M.T. and J.B. Stribling. 1991. Use of habitat assessment in evaluating the biological
integrity of stream communities. Pages 25-38 in Biological Criteria: Research and Regu-
lation. EPA 440/5-91-005. Off. Water, U.S. Environ. Prot. Agency, Washington, DC.
Karr, J.R. et al. 1986. Assessing Biological Integrity in Running Waters: A Method and Its
Rationale. Spec. Publ. 5. Illinois Nat. History Surv., Urbana, IL.
Karr, J.R. 1991. Biological integrity: a long-neglected aspect of water resource manage-
ment. Ecol. Appl. 1:66-84.
Leonard, P.M. and D.J. Orth. 1986. Application and testing of an index of biotic integrity
in small, coolwater streams. Trans. Am, Fish. Soc. 115:401-14.
Ohio Environmental Protection Agency. 1990. The Use of Biocriteria in the Ohio EPA
Surface Water Monitoring and Assessment Program. Columbus, OH
Platts, W.S., W.F. Megahan, and G.W. Minshall. 1983. Methods for Evaluating Stream,
Riparian, and Biotic Conditions. Gen. Tech. Rep. INT-138. Intermountain Res. Sta.,
Forest Serv., U.S. Dep. Agric., Ogden, UT,
Steedman, R.J. 1988. Modification and assessment of an index of biotic integrity to
quantify stream quality in southern Ontario. Can. J. Fish. Aquat. Sci. 45:492-501.
U.S. Environmental Protection Agency. 1983. Technical Support Manual: Waterbody
Surveys and Assessments for Conducting Use Attainability Analyses. Vol. 1-3. Off.
Water Reg. Stand., Washington, DC,
. 1990. Biological Criteria: National Program Guidance for Surface Waters. EPA-
440/5-90-004. Off. Water, Washington, DC.
92
-------
CHAPTER 6.
Multimetric Approaches
for Biocriteria
r
Development
Classical approaches to the assessment of biological integrity have
usually selected a single biological attribute that refers to a narrow
range of perturbations or conditions (Karr et al. 1986). Likewise, many
ecological studies have focused on a limited number of parameters, such
as species distributions, abundance trends, standing crops, or production
estimates, which are interpreted separately, then used to provide a sum-
mary statement about the system's overall health. These approaches are
limited because a single attribute may not reflect the overall ecological
health of the stream or region. An accurate assessment of biological integ-
rity requires a method that examines the pattern and processes of biotic re-
sponses from individual to ecosystem levels (Karr et al. 1986).
An alternative approach is to define an array of metrics, each of which
provides information on a biological assemblage and, when integrated,
functions as an overall indicator of the stream or river's biological condi-
tion. The strength of a multimetric assessment is its ability to integrate in-
formation from individual, population, community, and ecosystem levels
and evaluate this information, with reference to biogeography, as a single,
ecologically based index of water resource quality (Karr, 1991; Karr et al.
1986; Plafkin et al. 1989). Multimetric assessments provide detection capa-
bility over a broad range and nature of stressors. The Ohio EPA (1987) sug-
gests that the strengths of individual metrics taken in combination
minimize any weaknesses they may have individually.
Abel (1989), LaPoint and Fairchild (1989), and Karr (1991) do not rec-
ommend using a single metric. For the broad range of human impacts, a
comprehensive, multiple metric approach is more appropriate. Similarly,
each of the assemblages discussed in Chapter 4 has a response range to
disturbing events and impairments (degraded conditions). Therefore, bio-
surveys that target multiple assemblages provide the detection capability
that is needed to accomplish assessment objectives.
Karr (1991), Karr et al. (1986), Ohio EPA (1987), and Plafkin et al. (1989)
recommend use of a number of biological assemblages and metrics that
can, when combined and compared with expected conditions, give a more
complete picture of the relative biological condition of the study site.
Purpose:
To describe a
multimetric approach
for analyzing
biological data and to
provide guidance for
regional selection of
metrics.
An accurate
assessment of
biological integrity
requires a method
that examines the
patterns and
processes of biotic
responses from .
individual to
ecosystem levels.
93
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Core metrics should
represent diverse
aspects of structure,
composition,
individual health, or
processes of the
aquatic biota.
Metric Evaluation and Calibration
Core metrics should represent diverse aspects of structure, composition,
individual health, or processes of the aquatic biota. Together they form the
foundation for a sound integrated analysis of the biotic condition and esti-
mate of the system's biological integrity. Thus, metrics reflecting commu-
nity characteristics are appropriate in biocriteria programs if their
relevance can be demonstrated, their response range verified and docu-
mented, and the potential for program application exists. Regional vari-
ation in metric details are expected; nevertheless, the general principles
used to define metrics seem consistent over wide geographic areas (Miller
et al. 1988).
¦ Candidate metrics are determined from the biological data. Good-met-
rics have low variability with respect to the expected range and resppnse of
the metrics: it must be possible to discriminate between impaired and unim-
paired sites from the metric values. The use of percentiles is a useful tech-
nique to evaluate variability of metric performance within stream classes. In
operational bioassessment, metric values below the lower quartile of refer-
ence conditions are typically judged impaired to some degree (e.g., Ohio
Environ. Prot. Agency, 1990). The distance from the lower quartile can be
termed a "scope for detection" (Fig. 6-la), The larger this distance, com-
pared to the interquartile range, the easier it is to detect deviation from the
reference condition. Thus, we can define a "detection coefficient" as the ra-
tio of the interquartile range to the scope for detection (Gerritsen and Bow-
man, 1994). This coefficient is analogous to the coefficient of variation (CV),
and the smaller the value, the easier it is to detect the impairment.
Metrics with high variability, or scope for detection, compared to the
range of response should be used with caution. Many metrics (e.g.,
number of taxa) decrease in value with impairment and the detection coef-
ficient for reference sites is thus a good measure of the metrics' potential
discrimination ability. Some metric values (e.g., HBI, percent omnivores,
Max
Min
maximum
75th percentile
median
25th percentile
minimum
interquartile
range
scope for
detecting
impairment
Figure 6-1 a.—Metrics that decrease with Impairment.
94
-------
Multimetric Approaches
CHAPTERS:
for Biocriteria Development
100%-
0%
I
scope for
detecting
impairment
1
interquartile
range
Figure 6-1 b.— Metrics that increase with Impairment.
percent filterers) may increase under impaired conditions, and the scope
for detection would be from the 75th percentile to the maximum value
(Fig, 6-lb). The detection coefficient would be calculated the same way
and used to judge the discriminatory power of the metrics..
Certain metrics may exhibit a continuum of expectations dependent
on specific physical attributes of the reference streams. For example,
Fausch et al. (1984) determined that the total number of fish species
changes as a function of stream size estimated by stream order or water-
shed area (Fig. 6-2). They showed that when these data are plotted, the
points produce a distinct right triangle, the hypotenuse of which approxi-
mates the upper limit of species richness. Thus, a line with a slope fitted to
include about 95 percent of the sites is an appropriate approximation of a
maximum line of expectations for the metric in question and identifies the
upper limit of the reference condition. The area on the graph beneath the
maximum line can then be trisected or quadrisected to assign scores to a
range of metric values as illustrated in Figure 6-2. The scores provide the
transformation of values to a consistent measurement scale to group infor-
mation from several metrics for analysis.
When different stream classes have different expectations in metric
values and a covariate that produces a monotonic response in a metric, a
plot of survey data for each stream class may be useful (Fig. 6-3). For each
metric, the sites are sorted by stream class (e.g., ecoregion, stream type)
and plotted to ascertain the spread in data and the ability to discriminate
among classes (Fig. 6-4). If such a representation of the data does not allow
discrimination of the classes, then it will not be necessary to develop a
separate biocriterion for each class. That is, a single criterion will be appli-
cable to a set of sites that represent different physical classes. Conversely,
if differences in the biological attribute are apparent and appear to corre-
spond to the classification, then separate criteria are necessary. This tech-
nique is especially useful if the covariates are unknown or do not exist,
but a difference in stream class is apparent (Fig. 6-4).
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
in 30 —
G)
O
2L
CO
sz
Si
E
ra
o
t-
20—
10-
Maximum Species
Richness Line1
Stream Order
Figure 6-2.—Total number of fish species versus stream order for 72 sites along the
Embarras River in Illinois (Fausch et al. 1984).
O
MMrio
{•.g.ap^cfca
(e.g., Stream Size)
Figure 6-3.—Metrics plotted with a continuous covariate (hypothetical example).
Pilot studies or small-scale research may be needed to define, evaluate,
and calibrate metrics. Past efforts to evaluate the use of individual metrics
illustrate procedural approaches to this task (Angermeier and Karr, 1986;
Barbour et al. 1992; Boyle et al. 1990; Davis and Lubin, 1991; Karr and Ker-
ans, 1992; Karr et al. 1986; Kerans et al. 1992; Lyons, 1992; Resh and Jack-
son, 1993). Metrics can be calibrated by evaluating the response of metric
values to varying levels of stressors.
Sites must be carefully selected to cover the widest possible range of
suspected stressors. In general, impaired sites are selected that have im-
pacts from stressors singly and in combination. The selected impaired sites
96
-------
CHAPTER 6:
Multimetric Approaches for Biocriteria Development
A B C D E
Stream Class
Figure 6-4.—Box and whisker plots of metric values from hypothetical stream
classes. Shaded portions are above the median for each class. The box represents a
percentile, the vertical line is 1.5 times the Interquartile range, and the horizontal line
is the median of each distribution.
and the reference sites together are the basis for developing an empirical
model of metric response to stressors. Categories of land uses equated
with potential impairment are listed in Chapter 7. Candidate metrics that
do not respond to any of the stressors expected to occur in a region may be
eliminated.
As an example, the discriminatory power of macroinvertebrate metrics
was evaluated for Florida streams. The judgment criteria for discrimina-
tion were based on the degree of interquartile overlap between the least
impaired site category and the impaired site category for each metric. A
metric was judged excellent if no overlap existed in the interquartile range
(Fig. 6-5a); poor if the overlap was considerable, and no distinction be-
tween the impairment categories could be made (Fig. 6-5b). An analysis of
a metric's performance among all of the site classes indicated the metric's
strength in discriminating between "good" and "bad" conditions.
Additional research is needed to demonstrate the responses of metrics
to different stressors in different ecoregions or stream systems. However,
once these factors have been considered and demonstrated, the metrics
can be incorporated into localized biocriteria programs. It is also impor-
tant that the metrics and necessary survey methods be appropriate to the
logistical and budgetary resources of the investigating agency. Practical
application is the penultimate step in metric development. Continued
evaluation of metrics and indices is an essential feature of the use of
biocriteria.
Biocriteria Based on a Multimetric Approach
The validity of an integrated assessment using multiple metrics is sup-
ported by the use of metrics firmly rooted in sound ecological principles
(Fausch et al. 1990; Karr et al. 1986; Lyons, 1992). For biocriteria, a biologi-
cal attribute or metric is some feature or characteristic of the biotic assem-
blage that changes in a predictable way with increased human influence.'
A biological attribute
or metric is some
feature or
characteristic of the
biotic assemblage
that .reflects ambient
condition, especially
the influence of %
human actions.
7
-------
BIOLOGICAL CRITERIA;
Technical Guidance for Streams and Small Rivers
X
CD
a.
ui
20
16
12
ri
Reference Impaired
Other
6575A
Reference Impaired
Other
75BCD
Min-Max
~ 25%-75%
° Median value
Figure 6-5a.— Site discrimination for the number of Ephemeroptera, Piecoptera, and
Trichoptera {EPT Index) In Florida streams. (Reference = least impaired, other =
unknown, Impaired = determined Impaired a priori.)
28
24
20
£
H
0)
16
(0
•o
E
g
12
s
JZ
8
O
o
*
4
0
Reference Impaired
Other
6S75A
Reference Impaired
Other
75BCD
Min-Max
~Z! 25%-75%
° Median value
Figure 6-5b.— Site discrimination for the number of Chlronomidae taxa In Florida
streams. (Reference = least Impaired, other = unknown, Impaired = determined
Impaired a priori.)
The status of the biota as indicated by a composite of appropriate attrib-
utes (metrics) provides an accurate reflection of the biological condition at
a study site. A large number of attributes have been used (e.g., see Fausch
et al. 1990; Karr, 1991; Karr et al. 1986; Kay, 1990; Noss, 1990), and each is
essentially a hypothesis about the relationship between in-stream condi-
tion and human influence (Fausch et al. 1990). Gray (1989) states that the
three best-documented responses to environmental stressors are reduction
in species richness, change in species composition to dominance by oppor-
tunistic species, and reduction in mean size of organisms. But because
9.
-------
CHAPTER 6:
Multimetric Approaches for'Biocriteria Development
Regional data iflterpretabk wrthin conceptual model
Provides new, important insights not available from
existing programs or measures
Evaluation of costs and benefits
Responsiveness demonstrated pilot field study
- High variability la response to natural
envtmameatal pressure
Cost prohibitive for implemeataiio q
Not responsive to stressors of concern
- Redundant with superior measures
Temporally unstable within the index
period
Important within the ecological system
understudy
Low incremental cost
Responsive to stressors oa a regional scale
Methods believed feasible oa a
regional scale
CANDIDATE
Rejected
Figure 6-6.—Tiered metric development process (adapted from Holland, 1990).
each feature responds to different stressors, the best approach to assess-
ment is the incorporation of many attributes into the assessment process.
The development of appropriate metrics is dependent on the taxa to
be sampled, the biological characteristics at reference conditions, and to a
certain extent, the anthropogenic influences being assessed. They must be
pertinent to the management objectives to which the biocriteria will be ap-
plied. In many situations, multiple stressors impact ecological resources,
and specific "cause and effect" assessment may be difficult. However,
change over sets of metrics in response to perturbation by certain stressors
(or sets thereof) may be used as response signatures.
A broad approach for program-directed development of metrics may
be modeled after Barbour et al. (1992), Fausch et al. (1990), Holland (1990),
or Karr and Kerans (1992). Candidate metrics are selected based on knowl-
edge of aquatic systems, flora and fauna, literature reviews, and historical
data (Fig. 6-6). During the research process, these metrics are evaluated for
efficacy and validity. Only after careful evaluation should the metrics be in-
troduced into the biocriteria program. Less robust metrics or those not
well-founded in ecological principles are weeded out in this research proc-
ess. Metrics with little or no relationship to stressors are rejected. The re-
maining, or core, metrics are those that provide useful information in
differentiating among sites having good and poor quality biotic charac-
teristics.
The use of multiple metrics to develop a framework for biocriteria is a
systematic process involving discrete steps. The process includes site classi-
fication (Chapter 3), conduct of a biosurvey and.determination of metrics,
aggregation into indices, and the formulation of biocriteria. The conceptual
model for processing biological data into a biocriteria framework is adapted
The development of
appropriate metrics is
dependent on the
taxa to be sampled,
the biological
characteristics at
reference conditions,
and to a certain
extent, the
anthropogenic
influences being
assessed.
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
1. Classification
2, Survey of Biota and Habitat
3. Candidate Metric Evaluation
4. Core Metric Calibration
5. Index Development
6, Blocriteria Development
Stream
Class
Designate
Stream
Class
Designation
Stream
Class
Designation
Survey Sites
(Biological Oala)
Metric 1
Metric 2
j Metrio3
Value
Value
J Value
Evaluation and Calibration
Metric 1
Metric 2
Metric 3
Indicators
L!p5»;
S>**< Aggregation
index
Score
Biocntena .
Relative to >"=23
Stream
Class
Assessment
of Sites
Figure 6-7.—The conceptual process for proceeding from measurements to Indica-
tors to assessment condition (modified from Paulsen et al. 1991).
from Paulsen et al. (1991) and illustrated in Figure 6-7. A description of the
process is summarized in Table 6-1 and described as follows:
¦ Step 1 — Classification. Sites are classified as described in Chapter 3 to
determine the stream class designation and to ascertain the best and most
representative sites for each stream class. The reference condition will be
established from this step. Site classification is necessary to reduce and
partition variability in the biological data. Multistate collaboration is en-
couraged in the development of these calibration regions; a benefit is that
common methods and metrics can be established among states and cross-
state comparisons are enhanced.
¦ Step 2 — Biosurvey. Surveys of the best sites and those known to be
impaired are made for both biota and physical habitat to determine the
discriminatory power of the metrics using the impaired and best sites
within the stream class. The use of standardized methods (Chapter 4) pro-
vides a better interpretation of the raw data than does a conglomeration of
techniques. The raw data from a collection of measurements must be
evaluated within the ecological context that defines what is expected for
similar waterbodies (by reference to waterbody type and size, season, geo-
graphic location, and other elements).
100
-------
Multimetric Approaches
CHAPTER 6:
for Biocriteria Development
Table 6-1.— Sequential progression of the biocriteria process.
BIOCRITERIA PROCESS
Step 1. Classification to Determine Reference Conditions and Regional Ecological
Expectations
• stream class designation
• best and representative sites (reference sites representative of class categories
and natural background physical integrity)
Step 2. Survey Best Sites {reference sites)
• biota and physical habitat
• database consists of raw data {taxonomic lists, abundance levels, arid other
direct measures and observations)
Step 3. Candidate Metric Evaluation
• data analysis (data summaries) of biological attributes
• calculation of candidate metrics
Step 4. Core Metric Calibration
• testing and validation of metrics by stream class
• calibration of metrics to discriminate impairment
Step 5. index Development
• determination of biological endpoints
• aggregation of metrics
Step 6, Biocriteria Development
• adjustment by physiochemical covariates
• adjustment by designated aquatic life use
¦ Step 3 — Candidate Metrics Evaluation and Calibration. Analysis of
the biological data emphasizes the evaluation of biological attributes that
represent the elements and processes of the community. All potential met-
rics having ecological relevance are identified in this step,
¦ Step 4 — Core Metric Calibration. From the data analysis, metrics are
evaluated for relevance to the biological community and validated by
stream classes. Calibration of the metrics must address the ability to differ-
entiate between impaired and nonimpaired sites.
¦ Step 5 — Index Development. For aggregation purposes, transforma-
tion to scores from values of various scales of measurement relevant to in-
dividual metrics must be done. These scores are normally incorporated
into an index, such as the IBI, which, in turn, becomes part of the final as-
sessment process. The individual metrics may also be used as indicators of
biological condition in the overall assessment of those endpoints — to
support the aggregated index or as individual endpoints.
¦ Step 6 — Biocriteria Development, Biocriteria are formulated from the
indices (Chapter 7) for the stream classes and adjusted by physical and
chemical covariates and designated aquatic life uses. The biocriteria may
be based on a single aggregated index or established for several biological
endpoints.
101
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Potential Metrics for Fish and
Macroinvertebrates
A number of metrics have been developed and subsequently tested in
field surveys of benthic macroinvertebrate and fish assemblages (Karr,
1991). Because metrics have been recommended for fish assemblages
(Karr, 1981; Karr et al. 1986) and for benthic macroinvertebrates (Barbour
et al. 1992; Kerans et al. 1992; Ohio Environ. Prot. Agency, 1987; Plafkin et
al. 1989), they will not be reviewed extensively here. A list of the fish as-
semblage metrics used in the Index of Biotic Integrity (IBI) is presented in
Table 6-2, which represents variations in regional fauna. Karr (1991) sepa-
rates these metrics into three classes: (1) species richness and composition,
(2) trophic composition, and (3) abundance and condition. These classes of
characteristics generally agree with the areas of assemblage response de-
scribed as being technically supported (Gray, 1989): reduction in species
richness, shift to numerical dominance by a small number of opportunistic
species, and reduction in the mean body size of individuals.
Benthic metrics have undergone similar evolutionary developments
and are documented in the Invertebrate Community Index (ICI) (Ohio En-
viron. Prot. Agency, 1987), Rapid Bioassessment Protocols (RBPs) (Barbour
et al. 1992; Hayslip, 1992; Plafkin et al. 1989; Shackleford, 1988) and the
benthic IBI (Kerans and Karr, in press). Metrics used in these indices are
surrogate measures of elements and processes of the macroinvertebrate as-
semblage. Although several of these indices are regionally developed,
some are more broadly based; and individual metrics may be appropriate
in various regions of the country (Table 6-3). |
Figure 2-2 (see chapter 2) illustrates a conceptual structure for attrib-
utes of a biotic assemblage in an integrated assessment that reflects overall
biological condition. A number of these attributes can be characterized by
metrics within five general classes: community structure, taxa richness,
variety, dominance, and relative abundance. Community structure"can be
measured by variety and distribution of individuals among taxa. Taxa
richness, or the number of distinct taxa, reflects the diversity within an as-
semblage. Multimetric uses of taxa richness as a key metric include the In-
vertebrate Community Index (Ohio Environ. Prot. Agency, 1987), the Fish
Index of Biotic Integrity (Karr et al. 1986), the Benthic Index of Biotic In-
tegrity (Kerans and Karr, in press), and Rapid Bioassessment Protocols
(Plafkin et al. 1989). Taxonomic richness is also recommended as critical
information in assays of natural phytoplankton assemblages (Schelske,
1984). Taxa richness is usually species level but can also be evaluated as
designated groupings of taxa, often as higher taxonomic groups (e.g., fam-
ily and order, among others) in assessments of invertebrate assemblages.
Relative abundance of taxa refers to the number of individuals of one
taxon as compared to that of the whole community. Abundance estimates
are surrogate measures of standing crop and density that can relate to
both contaminant and enrichment problems. Dominance (e.g., "measured
as percent composition of dominant taxon" [Barbour et al. 1992]) or domi-
nants-in-common (Shackleford, 1988) is an indicator of community bal-
ance or lack thereof. Dominance roughly equates to redundancy and is an
important indicator when the most significant taxa are eliminated from
the assemblage or if the food source is altered, thus allowing a few species
A number of
attributes can be
characterized by
metrics within five
general classes:
community structure,
taxa richness, variety,
dominance, and
relative abundance.
102
-------
CHAPTER 6:
Multimetric Approaches for Biocriteria Development
Table 6-2.—Index of Biotic Integrity metrics used in various regions of North America.
ALTERNATIVE
IBI METRICS
MIDWEST
NEW
ENGLAND
ONTARIO
CENTRAL
APPALACHIA
COLORADO
FRONT RANGE
WESTERN
OREGON
SACRAMENTO/
SAN JAQUIN
WISCONSIN
1. Total number of species
It native fish species
tf salmonid age classes3
X
X
X
X
X
X
X
X
X
X
X
2. Number of darter species
# sculpin species
# benthicinsectivore species
If darter and sculpin species
# salmonid yearlings (individuals)3
% round-bodied suckers
It sculpins (individuals)
X
X
X
X
X
X
X
X
X
X
X
3. Number of sunfish species
It cyprinid species
If water column species
If sunfish and trout species
It salmonid species
tf headwater species
X
X
X
X
X
X
X
X
4. Number of sucker species
If adult trout species3
It minnow species
ft sucker and catfish species
X
X
X
X
X
X
X
X
X
5. Number of intolerant species
It sensitive species
ft amphibian species
Presence of brook trout
X
X
X
X
X
X
X
X
6. Percent green sunfish
% common carp
% white sucker
% tolerant species
) % creek chub
% dace species
X
X
X
X
X
X
X
X
7. Percent omnivores
% yearling salmonids3
X
X
X
X
X X
X
X .
8. Percent insectivorous cyprinids
% insectivores
% specialized insectivores
It juvenile trout
% insectivorous species
X
X
X
X
X
X
X
X
9. Percent top carnivores
% catchable salmonids
% catchable trout
% pioneering species
Density catchable trout
X
X
X
X
*
X
X
X
X
10. Number of individuals
Density of individuals
X
X
X
X
X
X
X
X*
11. Percent hybrids
% introducted species
% simple lithophills
H simple lithophills species
% native species
% native wild individuals
X
X
X
X
X
X
X
X
X
12. Percent diseased individuals
X
X
X
X
X
X
X
"Metric suggested by Moyle or Hughes as a provisional replacement metric in small western salmonld streams.
X = metric used in region. Many of these variables are applicable elsewhere.
'Excluding individuals of tolerant species.
Taken from Karr et al. (1986), Hughes and Gammon (1987), Miller et al. (1988), Ohio EPA (1987), Steedham (1988), Lyons (1992).
103
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Table 6-3.—Examples of metric suites used for analysis of macrolnvertebrate as-
semblages.
ALTERNATIVE
BENTH1C
METRICS
RBP*
ICI*
RBPb
RBPC
ID
OR
WA
BIBI*
1. Total number taxa
% change in total taxa richness
X
X
X
X
X
X
X
X
X
X
2. Number EPT taxa
# mayfly taxa
# caddlsfly taxa
# stonefly taxa
Missing taxa (EPT)
XXX
X
X
X
X
X
XXX
3. Number diptera taxa
# chlronomidae taxa
X
X
X
4, Number intolerant snail and mussel species
X
5. Ratio EPT/chironomidae abundance
Indicator assemblage index
% EPT taxa
% mayfly composition
% caddisfly composition
X
X
X
X
X
X
X
X
X
6. Percent Tribe Tanytarslni
X
7. Percent other diptera and noninsect
composition
X
8. Percent tolerant organisms
% corbicula composition •
% oligochaete composition
Ratio hydropsychidae/tricoptera
X
X
X
X
X
9. Percent individual dominant taxa
% individual two dominant taxa
Five dominant taxa in common
Commonjaxa index
X
X
X
X
X
X
X
X
X
X
10, Indicator groups
X
X
11. Percent individual omnivores and scavengers
X
12, Percent individual collector gatherers and filterers
. % individual filterers
X
X
X
13, Percent individual grazers and scrapers
Ratio scrapers/filterer collectors
Ratio scrapers/(scrapers + filterer collectors)
X
X
X
X
X
X
14. Percent individual strict predators
X
15. Ratio shredders/total ind. (+ % shredders)
X
X
X
16. Percent similarity functional feeding groups (GSI)
X
X
17. Total abundance
X
X
18. Pinkham-Pearson Community Similarity Index
Community Loss Index
Jaccard Similarity Index
X
X
X
X
19. Quantitative Similarity Index (taxa)
X
X
20. Hilsenhoff Biotic Index
Chandler Biotic Index-
X
X
X
X
X
21. Shannon-Weiner Diversity Index
Equitability
Index of Community Integrity
X
X
X
¦Ohio EPA {1987)
"Barbour et a!. (1992) revised from Plafkin et al, (1089)
'Shackelford (1988)
dHayslip (1992); ID = Idaho, OR = Oregon, WA = Washington (Note: These metrics in ID, OR, and WA
are currently under evaluation.)
'Kerans and Karr (in press)
104
-------
Multimetric Approaches
CHAPTER 6:
for Biocriteria Development
<
that are characterized as opportunists to become substantially more abun-
dant than the rest of the assemblage. As a general rule, dominance of one
or a few species increasing at a site indicates that the influence of human
activities has increased. Comparison to reference conditions provides an
important tool to evaluate the extent to which dominance may reflect hu-
man activities.
Taxonomic composition can be characterized by several classes of in-
formation, including identity and sensitivity. Identity is the knowledge of
individual taxa and associated ecological principles and environmental re-
quirements. Key taxa (i.e., those that are of special interest or ecologically
important) provide information important to the identity of the targeted
assemblages. The presence of exotics or nuisance species may be an im-
portant aspect of biotic interactions that relates to both identity and sensi-
tivity. Sensitivity refers to the numbers of pollutant tolerant and intolerant
species in the sample. The ICI and RBPs use a metric based on species tol-
erance values. A similar metric for fish assemblages is included in the IBI
(Table 6-2). Recognition of rare, endangered, or important taxa provides
additional legal support for remediation activities or recommendations.
Species status for response guilds of bird assemblages — for example,
whether they are threatened or endangered, their endemicity, or some
commercial or recreational value — also relates to the composition class of
metrics (Brooks et al. 1991).
Individual condition metrics characterize assemblage features that re-
sult from sublethal or avoidance response to contaminants. These metrics
focus on low-level chronic exposure to chemical contamination. The con-
dition of individuals can be rated by observation of their physical (ana-
tomical) or behavioral characteristics. Physical characteristics that can be
useful for assessing habitat contaminations result from microbial or viral
infection, teratogenic or carcinogenic effects arising during development
of that individual, or from a maternal effect. These characteristics are cate-
gorized as diseases, anomalies, or metabolic processes (biomarkers).
The underlying concept of the biomarkers approach in biomonitoring
is that contaminant effects occur at the lower levels of biological organiza-
tion (i.e., at the genetic, cell, and tissue level) before more severe distur-
bances are manifested at the population or ecosystem level (Adams et al.
1990). Biomarkers may provide a valuable complement to ecological met-
rics if they are pollutant specific and if the time and financial costs can be
reduced. Unusual behaviors regarding motion, reproduction, or eating
habits are often an indication of physiological or biochemical stress. Often
behavior measures are difficult to assess in the field.
A metric of individual condition is used for fish in the IBI as "percent
diseased individuals" (Table 6-2). The potential for development of
biomarkers in biological monitoring exists. McCarthy (1990) briefly dis-
cussed several studies that have shown biomarker responses to correlate
with predicted levels of contamination and site rankings, based on com-
munity level measures of ecosystem integrity.
Assemblage processes can be divided into several categories as poten-
tial metrics. Trophic dynamics'encompass functional feeding groups and
relate to the energy source for the system, the identity of the herbivores
and carnivores, the presence of detritivores in the system, and the relative
representation of the functional groups. Inferences on biological condition
105
-------
!
BIOLOGICAL CRITERIA:
1 Technical Guidance for Streams and Small Rivers
can often be drawn from a knowledge of the capacity of the system to sup-
port the survival and propagation of the top carnivore. This attribute can
be a surrogate measure for predation rate. Without relatively stable food
dynamics, populations of the top carnivore reflect stressed conditions.
Likewise, if production at a site is considered high based on organism
abundance or biomass, and high production is natural for the habitat type
under study (as per reference conditions), biological condition would be
considered good.
Process metrics have been developed for a number of different assem-
blages. For example, Table 6-2 indicates at least seven IBI metrics dealing
with trophic status or feeding behavior in fish, focusing on insectivores,
omnivores, or herbivores. Also, number or density of individuals of fish in
a sample (or an estimate of standing crop) may be considered a measure of
production and, thus, in the process class of metrics. Additional informa-
tion is gained from density measures when considered relative to size or
age distribution. Three RBP metrics for benthic macroinvertebrates focus
on functional feeding groups (Table 6-3; Barbour et al. 1992; Plafkin et al.
1989). Brooks et al. (1991) use trophic level as one category for rating avian
assemblages.
It may not be necessary to establish metrics for every attribute of the
targeted assemblage. However, the integration of information from several
attributes, especially a grouping of metrics representative of the four ma-
jor classes of attributes (Fig. 2-2), would improve and strengthen the over-
all bioassessment. These metrics can be surrogate measures of more
complicated elements and processes, as long as they have a strong ecologi-
cal foundation and allow biologists to ascertain the attainment or nonat-
tainment of biological integrity.
Index Development
Some investigators have suggested that the Index of Biotic Integrity and
similar multimetric indices have several problems, particularly the over-
simplification of decisions about impairment (Suter, 1993). It is, however,
important to consider how these indices are to be employed. Final deci-
sions on the causes of impairment or management actions are not made on
the single aggregated number alone; rather, if comparisons to established
reference values indicate an impairment in biological condition, then the
component parameters (or metrics) are examined for their individual ef-
fects on the aggregated value. For each metric, a statement is made de-
scribing (1) the derivation of the metric value, (2) the range of possible
values, and (3) the ecological implications and relevance of metric values
(either absolutely or relative to expectations based on defined reference
conditions).
The effects of various stressors on the behavior of specific metrics must
be understood. An often-stated concern is that IBI values will be mislead-
ing unless the relative sensitivity of the monitored populations to specific
pollutants is well known. These concerns are often directed at the use of
subjective tolerance values. In fact, field biologists who have extensive ex-
perience in local fisheries do know the distribution and ecological require-
ments of resident fish species. The general concept of integrating tolerance
information with distributional data has been used successfully in a vari-
106
-------
CHAPTER 6:
Multimetric Approaches for Biocriteria Development
ety of situations (Karr et al. 1986; Ohio EPA, 1987; Hilsenhoff, 1987; Plafkin
et al. 1989).
Normalization — and additive aggregation assumes — that each met-
ric has the same meaning (is weighted the same). It also assumes that a 50
percent change in one metric is of equal value to assessment as a 50 per-
cent change in another. Aggregation simplifies management and decision
making so that a single index value is used to determine whether action is
needed. The exact nature of the action needed (e.g., restoration, mitigation,
pollution enforcement) is not determined by the index value, but by analy-
sis of the component metrics.
The stream invertebrate index for Florida was developed by aggregat-
ing the metrics that proved responsive to independent (but imprecise)
measures of impacts. The approach was to develop expectations for the
values of each of the metrics from the reference data set, and to score met-
rics according to whether they are within the range of reference expecta-
tions. Metric values were normalized into unitless scores. Metrics have
different numerical scales (e.g., percent Diptera; Shannon-Wiener Index)
and must be normalized as unitless values to be aggregated. Metrics
within the range received a high score; those outside received a low score.
The index value was then the same as the metric scores. The index was
further normalized to reference condition, such that the distribution of in-
dex values in the reference sites formed the expectations for the region.
Table 6-4.— Index of Biotic Integrity metrics and scoring criteria based on fish
community data from more than 300 reference sites throughout Ohio applicable
only to boat (i.e., nonwadable) sites. Table modified from Ohio EPA (1987). For
further information on metrics see Ohio EPA (1987).
SCORING DIVISIONS
5
3
1
IBI Metric
METRIC VALUE RANGES
Total no. species
> 20
10-20
< .10
% round-bodied suckers
>38
19-38
<19
No. sunfish species
> 3
2-3
< 2
No. sucker species
> 5
3-5
< 3
No. intolerant species
> 3
2-3
< 2
% tolerant species
< 15
15-27
> 27
% omnivores
.< 16
16-28
> 28
% insectivores
> 54
27-54
< 27
% top carnivores
> 10
5-10
< 5
% simple lithophils*
> 50
25-50
< 25
% DELT anomalies
< 0.5
0
cn
1
CO
o
> 3
Fish numbers
<200
200 - 450
>450
" For sites of a drainage area ^ 600 miles2; for sites of of an area > 600 miles2, scoring cate-
gories vary with drainage area
107
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Table 6-5.— Ranges for Index of Biological Integrity values representing different
narrative descriptions of fish assemblage condition in Ohio streams. Site cate-
gory descriptions — wading, boat, and headwaters — indicate the type of site
and style of samp
ing done at tho
se sites. Mo
Sified from
Ohio EPA (1
987).
SITE CATEGORY
EXCEPTIONAL
GOOD
FAIR
POOR
VERY POOR
Wading
50-60
36-48
28-34
18-26
< 18
Boat
50-60
36-48
26-34
16-24
< 16
Headwaters
50-60
40-48
26-38
16-24 '
< 16
i2
34
30
26
22
18
14
10
~
~
~
Reference Other Impaired
6575A
Refcrenea Other l/npalred
75BCD
—I— Miri-Max
CHI 25%-75%
0 Median value
Figure 6-8,—Invertebrate stream index scores for Florida streams.
Ohio EPA (1987) establishes tables based , on some predetermined per-
centiles as discussed above. They recognize three categories of metric scor-
ing ranges for fish assemblage data collected at nonwadable sites (boat
sites) (Table 6-4). Ohio EPA (1987) compared individual metric values from
sites constituting the reference database to Table 6-4 or similar tables to de-
velop total site scores (aggregated values from 12 normalized metrics) for
each of three different types of sites: (1) wadable, nonheadwater streams;
(2) nonwadaHle channels requiring boats for sampling; and (3) headwater
streams. These total scores were then used to establish assessment catego-
ries (Table 6-5), which are the quantitative basis of biological criteria.
The test of the aggregated index is in the ability to strengthen the dis-
crimination between least impaired and impaired conditions beyond that
of the individual metrics. This concept is illustrated in Figure 6-8, as it was
done for Florida streams. In some state programs, e.g., Maine and North
Carolina, the metrics are treated as individual measures and are not aggre-
gated to form a composite index. For instance, Maine DEP uses as many as
30 biological metrics (macroinvertebrates) to assess attainment of its
aquatic life use classes. A threshold coefficient has been established for
each metric to be used in a linear discriminant model to test for class at-
tainment. In North Carolina, macroinvertebrate metrics of Taxonomic
108
-------
CHAPTER 6:
Multimetric Approaches for Biocriteria Development
Richness, Biotic Index, and EPT Index are the primary metrics of concern
in evaluating attainment of their bioclassification criteria for North Caro-
lina's three physiographic provinces.
Multivariate Approaches
An alternative approach to multimetric indices is multivariate analysis of
species composition (e.g., Wright et al. 1984; Moss et al. 1987; Furse et al.
1987). The approach consists of developing a model that predicts the ex-
pected species composition for sites given their physical and chemical
characteristics. Then the observed species composition at a site is com-
pared to the expected species composition predicted by the model. The
model characterizes reference conditions, and assessment sites are com-
pared to model-predicted reference conditions.
In the first step of this approach, a classification is developed from
species abundance data at reference sites using one or more multivariate
clustering or ordination techniques (Ludwig and Reynolds, 1988). Dis-
criminant analysis is then applied to the class assignments and the corre-
sponding physical-chemical data to develop the model for predicting class
membership from subsequent physical-chemical site data (Wright et al.
1984). An assessment site is then assigned to a class using the discriminant
functions, and its observed species composition is compared to the.ex-
pected species composition (Moss et al. 1987; Furse et al. 1987). An alterna-
tive to discriminant analysis is direct analysis of associations between
species composition and environmental variables using methods such as
canonical correlation analysis, canonical correspondence analysis,- or mul-
tidimensional scaling.
Such multivariate approaches for bioassessment are still under devel-
opment. A predictive model requires extensive physical-chemical data on
the reference sites, and there is no assurance that a discriminant model
will work well and produce a minimum of misclassifications. The better
discriminant models using the above approach misclassify in the range of
25 to 34 percent (Moss et al. 1987). Assessment thresholds and standard
procedures are not yet well developed for multivariate assessment, other
than professional judgment on missing taxa, similarity indices, or metrics.
Nonetheless, as this approach becomes more refined, it may prove to be a
viable option to multimetric indexing. In fact, Maine is presently using a
combination of the two with promising results.
Suggested Readings
Barbour, M.T. et al. 1992. Evaluation of EPA's rapid bioassessment benthic metrics: met-
ric redundancy and variability among reference stream sifes. J. Environ. Toxicol.
Chem. 11(4).
Brooks, R.P. et al. 1991. Selection of biological indicators for integrating assessments of
wetland, stream, and riparian habitats. Pages 81-89 in Biological Criteria: Research
and Regulation. EPA-440/5-91-005. Off. Water, U.S. Environ. Prot. Agency, Wash-
ington, DC.
Gray, J.S. 1989. Effects of environmental stress on species rich assemblages. Biolog. J.
Linnean Soc. 37:19-32.
109
Multivariate
approaches for
bioassessment are
still under
development. . . .
Nonetheless, as this
approach becomes
more refined, it may
prove to be a viable
option.
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Hayslip, G.A. 1992. EPA Region 10 In-stream Biological Monitoring Handbook for Wad-
able Streams in the Northwest. Draft. EPA-910/9-92-013. Environ. Serv. Div., Reg.
10, U.S. Environ. Prot. Agency, Seattle, WA,
Katr, J.R. et al. 1986. Assessing Biological Integrity in Running Waters: A Method and Its
Rationale. Spec. Publ. 5. Illinois Nat. History Surv., Urbana, IL.
Kerans, B.L., and J.R. Karr. In Press. A benthic index of biotic integrity (B-IBI) for rivers
of the Tennessee Valley. Ecol. Appl. 4.
Kerans, B.L., J.R. Karr, and S.A. Ahlstedt. 1992. Aquatic invertebrate assemblages: spa-
tial and temporal differences among sampling protocols. J. N. Am, Benthol.
Soc.ll(4): 377:90.
Miller et al. 1988. Regional applications of an index of biotic integrity for use in water
resource management. Fisheries 13(5): 12-20.
Ohio Environmental Protection Agency. 1987. Biological Criteria for the Protection of
Aquatic Life. Vol. 1-3. Surface Water Sec., Div. Water Qual. Monitor. Assess., Colum-
bus, OH.
Plafkin, J.L. et al. 1989. Rapid Bioassessment Protocols for Use in Streams and Rivers:
Benthic Macroinvertebrates and Fish. EPA/444/4-89-001. Off. Water, U.S. Environ.
Prot. Agency Washington, DC.
110
-------
CHAPTER 7.
Biocriteria Development
and Implementation
The first phase in a biocriteria program is the development of "narra-
tive biological criteria" (Gibson, 1992). These criteria are essentially
statements of intent incorporated in state water laws to formally consider
the fate" and status of aquatic biological communities. As stated in that
guidance, attributes of sound biological criteria include the following ob-
jectives:
1. Support the goals of the Clean Water Act to provide for the protec-
tion and propagation of fish, shellfish, and wildlife, and to restore
and maintain the chemical, physical, and biological integrity of the
nation's waters.
2. Protect the most natural biological community possible by empha-
sizing the protection of its most sensitive components.
3. Refer to specific aquatic, marine, and estuarine community charac-
teristics that must be present for the waterbody to meet a particu-
lar designated use, for example, natural diverse systems with their
respective communities or taxa indicated.
4. Include measures of community characteristics, based on sound
scientific principles, that are quantifiable and written to protect
and/or enhance the designated use.
5. In no case should impacts degrading existing uses or the biological
integrity of the waters be authorized.
Establishing Regional Biocriteria
The first decision that a resource agency must make is to determine the set
of sites or class to which a biocriterion applies. Site classification (Chapter
3) permits more refined characterization of the reference condition and
therefore better resolution in detecting impairment. Any characterization
of a reference condition should account for the variability in the biological
data used to establish the biocriteria. Thus, the reference condition can be
characterized by measures of central tendency (mean, median, trimmed
mean) and by variability (standard deviation, quartiles, ranges).
Purpose:
To'provide water
resource agencies
with guidance for
biocriteria
development and
implementation.
111
-------
BIOLOGICAL CRITERIA
Technical Guidance for Streams and Small Rivers
Statewide characterization of reference condition can be expected to'
exhibit high variance; however, successive intrastate classification will
partition the variance from within a large class to among several different
component classes. The goal of classification is to minimize within-class
variability by allocating the variability to among-class differences. When
this goal is achieved, it results in less variation per class and greater reso-
lution of the criteria.
Classification into aquatic types (regional or specific habitat types)
should partition overall variance (to achieve lower variability within each
class than among classes). The central tendency of each class may be ex-
pected to differ (otherwise variability would not be reduced within classes as
compared to all classes combined). Investigators for Ohio EPA chose to class-
ify by ecoregion and by aquatic life use. Thus, for each ecoregion and for each
aquatic life use within that region, they can characterize a central tendency
and variability for the reference condition (from their reference sites).
The more refined the classification, the more precisely the reference
condition can be defined; however, an agency also needs to decide when
enough classification is enough. Classification can be discrete, as in ecore-
gions, or continuous, as along a gradient where, for example, expected
species richness is a function of stream size.
Biocriteria programs can use discrete and continuous classifications si-
multaneously; Ohio EPA (1987) has biocriteria that vary by stream size
and drainage area within its established ecoregions. The agency's calibra-
tion procedures allow investigators to normalize the effects of stream size
so that index scores, such as the IBI, can be compared among all streams of
a region. For example, the ratio of fish species richness to stream size is an
empirical model that accounts for overall variation in species, regardless
of stream size. In evaluating whether a test site achieves its species rich-
ness potential (a possible biological criterion), one would surely like to
take into account the stream size factor. It would be unfair to expect a
small stream (with a limited capacity to support a species-rich fish biota)
to achieve a high species richness (relative to all streams). By the same to-
ken, it would not be good stewardship to allow a large stream (with ex-
pected high species richness) to meet attainment merely because its size
achieves the statewide criterion.
Designing the Actual Criterion
Having selected its classification scheme, reference sites, and metrics, the
agency now has the basic material needed to design the actual criterion.
What statistic should be used? A variety of choices are available for meas-
uring central tendency and variability. Two general approaches have
evolved, however, for the selection of a quantitative regional biocriterion:
the first uses an aggregate or index of metric values, each of which has
been assigned a percentile along the distribution of represented minimally
impaired sites (Ohio and Florida); the second, a multivariate analysis of
metrics or other basic biological data to develop expected thresholds or at-
tainment (Maine).
The percentile that is established for each metric in the first approach
is a threshold from which quartiles can be determined for a score ranking
system (see chapter 6). The aggregation of these scores for the reference
condition functions as the basis for biocriteria.
112
-------
CHAPTER 7;
Biocriteria Development and "Implementation
An example of the second approach is the hierarchical decision-mak-
ing technique used by Maine. It begins with statistical models (linear dis-
criminant analysis) to make an initial prediction of the classification of an
unknown sample by comparing it to characteristics of each class identified
in the baseline database (Davies et al. 1991), The output from analysis by
the primary statistical model is a list of probabilities of membership for
each of four classes (A, B, C, and nonattainment of Class C). Subsequent
models are designed to distinguish between a given class and any higher
classes as one group, and any lower classes as a second group (Fig. 7-1).
An important consideration is how conservative or protective the
agency wants to be. The more conservative the resource agency, the more
likely it is that the criterion will be set at the upper end of the condition
spectrum. The more liberal the agency is in assessing impairment and
maintaining the aquatic life use, the more liberal the criterion will be. Ex-
amining the variance structure in a manner similar to that described ear-
lier helps validate the extent to which particular biocriteria apply. If there
is little biotic variation evident among the initial regions, or if their differ-
ences can be associated with management practices that can be altered, it
seems wise to combine those regions to adhere to the same biocriteria.
In the absence of a strong case for subregional biocriteria, it is prob-
ably better to overprotect by setting high biocriteria over broad regions
than to underprotect by using too low a threshold. Procedures can then be
developed that allow for both regional and subregional deviations from
the broadly established biocriteria if, and only if, the deviation is justified
by natural anomalies.
In these instances, some site-specific rules of exception to regional biocrit-
eria are necessary to accommodate natural limitations. For example, certain
natural channel configurations, such as those flowing through bedrock or
those that have natural barriers to dispersal, do not offer the habitat diversity
of other channel configurations. They cannot, therefore, support the richness
Some site-specific
rules of exception to
regional biocriteria
are necessary to
accommodate natural
limitations
FIRST STAGE MODEL-
(^Noi^ttaiimienP^
SECOND STAGE MODELS-
C or Better Key
(A+ B + C}
v —
YS
Per Better Key
c~a + iT) vs
VS
Cj
B + C + NA
Figure 7-1.—Hierarchy of statistical models used In Maine's bjological criteria pro-
gram (taken from Davies et al. 1993).
113
-------
BIOLOGICAL CRITERIA
Technical Guidance for Streams and Small Rivers
mmm
I he objective in
setting biocriteria is to
improve the quality of
our water resources.
Therefore, criteria
must not be
predicated on
accepting the
existing, degraded
conditions as a matter
of course.
In significantly
impaired areas, the
lowest potentially
acceptable criterion is
the "best, most
natural condition
remaining in the
region."
and diversity of other nearby channel types. Other natural restrictions to
achievement can also be identified, but care must be taken that culturally
degraded conditions are not included as evidence for regional biocriteria
modification.
Biocriteria for Significantly Impacted Areas
A key element in setting biological criteria is to avoid establishing unduly
low thresholds. The objective is to improve the quality of our water re-
sources; therefore, criteria must not be predicated on accepting the existing
degraded conditions as a matter of course. In significantly impaired areas, the
lowest potentially acceptable criterion is the "best, most natural condition re-
maining in the region" as defined by a review of the classification data. The
upper range for such criteria should be the best condition that is physically
and economically achievable by restoration management activities.
This determination is best made by an objective and balanced panel of
experts representing the research community, industry, and local, state,
and federal water resources specialists using information developed from
current and historical data. The actual selection, that is, the point within
this range that will become the criterion, should also be established by this
panel. This criteripn is expected to move upward periodically as manage-
ment efforts improve the resource condition. A review process should be
keyed to the periodic calibrations of regional reference conditions con-
ducted by the states.
There may be no acceptable reference sites in significantly impaired re-
gions. In these areas, an ecological model based on (1) neighboring site
classes, (2) expert consensus, and (3) composite of "best" ecological infor-
mation, may be used (Fig. 3-1). The. resultant biocriteria may be an interim
or hypothetical expectation that will improve with restoration and mitiga-
tion.
Selecting the Assessment Site
Assessment sites should be established to evaluate the effects of human
activities on water resources. Potential assessment sites can be identified
from land use and topographic maps; specific information can be pro-
vided by state and county personnel familiar with the areas. Such sites are
generally selected to reflect the influence of known or suspected point and
nonpoint source pollution loadings. Final selection should be made only
after field reconnaissance by qualified staff at the site verifies that the
documented conditions are accurate.
For discrimination of sources and causes of impairment, an agency
may need to establish an "impaired" sites database with similar impair-
ments to compare with information at aquatic community test sites. These
comparisons can be made using biological response signatures (Yoder,
1991). A biological response signature is a unique combination of biological
attributes that identify individual impact types or the cumulative impacts
of several related human influences. For best results, this process requires
the development of an extensive database.
114
-------
CHAPTER 7:
Biocriteria Development and Implementation
¦ National Pollutant Discharge Elimination System (NPDES) Permit
Requests or Renewals. Public or private wastewater treatment plant admin-
istrators and industrial dischargers must apply for NPDES permits. If the
number of test sites prohibits annual or more frequent monitoring surveys, a
percentage can be surveyed on a rotational basis each year. Priorities can be
assigned to permits requiring the earliest renewal or peraiit award and those
in the same geographical area or watershed.. Other permitting programs in-
clude hazardous waste site regulation, Clean Water Act, section 404/401,
dredge and fill certification programs, and construction sites.
¦ Locations of Concentrated Commercial or Industrial Discharges. In
addition to specified permit locations, states may find it appropriate to es-
tablish nonspecific monitoring stations along the stream system. These
stations can be particularly helpful if located between clusters of commer-
cial, industrial, or municipal operations, to help distinguish among poten-
tial sources and between groups of users. In addition, the use of
nonspecific monitoring stations will help to distinguish discharge effects
from preexisting upstream impacts, a distinction particularly helpful
given the typical sequential placement of textile or lumber mill operations
along small river courses.
¦ Agricultural Concentrations. Areas of intensive and extensive farming
activities are appropriate for the placement of test sites because they can
help isolate potential nonpoint source loadings or impairments. Such ar-
eas of interest include croplands, rangelands, clearcuts, feedlots, animal
holding facilities, manure holding systems, convergent field drainings,
contiguous farms, and fertilizer, feed, and pesticide storage facilities.
County' agricultural extension agents can help determine site placements.
They can also identify high risk localities and farms engaged in coopera-
tive conservation programs and suggest appropriate remedial land use
practices and programs if and when problems are identified.
¦ Urban Centers. The locations of shopping centers, commercial districts,
and residential areas that include stormwater runoff concentrations are a
source of impact to watersheds. Also of interest are urban developments
in riparian zones (areas bordering waterbodies), whether or not they con-
tain wastewater treatment plants. On-site wastewater disposal is common
in older communities on small lots concentrated near the waterway. The
potential septic system problem in these communities can be compounded
by an overburdened stormwater drainage network.
¦ Transportation Services. Vehicle and other traffic modes also affect
water resources: major highway interchanges near a watercourse; streams
paralleled by extensive, heavily traveled roads or railroads; heavily trav-
eled bridge or overpass systems; pipelines; and maintenance facilities in-
cluding stockpiles of deicing salt located near a stream system. Airports
and railroad or truck marshaling yards may also generate surface runoff
problems for nearby stream systems.
¦ Mining and Logging Activities. Any area affected by cumulative and
sequential mining activities and effects including road construction, drill-
ing wells, logging prior to mineral extraction, and acid mine drainage
should be evaluated for test site placement. The basis for such decisions
will be state mining permit records and associated maps because the areas
For discrimination .of
sources and causes
of impairment, an
agency may need to
establish an
"impaired" sites
database with similar
impairments to
compare with
information at aquatic
community test sites.
115
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
of potential impact, especially from subsurface mining and abandoned
mines, may not be self-evident.
¦ Forest Management Activities. Any areas affected by logging and saw-
mill activities should be evaluated for test site placement. Instability cre-
ated by road construction in timber areas is especially damaging to water
resources. Effective forestry best management practices (BMPs) will be im-
portant influences in these areas. Protection of these areas is critical be-
cause many of the representative reference sites will be located in forested
lands. Federal and state foresters need to interact with state water quality
agencies for identification of sensitive areas.
¦ Disruptive Land Use Activities. This category will include a variety of
planned or existing construction projects: landfills; channelization or other
in-stream projects such as dams and flood control structures, fish hatcher-
ies, or aquaculture. Any of these activities on a significant scale or near
streams should be monitored and evaluated. If advance notice of these ac-
tivities is provided, states should establish both spatial and temporal
monitoring before, during, and after the activities for biological assess-
ments.
¦ Land Use Activities in Unsurveyed or Remote Areas. This category in-
cludes regions not previously surveyed for which no preexisting informa-
tion would be available in the event of a spill or major hydrological
calamity and remote sites for which development is planned in the near or
distant future. Long-term antecedent biological information should be a
component in new development planning.
Evaluating the Assessment Site
Statistically evaluating the test site(s) against the reference condition to as-
sess the extent and degree of impairment is the focus of another document
(Reckhow, in review); however, the basic question is this: What evidence
do we have that indicates impairment (or absence of impairment)? If the
assessment is based on a reference condition determined from a composite
of sites, the manager's confidence in the judgment is improved over that
from use of a single reference site — notwithstanding that some level of
precision may be lost (see Chapter 3).
The simultaneous comparison of an assessment site to a site-specific ref-
erence condition is an alternative that is generally undertaken as an up-
stream/ downstream or paired watershed approach. Presumably the
site-specific reference condition represents the best attainable condition of the
assessment site(s). In this approach, the percent-of-reference may be the most
appropriate criterion from which to assess impairment. States that have lim-
ited resources may wish to implement this approach as an interim until a
larger database is developed. The assessment of sites follows the same guide-
lines whether reference data are site-specific or regional (Table 7-1).
Assessment sites are points or reaches on a stream at which distur-
bance is suspected or from which information about the location's relative
quality is desired. In selecting assessment sites, the latitude of selection
compared to the choice of reference sites may be considerably reduced. If
the area is suspect, it must be investigated regardless of its stream charac-
Assessment sites are
points or reaches on
a stream at which
disturbance is
suspected or from
which information
about the location's
relative quality is
desired.
116
-------
CHAPTER 7:
Biocriteria De velopment and Implementation
Table 7-1 .—Sequential process for assessment of test sites and determination
of their relationship to established biocriteria. Refer to Chapter 6 for an
explanation of biocriteria establishment.
ASSESSMENT PROCESS
Step 1.
Determine Class
• same classification scheme as for reference sites
Step 2.
Survey Assessment Sites
« biota and physical habitat
Step 3.
Calculate Metrics
• convert raw data to metric values
Step 4.
¦ ¦
Aggregate Metrics to Form Indices
* use scoring rules established for metrics
» sum normalized metric values
Step 5.
Compare to Reference (Biocriteria)
• use established regional biocriteria for assessment
Step 6.
Statement of Condition
• characterize existence and extent of impairment
• diagnostics as to stressors
teristics or channel configuration. Thus, regionalized reference conditions,
while necessary for criteria development, may not always be sufficient to
serve as a foundation for expecting a specific biological condition. The in-
vestigator facing a potentially contentious situation may find it prudent to
augment the regional reference data with results of locally matched refer-
ence sites, such as upstream sites or sites in similar, nearby streams.
The assessment process is essentially a replication of the procedure de-
scribed earlier to develop multiple metrics (see Chapter 6 and Fig. 6-2).
Note, however, that the move from the development of metrics and indices
to their use in the assessment process leads directly to the development
and implementation of biocriteria. The assessment process, summarized in
Table 7-1 and illustrated in Figure 7-2, is described as follows:
Step 1 — Classification of Assessment Sites. Sites selected for assess-
ment are assigned to the appropriate classification derived from the
initial reference classification scheme. The assessment site is classified
according to the stream class designations, not the nature of a sus-
pected land use or point-source discharge impact. In other words,
similar receiving waters should be in the same classification whether
or not there are similar discharges to those waters.
Step 2 — Biosurvey. Stream or small river biological communities and
habitat characteristics should be measured using the same techniques
and equipment as were used at designated reference site(s). It will also
be necessary to gather data during the same time frame. This schedule
may not coincide with a predetermined indexing period. For example,
if a construction site is scheduled to open on a particular date or if a
critical period of operation is approaching, both the test and reference
site(s) will have to be surveyed accordingly.
Step 3 — Calculate Metrics. Many of the intermediate steps used in
the criteria development process become unnecessary at this point. In-
vestigators can simply enter the appropriate raw data from the refer-
117
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Fish Species
Counts a Weights
Expectations
for warm water
streams
2
Species
% Exotic
% Sensitive
Richness
Species
Species
Indicators
Aggregation
WB Score
Geographic region
and stream type
Statement of
Condition
Figure 7-2.—The process for proceeding from measurements of fish assemblage to
indicators such as the Index of Biotic integrity (IBI) or Index of Weil Being (IWB) — as
used to develop criteria and apply those criteria to streams (modified from Paulsen et
ai. 1991).
ence and test sites into a preselected format to generate current met-
rics. In all cases, the integrity of the raw data should be presumed for
support and as additional information for more definitive assessment.
Step 4 — Calculate Indices. Where indicated, these metrics are simi-
larly summarized in indices of relative biological condition and habitat
description. Some states do not use indices but evaluate the informa-
tion from the Individual metrics as independent measures of biological
condition.
Step S — Compare to Appropriate Biological Criteria. The biological
data from the site under assessment are compared to established crite-
ria to ascertain the status. Both the indices (aggregation of metrics) and
118
-------
CHAPTER 7:
Biocriteria Development and Implementation
the individual metrics are evaluated as part of the assessment. All
available information must be used to confirm the status of the bio-
logical condition and to diagnose the cause and effect relationship if
impairment is detected.
Step 6 — Statement of Condition. At this point, the assessment sites
are evaluated to determine whether they do or do not meet the crite-.
ria. The sites can also be placed in priority order using the details of
this evaluation to support management plans and resource allocations.
Further refinement of the data collected and additional investigations
can help determine cause and effect relationships among the stresses
identified by this process. Such information will be essential to suc-
cessful remedial management.
Overview of Selected State Biocriteria Programs
¦ Maine. In 1986, the State of Maine enacted legislation that mandated an
objective "to restore and maintain the chemical, physical, and biological
integrity" of Maine waters. In addition, a legislative water quality classifi-
cation system was established to manage and protect the quality of Maine
waters. The classification system established minimum standards for des-
ignated uses of water and related characteristics of those uses (Table 7-2).
Within each use-attainability class, the minimum condition of aquatic life
necessary to attain that class is described.
. The descriptions or narrative standards in this legislation range from
statements such as "Change in community composition may occur" (Class
C) to "Aquatic life as naturally occurs" (Class A and AA). The designated
use classes were recombined into four biologically discernible classes (Ta-
ble 7-2): Classes A and AA were combined, and a fourth class, nonattairi-
ment of Class C, was added.
The Maine Department of Environmental Protection has assessed a
large, standardized macroinvertebrate community database from samples
taken above and below all major point-source discharges, as well as sam-
ples from relatively undisturbed areas. Maine used this database as a cali-
bration dataset to develop discriminant functions for classifying sites
among the four analytical classes.
The calibration data set consisted of the general level of abundances
from 145 rock basket samples collected from first to seventh order streams
throughout Maine, and covering a wide range of relatively unimpacted
and impacted streams. General abundances were reduced to approxi-
mately 30 quantitative metrics.
The calibration data set was given to five stream biologists to assign
the 145 sites to the four classes (A, B, C, and NA) using professional judg-
ment. The biologists used only the biological data; they did not see loca-
tions, names, habitat, or site chemistry. Disagreements on class
assignments were resolved in conference.
The resultant metrics and class assignments were then used to develop
linear discriminant models to predict class membership of unknown as-
sessment sites. Two stages of discriminant models were developed from
the calibration data set: the first stage estimates the probability that a site
belongs to one of the four classes (A, B, C, or NA); the second stage esti-
119
-------
BIOLOGICAL CRITERIA
Technical Guidance for Streams and Small Rivers
Table 7-2.—Maine's water quality classification system for rivers and streams,
with associated biological standards (taken from Davies et al. 1993).
AQUATIC LIFE
USE CLASS
MANAGEMENT
BIOLOGICAL
STANDARD
DISCRIMINANT
CLASS
AA
i
High quality water for
recreation and ecological
interests. No discharges or
impoundments permitted.
Habitat natural and free
flowing. Aquatic life as
naturally occurs.
A
A
High quality water with limited
human interference.
Discharges restricted to
noncontact process water or
highly treated wastewater
equal to or better than the
receiving water,
impoundments allowed.
Habitat natural. Aquatic
life as naturally occurs.
A and AAare
indistinguishable
because biota
are "as
naturally
occurs."
B
Good quality water. Discharge
of well treated effluent with
ample dilution permitted.
Habitat unimpaired.
Ambient water quality
sufficient to support life
stages of all indigenous
aquatic species. Only
nondetrimental changes
in community
composition allowed.
B
C
Lowest water quality.
Maintains the interim goals of
the Federal Water Quality Act
(fishab le/swim mable).
Discharge of well-treated
effluent permitted.
Ambient water quality
sufficient to support life
stages of all indigenous
fish species. Change in
community composition
may occur but structure
and function of the
community must be
maintained.
C
NA
Not attaining
Class C
mates two-way probabilities that a site belongs to higher or lower classes
(i.e., A, B, C. vs. NA; A, B, vs. C, NA; and A vs. B, C, NA). Each model uses
different metrics.
In operational assessment, sites are evaluated with the two-step hier-
archical models. The first stage linear discriminant model is applied to es-
timate the probability of membership of sites into one of four classes (A, B,
C, or NA). Second, the series of two-way models are applied to distin-
guish the membership between a given class and any higher classes, as
one group (Fig. 7-1). Monitored test sites are then assigned to one of the
four classes based on the probability of that result, and uncertainty is ex-
pressed for intermediate sites. The classification can be the basis for man-
agement action if a site has gone down in class, or for reclassification to a
higher class if the site has improved.
Maine biocriteria thus establish a direct relationship between manage-
ment objectives (the three aquatic life use classes and nonattainment) and
biological measurements. The relationship is immediately viable for man-
agement and enforcement as long as the aquatic life use classes remain the
same. If the classes are redefined, a complete reassignment of streams and
a review of the calibration procedure will be necessary.
1
-------
CHAPTER 7:
Biocriteria Development and Implementation
B North Carolina. The North Carolina Department of Environment,
Health and Natural Resources, Division of Environmental Management,
Water Quality Section has written Standard Operating Procedures for the
collection of biological data and the bioclassification of each station sam-
pled. Biological criteria have been included in the North Carolina water
quality standards as written narratives. Narrative standards have been in
place since 1983. They support the use of biological assessments.in point
and nonpoint source evaluation, and help identify and protect the best uses
of North Carolina waters. High Quality Waters, Outstanding Resource Wa-
ters and Nutrient Sensitive Waters are assessed using biocriteria.
Phytoplankton, aquatic macrophytes, benthic macroinvertebrates, and
fish are routinely collected as part of North Carolina's biosurvey effort.
Only the macroinvertebrate biosurvey data and the associated bioclassifi-
cation system are summarized here.
Macroinvertebrates are sampled qualitatively by one of two methods:
a Standard Qualitative Method or the Ephemeroptera, Plecoptera, and
Trichoptera (EPT) Survey Method. When following the Standard Qualita-
tive Method, two kick net samples from cobble substrate, three dip-net
samples (sweeps) from vegetation and shore zones, one leaf pack sample,
two fine-mesh rock and/or log wash samples, one fine-mesh sand sample,
and visual inspection samples are taken.
The EPT survey method focuses on qualitative collection of Ephe-
meroptera, Plecoptera, and Trichoptera, by collecting one kick sample, one
sweep sample, one leaf-pack sample and visual collections. With both
methods, invertebrates are sorted in the field using forceps and white
plastic trays, and preserved in glass vials containing 5 percent ethanol. Or-
ganisms are sorted in approximate proportion to their relative abundance.
Currently, site-specific reference conditions are typically used when
conducting surveys. However, where site-specific reference sites are not
available, ecoregional reference conditions are used to define unimpaired
conditions. North Carolina is developing ecoregional reference conditions
based on the available land use information. The three major ecoregions
identified in North Carolina are Mountain, Piedmont, and Coastal Plain.
Specific macroinvertebrate metrics, including taxonomic richness, biotic
indices, an Indicator Assemblage Index (IAI), diversity indices (Shannon's
Index), and the Index of Community Integrity (ICI) are used to rate sites as
poor, fair, good/fair, good, and excellent. The ratings are conducted in addi-
tion to the narrative descriptions for biocriteria. These metrics are used as
independent measures rather than aggregated into an overall index.
Bioclassification criteria for the Mountain, Piedmont, and Coastal Plain
ecoregions in North Carolina have been developed for EPT taxa richness
values. This community metric has been developed using both the Standard
Qualitative Method and the EPT Survey Method. The bioclassification rat-
ings for the number of EPT taxa in each ecoregion for both the Standard
Qualitative Method and the EPT method are summarized in Table 7-3. Note
that the rating system has been developed solely on summer (June-Septem-
ber) collections. Samples collected in other seasons, therefore, must be sea-
sonally corrected before a bioclassification can be assigned.
The North Carolina claissification system was developed for chemical
impact assessment and does not address sedimentation or other habitat al-
teration effects. A special bioclassification rating has also been developed
121
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Table 7-3.—Bioclassification criteria scores for EPT taxa richness values for
three North Carolina ecoregions based on two sampling methods.
STANDARD QUALITATIVE METHOD
BIOCLASSIFICATION
MOUNTAIN
PIEDMONT
COASTAL PLAIN
Excellent
>41
>31
>27
Good
32-41
24-31
21-27
Fair
12-21
8-15
7-13
Poor
0-11
0-7
0-6
EPT QUALITATIVE METHOD
BIOCLASSIFICATION
MOUNTAIN
PIEDMONT
COASTAL PLAIN .
Excellent
>35
>27
>23 I
Good
28-35
21-27
18-23 I
Good-Fair
19-27
14-20
12-17 I
Fair
11-18
7-13
6-11 I
Poor
0-10
0-6
0-5 |
for "small, high quality mountain streams which naturally exhibit a re-
duced macroinvertebrate taxa number. Streams possessing these particu-
lar characteristics, having EPT taxa of a 29 (Standard Qualitative Method)
or ^ 26 (EPT Survey Method) are considered excellent.
¦ Ohio. Ohio's biological criteria program was developed for complete
integration with state water quality standard regulations. As such, biocrit-
eria in Ohio are fully integrated with typical water quality measures, and
address three key strategic goals:
• The protection of aquatic life in all Ohio waterways capable of support-
ing aquatic life is an immediate goal of the Ohio EPA to be accomplished,
wherever possible, through a "systems" (biological community re-
sponse) approach.
• Short- and long-range goals must be established for the control of toxic
substances in Ohio's surface waters.
• The protection of human health through the assurance of a "safe" level of
exposure to toxic substances in water and fish is an immediate goal of the
Ohio EPA.
To accomplish these goals, the Ohio EPA program combines biocrite-
ria, effluent toxicity, and water chemistry. This integrated approach has
significantly increased Ohio EPA's ability to detect degradation, particu-
larly in streams receiving point and nonpoint sources and both toxic and
conventional pollutants.
The Ohio EPA has employed the concept of tiered aquatic life uses in
the Ohio Water Quality Standards (WQS) since 1978. Aquatic life uses in
Ohio include the Warmwater Habitat (WWH), Exceptional Warmwater
Habitat (EWH), Cold-water Habitat (CWH), Seasonal Salmonid Habitat
(SSH), Modified Warmwater Habitat (three subcategories: channel-modi-
fied, MWH-C; affected by mines, MWH-A; and impounded, MWH-I),
Limited Resource Water (LRW) (Ohio EPA 1992). Each of these use desig-
nations are defined in the Ohio WQS.
122
-------
- CHAPTER 7:
Biocriteria Development and Implementation
Water quality standards constitute the numerical and narrative criteria
that, when achieved, will presumably protect a given designated use
(Ohio EPA 1992). Chemical-specific criteria serve as the "targets" for was-
teload allocations conducted under the TMDL (Total Maximum Daily
Load) process, which is used to determine water quality-based effluent
limits for point' source discharges and, theoretically, load allocations for
nonpoint sources (in connection with best management practices). Whole
effluent toxicity limits consist of acute and chronic endpoints (based on
laboratory toxicity tests) and a dilution method similar to that used to cal-
culate chemical-specific limits. The biological criteria are used to directly
determine aquatic life use attainment status for the EWH, WWH, and
MWH use designations as is stated under the definition of each in the
Ohio WQS.
The biological criteria designed for Ohio's rivers and streams incorpo-
rate the ecoregional reference approach. Within each of the State's five
ecoregions, criteria for three biological indices have been derived. The in-
dices include two measures of fish community structure and one measure
of the benthic macroinvertebrate community. The combined indices pro-
vide a quantitative measure that can be compared to regional reference in-
dices to assess use attainment.
The two fish community measures include the Index of Biotic Integrity
(TBI) and the modified Index of Well Being (IWB). Both indices incorporate
structural attributes of the fish community, while the IBI additionally in-
corporates functional (trophic) characteristics. The two indices incorporate
a range of fish community attributes much broader than only species rich-
ness and relative abundance. For macroinvertebrate community measure-
ments, Ohio EPA uses the Invertebrate Community Index (ICI). The ICI is
a modification of the IBI concept, but has been adapted for use with
macroinvertebrates. Like the IBI, ICI values incorporate functional aspects
of the community.
Derivation of the above indices requires extensive sampling to provide
the quantitative data necessary for analysis. The IBI and IWB require sam-
pling of approximately 500 meters of a river or stream by electroshocking
to characterize the community of fish. Data recording is extensive, and in-
cludes fish species, number of individuals per species, and various obser-
vations of fish condition. The ICI requires that quantitative
(Hester-Dendy) and qualitative macroinvertebrate samples be collected.
Laboratory analysis of these samples includes taxon determination to ge-
nus or species, and quantification of the organisms collected.
The Exceptional Warmwater Habitat (EWH) is the most protective use
assigned to warmwater streams in Ohio. Ohio's biological criteria for
EWH applies uniformly statewide and is set at the 75th percentile index
values of all reference sites combined. The Warmwater Habitat (WWH) is
the most widely applied use designation assigned to warmwater streams
in Ohio. The biological criteria for fish vary by ecoregion and site type and
are set at the 25th percentile index values of the applicable reference sites
in each ecoregion (Fig. 7-3a). A modified procedure was used in the exten-
sively modified Huron Erie Lake Plain (HELP) ecoregion.
The Modified Warmwater Habitat (MWH), first adopted in 1990, is as-
signed to streams that have had extensive and irretrievable physical habitat
modifications. The MWH use does not meet the Clean Water Act goals
and therefore requires a Use Attainability Analysis. There are three sub-
123
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Fish — Boat Sites
•y/yAvyi
34/8.6
^vT
?ygggg»gg
$40/8.6 ft
«»»«
<£«•«
«««&««?
y$Mm&
Fish — Headwater Sites
mmsr
<0
Fish — Wading Sites
(iBMwb)
IWH
40/8.3
gl44/8.4gg
r,y>y,tih>y?>y
40/8.1
EWH
Macroinvertebrates
( ICI )
j Huron Erie Lake Plain - HELP
1 Interior Plateau - IP
EWH
j Eastern-Ontario Lake Plain - EOLP
| Western Allegheny Plateau - WAP
P«
m>pv
Dt.
EWH
I Eastern Com Belt Plains - ECBP
Figure 7-3a.— Biological criteria In the Ohio WQS for the Warmwater Habitat (WWH)
and Exceptional Warmwater Habitat (EWH) use designations arranged by biological
Index, site type for fish, and ecoregion. Index values in the boxes on each map are
the WWH biocriterla that vary by ecoregion as follows: IBI/Mlwb for Boat Sites (upper
left), IBI/Mlwb for Wading Sites (upper right), IBI for Headwaters Sites (lower left), and
the ICI (lower right). The EWH criteria for each index and site type are located in the
boxes Just outside each map (Ohio EPA, 1992).
categories: MWH-A, non-acidic mine runoff affected habitats; MWH-C,
channel modified habitats; and MWH-I, extensively impounded habitats.
Biological criteria were derived from a separate set of modified reference
sites. The biocriteria were set separately for each of three categories of
habitat impact (Fig. 7-3b), The MWH-C and MWH-I subcategory biocrite-
ria were also derived separately for the HELP ecoregion. The MWH-A ap-
plies only within the Western Allegheny Plateau (WAP) ecoregion.
Costs for State Programs Developing
Bioassessments and Biocriteria
Biocriteria programs begin with the development of a bioassessment
framework. Expertise in ecological principles and resource investment by
the agency is required to develop this framework and to implement
biocriteria. State agencies will vary in their investment of resources and ef-
fort in this process.
124
-------
CHAPTER 7:
Biocriteria Development and Implementation
Fish — Boat Sites
( IBt/lwb)
20/5.7
p24/5.s|$
2A/55tSSi
HELP:22/5.7
Rest: 30/6.6
24/5,8
•Z't&P
Impounded
Fish
- Headwater Sites
C
miy&&
msmm
Fish — Wading Sites
flBl/lwb)
Mine Affected
22/5.6p®g>£
24/6.2
24/6.2
K*24/6.2E£i
a24/55K>4fji
wm&
24/6.2
Macroinvertebrates
J Huron Erie Lake Plain • HELP
1 Interior Plateau - IP
Eastern-Ontario Lake Plain - EOLP
Western Allegheny Plateau - WAP
Eastern Corn Belt Plains - ECBP
Figure 7-3b.—Biological criteria In the Ohio WQS for the Modified Warmwater Habitat
(MWH) use designation arranged by biological index, site type for fish, modification
type, and ecoregion. Index values in the boxes on each map are the MWH biocriteria
for the channelized modification type that vary by ecoregion as follows: IBI/Mlwb for
Boat Sites (upper left), IBI/Mlwb for Wading Sites (upper right), IBI for Headwaters
Sites (lower left), and the ICI (lower right). The MWH criteria for the impounded modi-
fication type is located in the box Just outside the Boat Sites map. The biocriteria for
the mine-affected modification type Is represented by the circled value located in the
WAP ecoregion on each map (Ohio EPA, 1992).
Several states that have initiated biocriteria programs were polled to
obtain estimates of their cost and resource needs. These cost estimates rep-
resent a range of program elements including assemblage selection (ben-
thic macroinvertebrates and fish) and geographical coverage (statewide or
targeted regions of the state). The following paragraphs briefly charac-
terize each of the state programs included in the poll before extrapolating
cost estimates in terms of funding and personnel.
¦ Delaware. The nontidal streams in Delaware are mostly low-gradient
coastal streams that drain agricultural lands. Delaware Department of
Natural Resources and Environmental Control (DNREC) developed a
modification of the EPA's rapid bioassessment protocols to sample benthic
macroinvertebrate from multihabitats in these streams. Technical issues
addressed in developing their bioassessment included standardized meth-
ods, level of subsampling, taxonomic level (family or genus), and the se-
125
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
lection of appropriate metrics. Samples are collected during a specified in-
dex period that extends from late summer through the fall season. Biosur-
veys done by department biologists include survey planning, collection,
processing, and data analysis. Consultants are used to assist in processing
benthic samples for large projects.
¦ Florida. Florida Department of Environmental Protection (DEP; for-
merly the Department of Environmental Regulation) used a combination
of in-house biologists, scientists from the EPA's Environmental Research
Laboratory in Corvallis, and consultants to develop a statewide stream
bioassessment program based on thorough site regionalization and meth-
ods development projects. Florida DEP samples benthic macroinverte-
brates from multiple stream habitats using a modified RBP method, and
assesses biological condition using a suite of metrics. The sampling sites
are classified into aggregated subecoregions for determination of appro-
priate reference conditions. Currently, the portions of Florida that are not
adequately delineated are south Florida, south of Lake Okeechobee, and
northeastern Florida around Jacksonville. Two index periods are used to
assess biological condition—August through September, and January
through February. Florida DEP biologists collect and process all samples.
Outside consultants are used to analyze the data and develop taxonomic
keys.
¦ Idaho. Both fish and benthic macroinvertebrates are surveyed by Idaho
Department of Environmental Quality (DEQ) as part of Idaho's monitoring
program. Their biological program is a relatively intense part of a multiyear
monitoring effort to assess nonpoint source impacts. Idaho DEQ is now
evaluating their current program and refining their biological methods.
Consultants are used to assist in this process. The field sampling and sam-
ple analysis are conducted by Idaho DEQ regional staff.
¦ Maine. Maine Department of Environmental Protection (DEP) uses
rock-filled baskets as-introduced substrate for macroinvertebrate coloniza-
tion. The statewide program uses aquatic life use designations to establish
reference conditions. Numeric biocriteria have recently been incorporated
in Maine's rules. Analysis is done using a tiered multivariate procedure
that incorporates information from up to 35 metrics. Maine's index period
is in the summer. Virtually all of its bioassessment is accomplished by
Maine DEP biologists.
¦ Nebraska. Both fish and benthic macroinvertebrates are sampled in Ne-
braska by the Department of Environmental Quality (DEQ). A multimetric
approach is used for both assemblages, based on the IBI for fish and EPA's
RBPs for benthos. Reference conditions have been determined for each
ecoregion in Nebraska and a summer index period is used to sample
streams. Nebraska's biological monitoring program was developed and is
maintained by DEQ biologists.
¦ North Carolina. The Department of Environment, Health, and Natural
Resources (DEHNR) of North Carolina has had an effective bioassessment
program in place for several years. A standardized macroinvertebrate
sampling procedure is used to sample multiple habitats in North Carolina
streams; metrics are used to assess biological condition, and judgment cri-
teria are based on the ecoregion level of site classification. The design and
126
-------
CHAPTER 7;
Biocriteria Development and Implementation
development of the program as well as all aspects of monitoring are con-
ducted by DEHNR biologists.
¦ Ohio. Ohio EPA has developed both a fish and benthic macroinverte-
brate protocol for conducting bioassessments in Ohio's streams and rivers.
A multimetric approach is used in both protocols that focuses on a sum-
mer index period. Site classification is by ecoregion with a given percent-
age of the sites monitored on an annual basis. Numeric biocriteria are
included in Ohio's water resource program. They were developed in a hi-
erarchical manner by aquatic life use and ecoregion. Ohio EPA staff de-
signed and developed the bioassessment program, and conducts the
annual sampling with in-house staff and summer interns.
¦ Oklahoma. The Oklahoma Conservation Commission (OCC) has devel-
oped a biological assessment program that includes benthic macroinverte-
brate, fish, and periphyton sampling to evaluate nonpoint source effects.
However, the benthic program is central and reflects the cost of develop-
ing the program which is statewide and loosely based on ecoregions. The
index period is summer, and monitoring during other seasons is depend-
ent on the case study. Technical consultants were used to help establish the
reference condition.
¦ Oregon. Oregon Department of Environmental Quality (DEQ) has de-
veloped a modified RBP approach for surveying benthic macroinverte-
brates and fish in streams in the Coastal Range. The other five ecoregions
have not been extensively sampled to date. Multiple metrics are calculated
and used to assess biological condition. A single fall index period (Septem-
ber, October, November) is emphasized. However, monitoring is done in
other seasons to evaluate specific impacts, for example, forest insecticide
application. The majority of the biosurvey and assessment is done by DEQ
biologists.
Turning now to costs: it is apparent from the states polled that a mini-
mum of two full-time equivalent staff are needed for the development of
an effective biological assessment program. The states of Ohio, Maine,
North Carolina, and Florida have invested the equivalent of 12 staff (or
more) to develop their programs (Table 7-4). However, Ohio EPA points
out that only 19 percent of their surface water monitoring program is de-
voted to biological monitoring (Yoder and Rankin, 1994). When consid-
ered on the basis of agencywide water programs, Ohio EPA allocates 6
percent to biological monitoring.
Cost investment will vary depending on the geographical coverage
(number of stream miles), the extent of coverage, biological approach and
targeted assemblages, and the extent of shared resources (e.g., other state
and federal agency assistance, and shared reference conditions). Nebraska
and Ohio have developed their program statewide for fish and benthos,
whereas other states polled emphasized only benthos and some have not
covered the whole state (Table 7-5). Although Delaware and Florida have
only partial coverage to date, their programs are relatively complete and
are pertinent for the majority of their state streams. A few of the states
have used contractor support, which ranged from $10,000 to $350,000.
Though self-reported, the costs reviewed here are typical costs in-
curred by state bioassessment programs.
127
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rive'rs
Value of Biocriteria in Assessing Impairment
Water resource agencies currently use several tools to assess impairment
and monitor changes. However, these tools can be separated into three
distinct categories: chemical analysis of water samples, toxicity testing of
selected species, and biosurveys. These tools, though not interchangeable
in all cases, are most effective when used in conjunction with each other.
Chemical and toxicity criteria, however, are only useful for assessing ad-
verse impacts from chemical discharges. Biosurveys and biocriteria are
more appropriate than other tools for measuring cumulative or synergistic
impacts, the status of the resources, and impairment from stressors other
than chemical contamination, such as habitat degradation.
Table 7-4.— The Investment of state water resource agency staff needed to develop bloassessment programs
as a framework for biocriteria.
FULL-TIME EQUIVALENT (FTE) STAFF
STATES
STANDARDIZE
METHODS
SITE
CLASSIFICATION
FIELD
SURVEY
REFERENCE
CONDITION
METRICS AND
INDICES
DEVELOPMENT
TOTAL
Benthos and Fish
[Statewide]
Nebraska
Ohio
0.04
2.0
0.73
1.0
0.88
2.7
0.28
2.5
0.49
3.0
2.4
11.2
Benthos
[Statewide]
Maine
N. Carolina
Oklahoma
1.0
8.0
0.05
8.0
1.0
0.5
1.5
4.0
0.25
2.0
0.75
3.0
1.0
0.25
13.5
16.0
1.8
Benthos
[Partial Coverage]
Delaware
Florida
Oregon
0.4
2.6
0.25
0.1
2.0
0.25
0.3
5.7
1.0
0.6
1.0
1.0
0.6
1.0
0.5
2.0
12.3
3.0
Table 7-5.— Costs associated with retaining consultants to develop bioassessment programs as a framework
for biocriteria. Dash Indicates work done by state employees or information not available; FTE costs for
contractors and state employees are not equivalent.
FULL-TIME EQUIVALENT (FTE) STAFF
STATES
STANDARDIZE
METHODS
SITE
CLASSIFICATION
FIELD
SURVEY
REFERENCE
CONDITION
METRICS AND
INDICES
DEVELOPMENT
TOTAL
Benthos and Fish
[Statewide]
Nebraska
Ohio
—
—
—
—
—
—
Benthos
[Statewide]
Maine
N. Carolina
Oklahoma
—
8
36
25
13
57
25
Benthos
[Partial Coverage]
Delaware
Florida
Oregon
55
100
5
210
10
75
40
75
100
350
10
128
-------
CHAPTER 7:
. Biocriteria Development and Implementation
Several comparison studies were conducted and documented in the
Technical Support Document for Water Quality-based Toxics Control (U.S.
Environ. Prot. Agency, 1991). These studies used biosurvey results to cali-
brate the judgment of impairment using toxicity testing.
The Agency conducted studies at eight freshwater sites in which ambi-
ent toxicity was compared to the biological impact on the receiving water.
These site studies were a part of the Complex Effluent Toxicity Testing
Program (CETTP). Testing was performed on-site concurrent with the field
surveys. Sites exhibiting biological impacts were included from Okla-
homa, Alabama, Maryland, West Virginia, Ohio, and Connecticut. Organ-
isms were exposed to samples of water from various stations and tested
for toxicity. Biological surveys (quantitative field sampling of fish, inverte-
brate, zooplankton, and periphyton communities in the receiving water
areas upstream and downstream of the discharge points) were made at
these stations at the same time the toxicity was tested to see how well the
measured toxicity correlated to the health of the community. These studies
have been reviewed and published in an EPA publication series (Mount et
al. 1984; 1985; 1986; 1986a; 1986b; Mount and Norberg-King 1985; 1986;
Norberg-King and Mount 1986).
A robust canonical correlation analysis was performed to determine
whether or not statistically significant relationships existed between the
ambient toxicity tests and in-stream biological response variables and to
identify which variables play an important role in that relationship (Dick-
son et al. 1992). Influential variables were then used to classify stations as
either impacted or not. Ceriodaphnia dubia productivity and/or Pimephales
promelas weight were used as the basis for predicting impact (U.S. Environ.
Prot. Agency, 1991). Fish richness was used to classify streams as impact
observed or impact not observed.
In this set of studies, agreement was obtained between the prediction
of in-stream toxicity using ambient toxicity testing and the observed bio-
logical impairment from the biosurvey results (Fig. 7-4). However, at 10
percent of the sampling stations, agreement was not reached. EPA (1991)
has said that this small difference in results would not significantly affect
the diagnosis of impairment.
Another study conducted by the North Carolina Division of Environ-
mental Management indicated the high accuracy of predicting receiving
water impacts from whole effluent toxicity tests. Forty-three comparisons
were made between freshwater flowing streams using the Ceriodaphnia du-
86.2%
Inst ream toxicity predicted.
Impairment observed.
Instream toxicity riot predicted,
impairment observed.
Instream toxicity predicted.
Impairment not observed.
Instream toxicity not predicted.
Impairment not observed.
Figure 7-4.—Comparison of ambient toxicity and fish richness surveys at eight sites
In various parts of the United States (taken from U.S. EPA, 1991).
129
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
65.0%
5.0%
7.0%
23.0%
Inst ream toxicity predicted.
Impairment observed.
Instream toxicity oat predicted.
Impairment observed.
Instream toxicity predicted.
Impairment nol observed
Instream toxicity no! predicted.
Impairment not observed.
Figure 7-5.—Comparison of effluent toxicity of receiving water impact using Carlo-
daphnla dubla chronic toxicity tests and freshwater receiving stream benthlc inverte-
brates at 43 point source discharging sites In North Carolina (taken from U.S. EPA,
1991),
48.1%
36.4%
Chemical criteria exceedances.
Biological Impairment observed.
No chemical criteria exceedances.
Biological impairment observed.
Chemical criteria exceedances.
Nq biological impairment.
No chemical criteria exceedances.
No biological Impairment.
Figure 7-6.—Comparison of chemical criteria exceedances and biosurvey results at
64S stream segments in Ohio.
bia chronic test and a qualitative macroinvertebrate sampling. The result
was an overall 88 percent accuracy of prediction (Fig. 7-5). However, in 12
percent of the cases, agreement was not reached. Both of these studies in-
dicate that some risk of error exists if impairment is predicted using toxic-
ity tests alone.
Chemical analyses are less accurate in predicting biological impair-
ment. In a study conducted by Ohio EPA, the prediction of impairment
from chemical analyses agreed with the biological survey results in only
47 percent of the cases (Fig. 7-6). Chemical analyses were unable to detect
the impairment measured by biocriteria at 50 percent of the sites. Ohio
EPA (1990) stated that the absence of detected chemical criteria ex-
ceedances when biological criteria impairment was indicated may result
from several possibilities: (1) chemical parameters other than those sam-
pled have been exceeded, (2) impairments of a nontoxic nature exist, (3)
impairments stemming from physical impacts (e.g., habitat modification,
flow alteration) exist, and/or (4) impairments related to biological interac-
tions (e.g., exotics, disease) exist. None of these scenarios would be de-
tected or fully understood using chemical criteria assessments alone.
The Delaware Department of Natural Resources and Environmental
Control assessed the attainment of their aquatic life use class for nontidal
streams in 1994 using both their dissolved oxygen criteria and a biological
endpoint. Results indicated that the use of the dissolved oxygen criteria
130
-------
• . CHAPTER 7:
Blocriteria Development and Implementation
was inadequate to detect impairment to the aquatic life. Documentation of
exceedances to the dissolved oxygen criteria suggested that only 9 percent
of Delaware's nontidal streams failed to meet attainment (Fig. 7-7).
Whereas the habitat and biological assessment approach indicated that 78
percent of the nontidal streams were not attaining their designated use.
These experiences support the observation that biological criteria are
ah excellent assessment tool and one that covers environmental variables
not necessarily addressed by other chemical, physical, or effluent toxicity
studies. While not yet advocated as a method for setting regulatory
NPDES permit limits, the biocriteria process is clearly an essential means
of environmental assessment and has in fact been used to review these
permits and other management efforts in several states including Ohio, .
Maine, and North Carolina.
No
Fixed Stations - Dissolved Oxygen
(No statistical confidence)
91.0%
Yes
Probabilistic - Habitat/Biology
(95% Confidence Interval +//- 5-6%)
Figure 7-7.—Assessment of nontidal stream aquatic life use attainment In Delaware,
(taken from the state's 305[b] report,1994).
131
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Suggested Readings
Berthouex, P.M. and I. Hau. 1991. Difficulties related to using extreme percentiles for
water quality regulations. Res. J. Water Pollut. Control Fed. 63(6):873-79.
Dickson, K.L., W.T. Waller, J.H.. Kennedy, and L.P. Ammann. 1992. Assessing the rela-
tionship between ambient toxicity and instream biological response. Environ. Toxi-
col. Chem. 11:1307-22.
Fausch, K.D., J.R. Karr, and P.R. Yant. 1984. Regional application of an index of biotic in-
tegrity based on stream fish communities. Trans. Am. Fish. Soc. 113:39-55.
Gibson, G.R. 1992. Procedures for Initiating Narrative Biological Criteria. EPA-822-8-92-
002. Off. Sci. Technol. U.S. Environ. Prot. Agency, Washington, DC.
Karr, J.R. et al. 1986. Assessing Biological Integrity in Running Waters: A Method and Its
Rationale. Spec. Publ. 5. Illinois Nat. History Surv., Champaign, IL.
Mount, D.I., and T.J. Norberg-King, editors. 1985. Validity of Effluent and Ambient Tox-
icity Tests for Predicting Biological Impact, Scippo Creek, Circleville, Ohio.
EPA/600/3-85/044. U.S. Environ. Prot. Agency Washington, DC.
. 1986. Validity of Effluent and Ambient Toxicity Tests for Predicting Biological
Impact, Kanawha River, Charleston, West Virginia. EPA/600/3-86/006. U.S. Envi-
ron. Prot. Agency, Washington, DC.
Mount, D.I., T.J. Norberg-King, and A.E. Steen, editors. 1986. Validity of Effluent and
Ambient Toxicity Tests for Predicting Biological Impact, Naugatuck River, Water-
bury, Connecticut. EPA/600/8-86/005. U.S. Environ. Prot. Agency, Washington,
DC. '
Mount, D.I., A.E. Steen, and T.J. Norberg-King, editors. 1986. Validity of Effluent and
Ambient Toxicity Tests for Predicting Biological Impact, Back River, Baltimore Har-
bor, Maryland. EPA/600/8-86/001. U.S. Environ. Prot. Agency, Washington, DC.
Mount, D.I., N. Thomas, M. Barbour, T. Norberg, T. Roush, and R. Brandes. 1984. Efflu-
ent and Ambient Toxicity Testing and Instream Community Response on the
Ottawa River, Lima, Ohio. EPA/600/8-84/080. U.S. Environ. Prot. Agency, Permits
Div. and Off. Res. Dev., Duluth, MN.
Mount, D.I. et al., editors. 1985. Validity of Effluent and Ambient Toxicity Tests for Pre-
dicting Biological Impact, File Mile Creek, Birmingham, Alabama. EPA/600/8-
85/015. U.S. Environ. Prot. Agency, Washington, DC.
Norberg-King, T.J., and D.I. Mount, editors. 1986. Validity of Effluent and Ambient Tox-
icity Tests for Predicting Biological Impact, Skeleton Creek, Enid, Oklahoma.
EPA/600/8-86/002. U.S. Environ. Prot. Agency, Washington, DC.
Paulsen, S.G. et al. 1991. EMAP-Surface Water Monitoring and Assessment Pro-
gram—Fiscal Year 1991. Off. Res. Dev., U.S. Environ. Prot. Agency, Washington, DC.
Reckhow, K. In Press. Biological Criteria: Technical Guidance for Survey Design and
Statistical Evaluation of Biosurvey Data. Off. Sci. Technol. and Off. Res. Dev., U.S.
Environ. Prot. Agency, Washington, DC.
132
-------
CHAPTER 8.
Applications of the
Biocriteria Process
Biocriteria, a critical tool for state agencies to use in protecting the
quality of water resources, serve several important purposes: they
help (1) characterize and classify aquatic resources, (2) refine aquatic life
use categories, and (3) judge use impairment (1,6., they help determine at-
tainment and nonattainment of designated uses). Additionally, biocriteria
are used for (4) identifying possible sources of impairment (e.g., habitat
degradation, flow regime changes, chemical contamination, energy altera-
tions, or biological imbalance); (5) problem screening; (6) ranking and es-
tablishing priorities for needed remedial actions; and (7) assessing the
results of new management practices. Other applications of the process in-
clude evaluating the adequacy of NPDES permits, and trend reporting for
305(b) reports.
Stream Characterization and Classification
The process of biocriteria development requires that streams be classified
according to type to determine which reference conditions and criteria are
required. This classification must be done in each of the nation's eco-
regions — as defined by climate, geographic, and geologic characteristics.
Then, within these regions, the streams should be further categorized and
their classes either combined or subdivided depending on whether they
have similar or distinctive biotic compositions.
Initial classifications can be confirmed, refined, or revised on the basis
of subsequent biological data. This continued monitoring makes the refer-
ence sites and derived biological criteria more certain, and helps the re-
source managers and biologists identify unique or particularly sensitive
streams for special attention or protection. The following case study from
North Carolina illustrates this point.
CASE STUDY — North Carolina
STATE
LOCATION
DATES
North Carolina
South Fork of New River
March-August 1990
The South Fork of New River forms the headwaters of the New River in
North Carolina. The entire South Fork New River catchment is mountain-
ous with generally steep, forested slopes. The floodplain is broad with
Purpose: .
To illustrate the
importance of
biocriteria in various
areas of water
resource
management.
133
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
The classification
and definition of
designated uses of
streams and rivers are
important in
developing and using
biocriteria. Similarly,
as biocriteria become
established, the
expanded database
helps refine these
classifications.
rolling hills; and land uses in the area are primarily rural and agricultural,
including crop and dairy pasture production. Nonpoint source runoff
from these uses has a high potential for water quality problems (NC Dep.
Environ. Manage. 1978).
The North Carolina Environmental Management Commission classi-
fies certain waters of the state as "outstanding resource waters" (ORW) if
such waters have an exceptional recreational significance and exceptional
water quality. Determining whether a North Carolina stream qualifies for
reclassification as an ORW depends primarily on data collected by the Bio-
logical Assessment Group, which is part of North Carolina's biocriteria
program.
To evaluate an ORW request for the New River, the Biological Assess-
ment Group collected benthic macroinvertebrate samples from 21 riverine
and tributary locations within the New River catchment. Main-stem river
locations (the South and North Forks of the New River) were sampled us-
ing the Group's standardized qualitative collection method, which.uses a
wide variety of collection techniques (and 10 samples) to inventory the
aquatic fauna. The primary output is a taxa list with some indication of
relative abundance for each taxon (i.e., abundant, common, or rare). The
combined number of species in the pollution-intolerant insect orders of
Ephemeroptera, Plecoptera, and Trichoptera (EPT Index) is used with de-
partment criteria to assign water quality ratings. Unimpaired or minimally
impaired streams and rivers have many species, while polluted areas have
fewer species.
Based on analyses of the biological data (Fig. 8-1), excellent water
quality was found at the ambient monitoring location on the South Fork
New River near Scottsville and Old Field Creek, a tributary of the South
Fork New River. Prior data have also consistently shown excellent water
SeottsKllto (S.Fk. New Rlur)
Amelia (S.Fk. New River)
"Seasonal adjustment factor for winter and spring developed for EPT Index after 1990
Figure 8-1.—EPT Index (number of taxa of Ephemeroptera, Plecoptera, and Trichop-
tera) for two locations on the South Fork of the New River, North Carolina.
134
-------
CHAPTER 8:
. Applications of Biocriteria
quality at the South Fork New River near Jefferson and for the New River
itself, below the confluence of the North and South Forks. A site on the
North Fork New River also had excellent water quality, but repeated sam-
pling at this site revealed that its samples fluctuate between good and ex-
cellent quality on a temporal basis. Until it achieves a more consistent
water quality rating, this site on the North Fork will not be recommended
for an ORW classification.
Old Field Creek has an outstanding brook trout resource. The South
Fork of the New River has been designated as a Natural and Scenic River
from the confluence of Dog Creek in the documented excellent reach of
the river to its confluence with the New River. The New River — accord-
ing to information provided by local canoeing outfitters — supports an
unusually high level of water-based recreation.
It was, therefore, recommended that the South Fork New River from
the confluence of Dog Creek to the New River, and the New River itself, to
the last point at which it crosses the North Carolina-Virginia state line be
designated ORW. The west prong of Old Field Creek (Call Creek) from its
source to Old Field Creek, and Old Field Creek below its confluence with
the west prong to the South Fork New River was also designated ORW.
On the basis of biological data, the recommendation was accepted. The
Commission reclassified these streams in December 1992, thereby ensur-
ing that stricter point and nonpoint source regulations would be enforced
in this region.
Refining Aquatic Life Uses
As a biocriteria program grows, the accumulated information helps state
or tribal biologists refine the aquatic life use categories initially developed.
That is, the additional information about the distribution and status of bi-
ota helps resource managers refine their categories of aquatic life use. The
development of the "outstanding resource waters" category in North
Carolina is an illustration of this process in which a less natural and di-
verse community characterizes the aquatic life use. Information obtained
through biological surveys is used to explicitly characterize each aquatic
life use. Other examples follow.
Oregon is presently developing state surface water categories based
on aquatic life classifications. The proposed language for biological criteria
in Oregon separates water resources into two categories. The first classifi-
cation ("Outstanding Resource Waters") is for waters that shall be man-
aged so that "resident biological communities . ... remain as they naturally
occur and all indigenous aquatic species are protected and preserved."
The second category is for all other waters of Oregon. Waters in this
class meet their use requirement if and when the following statement is
applicable: "other waters of the state, including waters outside designated
mixing zones, shall be of sufficient quality to support aquatic species with-
out detrimental changes in the resident biological communities" (Oregon
Dep. Environ. Qual. 1991).
Maine has establisl^ed four classes of water quality for streams and
rivers (Table 7-2). The "high quality waters" of Maine are separated into
two categories: one caitegory contains waters meeting the highest goal of
135
' a
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Wisconsin
19%
Nebraska
68%
Ohio
61%
Vermont
23% '
Maine
4%
Percent of
streams that do
not sustain fish
and aquatic
insect life.
Connecticut
54%
New Jersey
61%
Kansas
96%
Delaware
87%
Missouri
52%
Florida
35%
Kentucky
22% .
Source: State 305(b) Reports. 1992-1994
Figure 8-2.—Examples from some states using biological assessments to determine
aquatic life use support in rivers and streams. Failure to sustain fish and aquatic life
is defined with respect to the reference condition in that state.
The blocriteria
process is a
fundamental tool for
assessing aquatic life
use impairment. 1 •
the Water Quality Act (no discharge, Class AA); the other contains waters
of high integrity but minimally impaired by human activity (Class A).
"Good quality water" is assigned to the second category: Class B. Waters
in Class B meet their aquatic life use requirement if and when all indige-
nous aquatic species are supported and only nondetrimental changes in
community composition occur. The fourth category Class C, is reserved for
the lowest quality waters. Waters in this class also meet their use require-
ment if and when all indigenous aquatic species are supported. However,
changes in species composition may occur in Class C waters, even though
the structure and function of the aquatic community must be maintained
(Davies et al. 1991).
These classifications and their refinement depend on a well-estab-
lished biocriteria program supported by regular, representative biosur-
veys. In fact, the procedure has been so successful that some states are
' shifting from only chemical sampling to an emphasis on biological moni-
toring for their 305(b) assessments. In their water quality assessment re-
ports to Congress in 1992 and 1994, several states used biological
assessments to determine the extent of attainment or nonattainment of the
aquatic life use designations for their streams (Fig. 8-2). These data should
not be used for comparing one state to another because the data — and
hence the figures listed in Figure 8-2 — refer to assessed waters only, not
to all waters in a given state.
Judging Use Impairment
A key element of water resource management under the Clean Water Act
is the establishment and enforcement of standards to protect the nation's
surface waters. If these state-developed standards are not met, legal action
may be taken against dischargers to protect or restore the water resource.
Criteria are scientifically based benchmarks upon which the standards are
based, and biological criteria are benchmarks arrived at from direct meas-
136
-------
CHAPTER B:
Applications of Biocriteria
urements of the responses of resident fish and other organisms to condi-
tions in the water. Chemical, physical, and whole effluent criteria are indi-
rect or surrogate measurements of degradation based on the amount of
pollutant present in the waters, not the actual condition of the biota.
Biocriteria are designed to reflect the designated use of the water re-
source selected by the state so failure to meet these criteria is a violation of
the standards derived from them. Thus, the biocriteria process is a funda-
mental tool for directly assessing aquatic life use impairment.
In Ohio, use attainment or nonattainment is determined using biocrit-
eria based on both macroinvertebrates and fish. Full use attainment occurs
if all criteria are met. Partial use attainment occurs if one assemblage
meets its criteria though the other does not. The status is nonattainment if
none of the biocriteria are met, or if one assemblage indicates poor or very
poor performance, even though the other indicates attainment.
CASE STUDY —Ohio
STATE
LOCATION
DATES
Ohio
Upper Hocking River
. 1982-1991
The Hocking River basin covers 1,197 square miles in southeast Ohio, and
flows through the cities of Lancaster, Logan, Nelsonville, and Athens; each
city maintains wastewater treatment facilities (WWTPs) that discharge
into the river (Clayton Environmental Consultant, 1992). Historically, the
upper Hocking River near Lancaster has been one of the most severely de-
graded river segments in the state (Ohio Environ. Prot. Agency, 1982).
Throughout the 1970s and early 1980s, the river was severely impacted by
industrial effluent, combined sewer overflows (CSOs) and inadequate
treatment at the Lancaster WWTP (Ohio Environ. Prot. Agency, 1985). The
severe chemical impacts — low dissolved oxygen, and high levels of am-
monia, lead, cyanide, cadmium, and phenolics — resulted in gross organic
enrichment, heavy metal contamination, significant'levels of in-stream
toxicity, and periodic fish kills. Invertebrate studies of this portion of the
river revealed a severely degraded biological condition with little down-
stream recovery (Fig. 8-3).
Consequently, the city of Lancaster began upgrading its WWTP in
1986 and reached full operation in 1989. The upgrades, sewer rehabilita^
tion, elimination of bypasses, and the addition of a pretreatment program
to remove metals, substantially improved both the water quality and the"
resident aquatic communities.
The Upper Hocking River has since exhibited the greatest improve-
ment in biological performance of any river system in the state, although
its recovery is not yet complete. In 1982, the biological communities down-
stream of the Lancaster WWTP and CSOs reflected the grossly polluted
and acutely toxic conditions. None of the 20.5 miles from Lancaster to
Logan attained their WWH standard, and 75 percent of them were in poor
or very poor condition. In 1990, only 8.7 miles were still in the nonattain-
ment category, while the rest achieved partial or full attainment and the
average ICI score for that portion of the river rose from 6.9 to 42, a seven-
fold improvement in the invertebrate community index (ICI).
Macroinvertebrate community performance (as measured by the ICI)
improved dramatically, largely in response to the improved water quality.
The fish community has substantially improved as well, although serious
Biocriteria establish
conditions based on
attributes of the
resident biota which
protect the level of
aquatic life
designated for the
\ water resource by a
state or tribe. Failure
to meet the biocriteria
is evidence of an
impaired water
resource.
-------
BIOLOGICAL CRITERIA;
Technical Guidance for Streams and Small Rivers
60
50
— 40
& 30
20
10
0
100 95' 90 85 80 75 70 65
River Mile
Source: Ohio EPA
Figure 8-3.—Temporal trends in the improvement of the Upper Hocking River, 1982 -
1990 (adapted from Ohio EPA).
habitat alterations (e.g., channelization, bank erosion, and siltation) con-
tinue to inhibit silt-sensitive species. As seen in Figure 8-3, the biocriteria
process with its well-defined criterion, careful surveys, and documented
biotic indices clearly reveals not only impairment, but management re-
sponse efforts and the magnitude of the subsequent recovery.
Diagnosing Impairment Causes
An underlying theme of biosurveys and biocriteria is to demonstrate the
type and extent of impairment at the sites being evaluated so that proper
management can be initiated. This demonstration can be done by compar-
ing the attributes of aquatic communities at these sites with those found at
sites that are unimpaired or minimally impaired. All human-induced al-
terations affect biological integrity simply by impacting the five environ-
mental factors that affect and determine water resource quality. As
discussed in chapter 5, the environmental factors of importance to the
stream biota are the site's
• energy base
• chemical constituents
• habitat structure
• flow regime, and
• biotic interactions.
These factors not only influence the aquatic biota; they also affect other
elements and processes that normally occur along the stream or river gra-
dient.
Their identification provides an important indicator of the type, locale,
and extent of remedial or protective management efforts that should be
Lancaster Area
Urban Sources
: Exceptional
. Good
Ecoregioriai
Biocriterion
(ICI=34)
: Fair
1990
1982
: Poor
An underlying theme
of biosurveys and
biocriteria is to
demonstrate the type
and extent of
impairment at study
sites so proper
management can be
initiated.
138
-------
CHAPTER. 8:
Applications of Biocriteria
taken. For example, anthropogenic impairment may result from nutrient
runoff of fertilizers; improper use or disposal of chemical toxins; conver-
sion to cropland or other land use modifications; flow alterations; or over-
fishing. The evaluation of biological and habitat data collected in the
biosurvey-biocriteria process can help reveal these causative elements. For
example, the biological data will suggest whether overfishing or stocking
are factors, or whether disease (which is not strictly anthropogenic) may
also be a contributing factor. The habitat data will divulge any structural
or sedimentation rate changes, and attendant or subsequent water quality
tests will further define toxic or other problems of chemical origin.
An example in West Virginia involved stream degradation resulting
from sewage, mining, and urbanization (Leonard and Orth, 1986). Here
fish assemblage measurements were indexed in a "cultural pollution in-
dex" or CPI (derived from the IBI) to assess watershed and stream quality
based on the assumption that assemblage features change consistently
with stream degradation. Some fish community attributes respond more
quickly than others to stream degradation (Angermeier and Karr, 1986;
Karr et al. 1986). However, each metric of the index is sensitive within a
different range of stream degradation. In these small coolwater streams of
West Virginia, the CPI was sufficiently broad to rank the degree of degra-
dation variously caused by mining, sewage, and urbanization. This study
indicates that biotic indexes and criteria can be developed to reflect both
the characteristics of regional fish populations and the particular forms of
pollution or disruption they encounter.
CASE STUDY —Delaware
STATE LOCATION DATES
. Delaware Statewide 1991-1994 _
In 1994, the Delaware Department of Natural Resources and Environ-
mental Control (DNREC) completed an assessment of the physical habitat
conditions of nontidal streams throughout the state. Based on a sampling
of 189 sites, only 13 percent were found to be in "good" condition while 87
percent were found to be in either "fair" or "poor" condition. "Good" con-
ditions were defined as comparable to reference conditions. These results
have a 95 percent confidence interval of plus or minus 6 to 8 percent. Re-
sults were also reported separately for each of the three Delaware counties
and for the Piedmont and Coastal Plain ecoregions. The impairment in the
Piedmont ecoregion was caused by urbanization and stormwater while
the impairment in the Coastal Plain was caused by agriculture and chan-
nelization. This assessment is published as Appendix D of the state's 1994
305(b) report.
This information builds on biological data collected at the sites in the
Coastal Plain in 1991 and published in the state's 1992 305(b) report. This
report concluded that 72 percent of the nontidal streams in Kent and Sus-
sex Counties (Coastal Plain ecoregion) had "good" macroinvertebrate
communities compared to 28 percent that were determined to be in "fair"
or "poor" condition. Further analysis has shown that degraded physical
habitat was the principle cause of the biological impairment; 81 percent of
the sites with "poor" biology had "poor" physical habitat (Fig. 8-4). Fur-
ther water quality studies have implicated the loss of shade and its effects
on dissolved oxygen and temperature as key factors that contributed to
Human-induced
alterations may occur
as chemical
contamination (point
or nonpoint) or as a
variety of other effects
'such as flow
alteration or habitat
modification.
139
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Fair 4).o*
POOr 28.0%
Habitat
81%
Other
19%
Good 3i.o%
Biological Quality
Stressor Evaluation
(Margin of error +/- 6-8%; 95% confidence
Source: Delaware, 1992
Figure 8-4,—Assessment summary, Kent and Sussex Counties, Delaware, 1991.
Yes
87%
Fixed Stations - Dissolved Oxygen
(no statistical confidence)
(not resource based)
Probabilistic - Habitat/Biology
(95% confidence interval +/- 5-6%)
(resource based)
Source: Delaware, 1994
Figure 8-5.—State of Delaware 1994 305(b) report, aquatic life use attainment —
all nontldal streams.
the biological impairment. A statewide survey of the biological condition
of nontidal streams is currently under development.
Prior to the use of biological and physical habitat measures, Delaware
used dissolved oxygen (DO) to judge attainment or nonattainment of
aquatic life uses. In the 1994 305(b) report, the state reported that 13 per-
cent of its streams were not attaining aquatic life uses based on DO data.
However, 87 percent were found to be impaired based on biological and
physical habitat measures (Fig. 8-5). The lower estimate of impairment us-
ing DO results from (1) sampling during the day when DO levels are the
highest, (2) disproportionate sampling of larger streams with better habi-
tat and more assimilative capacity than smaller streams, and (3) a focus on
point sources many of which are meeting permit limitations. The higher
estimate of impairment using biological criteria and supporting biological
community measurements helped reveal a cause of degradation that
might not have been identified by other methods. It reflects the impact of
nonpoint source activities, primarily urbanization (stormwater) and agri-
culture, on the state's nontidal streams.
140
-------
CHAPTERS:
Applications of Biocriteria
Problem Identification
Monitoring the status and condition of resident communities over time is
important to assess trends in the quality of the biota, whether to guard
against further degradation or to measure improvement. In the course of
such routine monitoring, new problems or conditions are often discov-
ered, In fact, the Florida Department of Environmental Regulation has a
specific (unpublished) program underway to determine the environ-
mental damage (or lack thereof) caused by all significant point source dis-
charges in the state. When the Florida DER began permitting point source
discharges, staff relied mainly on compliance with numerical chemical
standards. Over time, the need to evaluate the effects of these discharges
on receiving waters has increased, both to ensure adequate environmental
protection and to set priorities for enforcement or remedial action. Empha-
sis will be placed on detecting losses of biotic integrity through measures
of imbalance in the flora and fauna, effects of toxic materials, dominance
of nuisance species, and high populations of microbiological indicators.
A two-tiered approach is being used in the Florida program to detect
environmental disturbances in receiving waters. Preliminary investiga-
tions (screening phase) involve qualitative sampling and analysis of ben-
thic macroinvertebrate assemblages. A reference or background station is
established for comparison with an area downstream of a discharge. Using
the results of this relatively low intensity investigation, site impairment is
ranked from "no" to "moderate" to "severe." If necessary, subsequent
studies on dischargers (definitive phase) will use a more quantitative,
multiparameter sampling regime. According to the Florida Department of
Environmental Regulation, study parameters (such as macroinvertebrates,
periphyton, macrophytes, bacteria, bioassays, sediment analysis, and
physical and chemical analyses) are well suited for detection of violations.
The Arkansas Department of Pollution Control and Ecology addresses
screening level monitoring using rapid bioassessment at paired stations
that bracket pollutant sources for impact identification. As was shown in
Figure 5-2, the initial rapid bioassessment screening may result in the ap-
plication of other biological and chemical methods, after which an on-site
decision can be made for subsequent action. In situations where "no im-
pairment" or "minimal impairment" classifications• are met, field efforts
are discontinued until further information indicates a problem. Streams
classified as "substantially" or "excessively" impaired trigger additional
investigative steps that employ a variety of methods (Shackleford, 1988).
CASE STUDY —Maine
STATE
LOCATION
DATES
• Maine
. Piscataquis River
1984-1990
The Piscataquis River, with a drainage area of about 250 square miles
northwest of Bangor, runs near the town of Guilford (Clayton Environ-
mental Consultants, 1992). For many years, untreated manufacturing
water from a textile mill and untreated domestic sewage from Guilford
significantly impacted the river. In an attempt to improve the quality of
the waterbody, the town of Guilford constructed a publicly owned treat-
ment works (POTW), which was completed in June 1988. The POTW has
aerated lagoons (detention time of 50 days) and a flow of 0.75 million gal-
Monitoring the status
and condition of
resident communities
over time is important
to assess trends in
the quality of the
biota, whether to
guard against further
degradation or to
measure
improvement.
141
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Ions per day (mgd). Seventy-five percent of the total inflow into the plant
comes from textile mill waste; the remaining 25 percent from domestic
sewage.
Maine's water quality standards designate a specific level of biological
integrity that each class of water must maintain. To meet the standards for
a Class A water, the aquatic community must be "as naturally occurs" and
specific definitions are used to identify ecological attributes that may be
tested to determine if the standards are being achieved.
Maine's Department of Environmental Protection uses a multivariate
statistical model to predict the probability of attaining each classification.
The model uses 31 quantitative measures of community structure, includ-
ing the Hilsenhoff Biotic Index, Generic Species Richness, EPT, and EP
values.
Monitoring of the Piscataquis River occurred at sites upstream and
downstream of the textile mill in 1984,1989, and 1990, and at a site down-
stream from the POTW in 1989 and 1990. Before 1988, benthic macroinver-
tebrate samples collected downstream of the mill revealed a severely
degraded community consisting primarily of pollutant tolerant organisms.
The macroinvertebrate samples indicated that the waterbody failed to
meet the lowest aquatic life standards allowed by the state, although
chemical water quality parameters (e.g., biochemical oxygen demand) col-
lected at the site were meeting standards. Chemical parameters alone are
insufficient to detect every water quality impairment.
Following the rerouting of the textile mill waste and the completion of
the POTW in 1988, the river recovered quickly. Monitoring data, collected
during the summer of 1989, revealed a substantially improved macroin-
vertebrate community (Fig. 8-6). Pollution-sensitive organisms were abun-
dant and EPT values had increased from 1 in 1984 to 17 to 20 in 1989 and
1990. The generic richness improved from 6.35 in 1984 to 38 in 1990. The
site now fully supports the aquatic life standards of Class A waters.
Other Applications of the Process
¦ Regulatory Assessments. The biocriteria process is excellent for assess-
ing the adequacy of NPDES permits to accomplish their intended purpose.
As indicated earlier in this text, biological parameters are not recom-
mended as permit limits at this time. But an ideal way to evaluate the suc-
cess of the permit is to compare downstream biota to upstream or regional
reference conditions and biological criteria. If the biota are not sufficiently
protected as indicated by a downstream survey, the permit should be re-
viewed and perhaps revised. This biological review should be scheduled
each time a permit is due for renewal.
¦ Management Planning. This application was implied in several of the
examples used in this chapter. Streams in a particular ecoregion can be
ranked on the basis of their index scores and relative compliance with
biocriteria. The natural resource manager can then assign priorities to in-
dividual streams or groups of streams for protection, further investiga-
tions, or remedial management depending on the availability of personnel
and funding resources. That is, a rational decision with a reasonable ex-
142
-------
CHAPTER 8:
Applications of Biocriteria
60
50..
Site #2 - Below Textile Mill, Above POTW
New POTW operational (June 1988)
Site #3- Below POTW
1984 Data not
available
New POTW operational(June 1988)
INDICES
Generic Richness mm EPT
EP
Source: Maine DEP
Figure 8-6.—Macroinvertebrates in the Piscataquis River, Maine, 1984-1990. New sew-
age treatment plant became operational in June 1988 (arrow).
pectation of results can be used to determine which streams will receive
attention in any given year.
¦ Water Quality Project and Techniques Evaluation. When a. manage-
ment plan is implemented, the changed land use practices, bank erosion
control structures, and effluent diversion or treatment practices applied
can be evaluated for effectiveness by applying the biocriteria process as a
"before," "during," and "after" monitoring scheme. If results are as hoped
for — as they were, for example, in the Maine case study — the manager
can apply the technique to similar problems on other streams. If there is
little or no change in the biota, more work is indicated and the technique
obviously is not ready for application elsewhere.
¦ Status and Trends Documentation. This task is one of the primary
functions of the biocriteria process and should not be overlooked in dis-
cussing other uses of the approach. As an ongoing program, the biosur-
vey-biocriteria process provides perhaps the best, most direct and
comprehensive assessment of water resource condition available to us.
Annual surveys of the biota not only refine the biocriteria, but are the ba-
sis of state and EPA reports to the nation on the status of surface waters
and on our relative success or failure to protect these valuable resources.
14,
-------
BIOLOGICAL CRITERIA.
Technical Guidance for Streams and Small Rivers
Suggested Readings
Davies, S.E, L. Tsomides, D.L. Courtemanch, and F. Drummond. 1991. Biological Moni-
toring and Biocriteria Development. Prog. Sum. Maine Dep. Environ. Prot.,
Augusta, ME.
Idaho Department of Health and Welfare. 1991. State funded 319 project: biological met-
ric development study plan. Pages 15-18 in Rapid Bioassessment Protocol Develop-
ment, Boise, ID.
Leonard, P.M. and D.J. Oxth. 1986. Application and testing of an index of biotic integrity
in small, coolwater streams. Trans. Am. Fish. Soc. 115:401-14.
Montana Department of Health and Environmental Sciences. 1990. Montana's Ap-
proach to Honpoint Assessment and Monitoring. Outline. Water Qual. Bur.,
Helena, MT.
North Carolina Division of Environmental Management. 1978. 208 Phase I Results.
Raleigh, NC.
Ohio Environmental Protection Agency. 1990. The Use of Biocriteria in the Ohio EPA
Surface Water Monitoring and Assessment Program. Columbus, OH.
. 1991. Biological and Water Quality Study of the Hocking River Mainstem and
Selected Tributaries: Fairfield, Hocking, and Athens County, Ohio. Columbus, OH.
Oregon Department of Environmental Quality. 1991. Biological Criteria-Implementation
Plan. Draft. Portland, OR.
Primrose, N.L. 1989. Routine Benthic Biomonitoring Protocol: A Proposal. Maryland
Dep. Environ., Annapolis, MD.
Rankin, E.T. and C.O. Yoder. 1991. Calculation and uses of the area of degradation value
(ADV). In Ohio Water Resource Inventory, Executive Summary and Volume 1. Ohio
Environ. Prot. Agency, Columbus, OH.
v 1992. Summary status, and trends. In Ohio Water Resource Inventory, Volume 1.
Ohio Environ. Prot. Agency Columbus, OH.
Shackleford, B. 1988. Rapid Bioassessment of Lotic Macroinvertebrate Communities:
Biocriteria Development. Arkansas Dep. Pollut. Control Ecol., Little Rock, AR.
U.S. Environmental Protection Agency 1990. Biological Criteria: National Program
Guidance for Surface Waters. EPA-440/5-90-004. Off. Water, Washington, DC.
. 1991c. Technical Support Document for Water Quality-Based Toxics Control.
EPA-505/2-90-001. Off. Water, Washington, DC.
. 1991d, Biological Criteria: Research and Regulation Proceedings of a Sympo-
sium. EPA-440 / 5-91-005. Off. Water, Washington, DC.
Contacts for Case Studies
David Penrose, North Carolina DEM, 919/733-6946
Chris Yoder, Ohio EPA, 614/728-3382
John Maxted, Delaware DNREC, 302/739-4590
David Courtemanch, Maine DEP, 207/287-7889
144
-------
Glossary
The development of water quality standards and criteria requires clear
understanding of key terms and concepts. Foremost is the differentia-
tion between water quality standards and criteria. A standard is a legally
established state regulation consisting of two parts: (a) designated uses
and (b). criteria. A designated use is a classification designated in water
quality standards for each waterbody or segment that defines the optimal
purpose for that waterbody. Examples of designated uses for particular
waterbodies are drinking water use and aquatic life use. Criteria are state-
ments of the conditions presumed to support or protect the designated use
or uses. In practice, if the conditions specified by the criteria are met, the
designated use should be supported.
Biocriteria require additional understanding and a common frame of
reference,for effective development and use in a water quality standards
framework. The following definitions provide this frame of reference, and
should be carefully considered to ensure consistent interpretation of con-
cepts and terminology.
An acceptable/unacceptable threshold is the minimum measured level at
which some condition can be differentiated such that the target loca-
tion is or is not considered reasonable for maintenance of the desig-
nated use. The magnitude of impairment is not addressed with a
threshold determination.
Ambient monitoring is sampling and evaluation of receiving waters not nec-
essarily associated with episodic perturbations.
An aquatic assemblage is an association of interacting populations of organ-
isms in a given waterbody, for example, fish assemblage or a benthic
macroinvertebrate assemblage.
Aquatic biota is the collective term describing the organisms living in or de-
pending on the aquatic environment.
An aquatic community is an association of interacting assemblages in a
given waterbody, the biotic component of an ecosystem (see also
aquatic assemblage).
Assemblage structure is the make-up or composition of the taxonomic
grouping such as fish, algae, or macroinvertebrates relating primarily
to the kinds and number of organisms in the group.
145
-------
BIOLOGICAL CRITERIA
Technical Guidance for Streams and Small Rivers
Autotrophic refers to the trophic status, the balance between production
and consumption where production within the system exceeds respi-
ration.
Autotrophic systems are those systems for which the primary nutrient
source of fixed carbon is intrinsic, such as streams in which there is
abundant growth of algae or macrophytes.
A biogeographic region is any geographical region characterized by a dis-
tinctive flora and/or fauna (see also ecoregion).
A bioindicator is an organism, species, assemblage, or community charac-
teristic of a particular habitat, or indicative of a particular set of envi-
ronmental conditions.
• Biological assessment is an evaluation of the condition of a waterbody using
biological surveys and other direct measurements of the resident biota
in surface waters.
Biological criteria, or biocriteria, are numerical values or narrative expres-
sions that describe the reference biological condition of aquatic com-
munities inhabiting waters of a given designated aquatic life use.
Biocriteria are benchmarks for water resources evaluation and man-
agement decision making.
Biological integrity is functionally defined as the condition of an aquatic
community inhabiting unimpaired waterbodies of a specified habitat
as measured by an evaluation of multiple attributes of the aquatic bi-
ota. Three critical components of biological integrity are that the biota
is (1) the product of the evolutionary process for that locality, or site,
(2) inclusive of a broad range of biological and ecological charac-
teristics such as taxonomic richness and composition, trophic struc-
ture, and (3) is found in the study biogeographic region.
Biological monitoring, or biomonitoring, is the use of a biological entity as a
detector and its response as a measure to determine environmental
conditions. Toxicity tests and ambient biological surveys are common
biomonitoring methods.
A biological response signature is a unique combination of biological attrib-
utes that identify individual impact types or the cumulative impacts of
several human influences.
A biological survey, or biosurvey, consists of collecting, processing, and ana-
lyzing representative portions of a resident biotic community.
A biomarker is any contaminant-induced physiological or biochemical
change in an organism that leads to the formation of an altered struc-
ture (a lesion) in the cells, tissue, or organs of that individual or
change in genetic characteristics.
Channelization is the procedure of deepening and straightening stream or
river channels through dredging. In some states, channelization in-
cludes complete concrete lining of channel bottom, sides, and ease-
ments.
146
-------
CHAPTERS;
Glossary
A community component is any portion of a biological community. The com-
munity component may pertain to the taxonomic group (fish, inverte-
brates, algae), the taxonomic category (phylum, order, family, genus,
species, stock), the feeding strategy (herbivore, omnivore, predator),
or the organizational level (individual, population, assemblage) of a
biological entity within the aquatic community.
A confidence interval is an interval that has the stated probability (e.g., 95
percent) of containing the true value of a fixed (but unknown) pa-
rameter.
Data quality objectives (DQOs) are qualitative and quantitative statements
developed by data users to specify the quality of data needed to sup-
port specific decisions; statements about the level of uncertainty that a
decisionmaker is willing to accept in data used to support a particular
decision. Complete DQOs describe the decision to be made; what data
are required, why they are needed, the calculations in which they will
be used; and time and resource constraints. DQOs are used to design
data collection plans.
Degradation is any alteration of ecosystems such that chemical, physical, or
biological attributes are adversely affected.
Degree days are units used in measuring the duration of a life cycle or
growth stage of an organism; they are calculated as the product of
time and temperature averaged over a specified interval;
A designated use is a classification specified in water quality standards for
each waterbody or segment relating to the level of protection from
perturbation afforded by the regulatory agency.
Diversity is the absolute number of species in an assemblage, community,
or sample; species richness (see also taxa richness).
Ecological assessment is a detailed and comprehensive evaluation of the
status of a water resource system designed to detect degradation and,
if possible, identify the causes of that degradation.
Ecological health is the degree to which the inherent potential of a biological
system is realized, the dynamic equilibrium of system processes is
maintained, and a minimal amount of external support for manage-
ment is needed.
Ecological integrity is the condition of an unimpaired ecosystem as meas-
ured by combined chemical, physical (including habitat), and biologi-
cal attributes.
Ecoregions, or regions of ecological similarity, are defined by similarity of cli-
mate, landform, soil, potential natural vegetation, hydrology, or other
ecologically relevant variables.
Ecoregionalization — See regionalization.
Elements are the richness of items that make up biological systems, meas-
ured as number of kinds.
147
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Generalists are organisms that can utilize a broad range of habitat or food
types.
Heterotrophic input refers to the trophic status, the balance between pro-
duction and consumption where respiration within the system ex-
ceeds production.
Heterotrophic systems are those systems for which the primary nutrient
source of fixed carbon is extrinsic, such as streams for which the main
source of organic input is from riparian vegetation in the form of leaf
litter and woody material.
Historical data are datasets existing from previous studies, which can range
from handwritten field notes to published journal articles.
Hyporheic pertains to saturated sediments beneath or beside streams and
rivers.
An impact is a change in the chemical, physical (including habitat), or bio-
logical quality or condition of a waterbody caused by external sources.
An impairment is a detrimental effect on the biological integrity of a water-
body caused by an impact that prevents attainment of the designated
use.
Level of uncertainty pertains to the confidence, or lack thereof, that data
from an assessment will support the conclusions.
Macroinvertebrates are animals without backbones of a size large enough to
be seen by the unaided eye and which can be retained by a U.S. Stand-
ard No. 30 sieve (28 meshes per inch, 0.595 mm openings).
Macrophytes are large aquatic plants that may be rooted, unrooted, vascu-
lar, or algiform (such as kelp); includes submerged aquatic vegetation,
emergent aquatic vegetation, and floating aquatic vegetation.
A metric is a calculated term or enumeration representing some aspect of
biological assemblage structure, function, or other measurable aspect;
a characteristic of the biota that changes in some predictable way with
increased human influence; combinations of these attributes or metrics
provide valuable synthetic assessments of the status of water re-
sources.
Minimal effluent dilution occurs in low flow conditions in which there is a
lower quantity of water and thus a decreased ability for receiving wa-
ters to lower concentration levels of discharged compounds.
Minimally impaired is a term used to describe sites with slight anthropo-
genic perturbation relative to the overall region of study.
Mutualism is a form of symbiotic relationship in which both organisms
benefit, frequently entailing complete interdependence.
Narrative biocriteria are general statements of attainable or attained condi-
tions of biological integrity and water quality for a given use designa-
tion (see also biocriteria).
14
-------
CHAPTER 9:
Glossary
Nonpoint source is the origin of pollution in diffuse sources such as agricul-
ture, forestry, and urbanization. Such pollution is transported by rainfall
or snowmelt runoff carrying pollutants overland or through the soil.
Numeric biocriteria are numerical indices that describe expected attainable
community attributes for different designated uses (see also biocriteria).
Organic pollution results from the presence of living substances in a stream
or other waterbody at higher than natural background levels because
of anthropogenic activities.
Paleoecological data are records derived from ancient or fossil remains dis-
covered in lake sediments, including, for example, the fossilized re-
mains of diatoms, pollen, seeds, or arthropod exoskeletal fragments.
(Arthropoda are the phylum of invertebrate animals with jointed
limbs, such as crustaceans and spiders.)
Performance effect criteria are judgment criteria that weigh the effectiveness
of a project activity or function; determination of proper functioning.
Periphyton is a broad organismal assemblage composed of attached algae,
bacteria, their secretions, associated detritus, and various species of
microinvertebrates.
Processes (or biotic processes) pertain to ecological and evolutionary activi-
ties that naturally organize and regulate biological systems at all levels
from genetic to landscape; examples are production, food acquisition,
biotic interactions, and recruitment.
Production is the increase in biomass (somatic growth plus reproduction)
of an individual, population, or assemblage.
Point source is the origin of pollutant discharge that is known and specific,
usually thought of as effluent from the end of a pipe.
A population is an aggregate of individuals of a biological species that are
geographically isolated from other members of the species and are ac-
tually or potentially interbreeding.
Quality assurance (QA) includes quality control functions and involves a
totally integrated program for ensuring the reliability of monitoring
and measurement data; the process of management review and over-
sight at the planning, implementation, and completion stages of envi-
ronmental data collection activities. Its goal is to assure that the data
provided are of the quality needed and claimed.
Quality control (QC) refers to the routine application of procedures for ob-
taining prescribed standards of performance in the monitoring and
measurements process; focuses on the detailed technical activities
needed to achieve data of the quality specified by data quality objec-
tives. Quality control is implemented at the bench or field level.
Range control refers to quality control activity through which measurement
values are kept within the range of natural or normal variability; con-
trol of operator variability.
Reasonably attainable refers to the ability of an aquatic resource to attain its
expected potential.
14,
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
A reference condition is the set of selected measurements or conditions of
minimally impaired waterbodies characteristic of a waterbody type in
a region.
A reference site is a specific locality on a waterbody which is minimally im-
paired and is representative of the expected ecological integrity of
other localities on the same waterbody or nearby waterbodies.
Regionalization or ecoregionalization is a procedure for subdividing a geo-
graphic area into regions of relative homogeneity in ecological systems
or in relationship between organisms and their environment.
Regulated flow of a stream or river is that for which the quantity of water
moving within its banks is a function of anthropogenic activity, usu-
ally associated with dams and reservoirs.
Residuals are the differences between a value predicted by regression and
an observed value.
Respiration is the energy expenditure for all metabolic processes. Matter
and energy are returned to the environment by respiration; matter as
CO2 and water, and energy as heat.
A riparian zone is an area that borders a waterbody.
Streams, as defined for the purpose of this document, are small lotic sys-
tems that can be waded by field investigators.
Targeted assemblage approach refers to an assessment procedure that has as its
focus of sampling a selected component of the biological community.
A targeted community segment is the component of the community, such as a
taxonomic category, trophic level, guild, or other designation, that is
the focus of a bioassessment.
Taxa richness refers to the number of distinct species or kinds (taxa) that are
found in an assemblage, community, or sample (see also diversity).
Termination control points are quality control elements that indicate when
and where nonvalid procedures are being used or data are being col-
lected and indicate necessary changes in procedures.
A test site is the location under study of which the condition is unknown
and suspect of being adversely affected by anthropogenic influence.
A vegetated buffer zone is a planted or naturally vegetated strip of land be-
tween some feature (usually a waterbody) and another landform or
habitat that has been altered by human activity (e.g., agricultural
fields, roadways, asphalt parking lots, residential areas).
A water resource assessment is an evaluation of the condition of a waterbody
using biological surveys, habitat quality assessments, chemical-spe-
cific analyses of pollutants in waterbodies, and toxicity tests. These en-
vironmental assessments may be diverse or narrowly focused
depending on the needs of the evaluation, and the probable sources of
degradation.
Zooplankton refers to animals which are unable to maintain their position
or distribution independent of the movement of water or air.
150
-------
References
Abel, P.D. 1989. Water Pollution Biology. In Ellis Horwood Series in Wastewater Tech-
nology. Ellis Horwood, Chichester, England. •
Adams, S.M., L.R. Shugart, G.R. Southworth, and D.E. Hinton. 1990. Application of
bioindicators in assessing the health of fish populations experiencing contaminant
stress. In J.F. McCarthy and L.R. Shugart, eds. Biomarkers of Environmental Con-
tamination. Lewis Publ., CRC Press, Boca Raton, FL.
Anderson, T.M. 1986. The Relationship Between Basin Cultivation and Stream Insect
Composition in Small Agricultural Streams in Kansas. M.A. Thesis, Univ. Kansas,
Lawrence.
Angermeier, P.L. and J.R. Karr. 1986. Applying an index of biotic integrity based on
stream-fish communities: considerations in sampling and interpretation. N. Am. J.
Fish. Manage. 6:418-29.
Armour, C.L., K.P. Burnham, and W.S. Platts. 1983. Field methods and statistical analysis
for monitoring small salmonid streams. FWS/OBS-83/33. U.S. Fish Wildl. Serv.,
Dep. Inter., Washington, DC.
Atkinson, S.F. 1985. Habitat-based methods for biological impact assessment. Environ.
Prof. 7:265-82. "
Austin, A., S. Lang, and M. Pomeroy. 1981. Simple methods for sampling periphyton
with observations on sampler design criteria. Hydrobiologica 85:33-47.
Austin, M.E. 1972. Land resource-regions and major land resource areas of the United
States. Agriculture Handbook 296. Soil Conserv. Serv., U.S. Dep. Agric., Washing-
ton, DC.
Bailey, R.G. 1976. Ecoregions of the United States. Map (scale 1:7,500,000). Intermoun-
tain Reg., Forest Serv., U.S. Dep. Agric., Ogden, UT.
Bain, M.B., J.T. Finn, and H.E. Booke. 1988. Streamflow regulation and fish community
structure. Ecology 69(2):382-92.
Ball, J. 1982. Stream classification guidelines for Wisconsin. In Water Quality Standards
Handbook. Off. Water Reg. Stand., U.S..Environ. Prot. Agency, Washington, D.C.
Barbour, M.T. et al. 1992. Evaluation of EPA's rapid bioassessment benthic metrics: met-
ric redundancy and variability among reference stream sites. J. Environ. Toxicol.
Chem. 11:437-49.
Barbour, M.T. and J.B. Stribling. 1991. Use of habitat assessment in evaluating the bio-
logical integrity of stream communities. Pages 25-38 in Biological Criteria: Research
and Regulation. EPA-440/5-91-005. Off. Water, U.S. Environ. Prot. Agency, Wash-
ington, DC.
Barbour, M.T. and S. Thornley. 1990. Quality assurance issues related to field assessment
of aquatic ecosystems. Pages 167-83 in D. Hart, ed. Proc. Third Annu. Ecol. Qual.
Assur. Workshop, Burlington, Ontario, Canada.
151
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Barton, D.R., W.D. Taylor, and R.M. Biette. 1985. Dimensions of riparian buffer strips re-
quired to maintain trout habitat in southern Ontario streams. N. Am. J. Fish. Man-
age. 5:364-78.
Batschelet, E. 1976. Introduction to Mathematics for Life Sciences. 2d ed. Springer-Ver-
lag, New York.
Bazata, K. 1991. Nebraska stream classification study. Rep. Surv. Water Sec., Water Qual.
Div., Nebraska Dep. Environ. Control, Lincoln, NE.
Benke, A.C., T.C. Van Arsdall Jr., and D.M. Gillespie. 1984. Invertebrate productivity in a
subtropical blackwater river: the importance of habitat and life history. Ecol.
Monogr. 54(l):25-63.
Berthouex, EM. and I. Hau. 1991. Difficulties related to using extreme percentiles for
water quality regulations. Res. J. Water Pollut. Control Fed. 63(6):873-79.
Borror, D.J., C. A. Triplehom, and N.R Johnson. 1989. An Introduction to the Study of In-
sects. 6th Ed. Saunders College Publ., Philadelphia, PA.
Boyle, T.P., G.M. Smillie, J.C. Anderson, and D.P. Beeson. 1990. A sensitivity analysis of
nine diversity and seven similarity indices. J. Water Pollut. Control Fed. 62:749-62.
Bramblett, R.G. and K.D. Fausch. 1991. Variable fish communities and the index of bi-
otic integrity in a western Great Plains river. Trans. Am. Fish. Soc. 120:752-69.
Brooks, R.P. et al. 1991. Selection of biological indicators for integrating assessments of
wetland, stream, and riparian habitats. Pages 81-89 in Biological Criteria: Research
and Regulation. EPA-440/5-91-005. Off. Water, U.S. Environ. Prot. Agency Wash-
ington, DC. >
Cairns Jr., J. 1982. Artificial Substrates. Ann Arbor Science Publ., Ann Arbor, MI.
Cairns Jr., J. and J.R. Pratt. 1986. Developing a sampling strategy. Pages 168-86 in B.G.
Isom, ed. Rationale for Sampling and Interpretation of Ecological Data in the As-
sessment of Freshwater Ecosystems. ASTM STP 894. Am. Soc. Test1 Mater., Philadel-
phia, PA.
Campbell, J.M. and W.J. Clark. 1983. Effects of microhabitat heterogeneity on the spatial
dispersion of small plant-associated invertebrates. Freshw. Invertebrate Biol.
2(4):180-85.
Caper, Jv G. Power, and F.R. Shivers, Jr. 1983. Chesapeake Waters, Pollution, Public
Health, and Public Opinion, 1607-1972. Tidewater Publ, Centreville, MD.
Carpenter, S.R. and D.M. Lodge, 1986. Effects of submerged macrophytes on ecosystems
processes. Aquat. Bot. 26:341-70.
Chambers, P.A., E.E. Prepas, H.R. Hamilton, and M.L. Bothwell. 1991. Current velocity
and its effect on aquatic macrophytes in flowing waters. Ecol. Appl. l(3}:249-57.
Chutter, F.M. 1972. An empirical biotic index of the quality of water in South African
streams and rivers. Water Resour. 6:19-20.
Coleman, R.L, and C.N. Dahm. 1990. Stream geomorphology: effects of periphyton
standing crop and primary production. J. N. Am. Benthol. Soc. 9:293-302.
Conquest, L.L., S.C. Ralph, and R.J. Naiman. 1994. Implementation of large-scale stream
monitoring efforts: sampling design and data analysis issues. Pages 69-90 in S.L.
Loeb and A. Spacie, eds., Biological Monitoring of Aquatic systems. Lewis Publ.,
Ann Arbor, MI.
Cooper, J.E. 1984. Vanishing species: The dilemma of resources without price tags. Pages
7-32 in A.W. Norden et al. eds. Threatened and Endangered Plants and Animals of
Maryland. Spec. Publ. 84-1. Maryland Nat. Heritage Prog.
Croonquist, M.J. and R.P. Brooks. 1991. Use of avian and mammalian guilds as indica-
tors of cumulative impacts in riparian-wetland areas. Environ. Manage. 1(5):701-14.
Crowley J.M. 1967. Biogeography. Can. Geogr. 11:312-26.
1.52
-------
CHAPTER 10:
References
Cuffney, T.E., M.E. Gurtz, and M.R. Meador. 1993. Methods for collecting benthic inver-
tebrate samples as part of the National Water Quality Assessment program. Open
file Rep. 93-406. U.S. Geological Survey, Raleigh, NC.
Cummins, K.VV. 1983. The natural stream ecosystem. In J.V. Ward and J.A. Stanford, eds.
The Ecology of Regulated Streams. Plenum Press, New York, NY.
Cummins, K.W. and M.J. Klug. 1979. Feeding ecology of stream invertebrates. Annu.
Rev. Ecol. Systematics 10:147-72.
Davies, S.P., L. Tsomides, D.L. Courtemanch, and F. Drummond. 1993. Biological Moni-
toring and Biocriteria Development Program. Maine Dep. Environ. Prot. Augusta,
ME.
Davis, W.S. and A. Lubin. 1991. Statistical validation of Ohio EPA's invertebrate com-
munity index. In Proc. 1989 Midw. Pollut. Control Biol. Meet., February 14-17,1989.
EPA-905/9-89-007. Chicago, IL.
Dickson, K.L., W.T. Waller, J.H. Kennedy, and L.P. Ammann. 1992. Assessing the Rela-
tionship between Ambient Toxicity and In-stream Biological Response. Environ.
Toxicol. Chem. 11:1307-22.
Dudley, D.R. 1991. A state perspective on biological criteria in regulation. Pages 15-18 in
Biological Criteria: Research and Regulation. Symp. Proc. EPA 440/5-91-005. Off.
Water, U.S. Environ. Prot. Agency, Washington DC,
Dudley, T.L., S.D. Cooper, and N. Hemphill, 1986. Effects of macroalgae on a stream in-
vertebrate community. J. N. Am. Benthol. Soc. 5:93-106.
EA, Inc., and Tetra Tech. 1994. Bioassessment for the Nonpoint Source Program. Rep.
Prep, for Florida Dep. Environ. Prot,, Tallahassee.
Edwards, C.J. and R.A. Ryder. 1990. Biological surrogates of mesotrophic ecosystem
health in the Laurentian Great Lakes. Rep. Great Lakes Sci. Advis. Board, Int. Joint
Comm., Windsor, Ontario, Canada.
Fassett, N.C. 1957. A Manual of Aquatic Plants. Univ. Wisconsin Press, Madison.
Fausch, K.D., J.R. Karr, and P.R, Yant. 1984. Regional application of an index of biotic in-
tegrity based on stream fish communities. Trans. Am. Fish. Soc. 113:39-55.
Fausch, K.D., C.L. Hawkes, and M.G. Parsons. 1988. Models that predict standing crop
of stream fish from habitat variables: 1980-85. Gen. Tech. Rep. PNW GTR-213, Pa-
cific Northw. Res. Sta., Forest Serv,, U.S. Dep. Agric., Portland, OR,
Fausch, K.D., J. Lyons, J.R. Karr, P.L. Angermeier. 1990. Fish communities as indicators
of environmental degradation. Am. Soc. Symp. 8:123-44.
Fenneman, N.M. 1946. Physical divisions of the United States. Map (scale 1:7,500,000).
U.S. Geol. Surv., Reston, VA.
Frey, D.G. 1975, Biological integrity of water - An historical approach. Pages 127-40 in
R.K. Ballentine and L.J. Guarraia, eds. The Integrity of Water. Off. Water Hazard,
. Mater., U.S. Environ. Prot. Agency, Washington, DC.
Gabler, R.E., R.J. Sager, S. Brazier, and D.L. Wise. 1976. Essentials of Physical Geography.
CBS College Publ., Dryden Press, New York, NY.,
Gallant, A.L, et al. 1989. Regionalization as a tool for managing environmental resources.
EPA 600/3-89/060. Environ. Res. Lab., U.S. Environ. Prot. Agency, Corvallis, OR.
Gerritsen, J. and M.L. Bowman. 1994. Periphytic Diatom Assemblages of high elevation
Rocky Mountain Lakes: Characterization of Reference Conditions. Rep. prep, for
Montana Dep. Health Environ. Sci., Helena, MT.
Gerritsen, J,, J. Green, and R. Preston. 1993. Establishment of regional reference condi-
tions for stream biological assessment and watershed management. Pres. at Water-
shed '93: A National Conference on Watershed Management, March 21-24, 1993.
Alexandria, VA.
153
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Gibson, G.R. 1992. Procedures for Initiating Narrative Biological Criteria. EPA-822-B-92-
002. U.S. Environ. Prot. Agency, Off. Sci. Technol., Washington, DC.
Gislason, J.C. 1985. Aquatic insect abundance in a regulated stream under fluctuating
and stable diel flow patterns. N. Am. J. Fish. Manage. 5:39-46.
Gorman, O.T. 1988. The dynamics of habitat use in a guide of Ozark minnows. Ecol.
Monogr. 58(1):1-18.
Gray J.S. 1989. Effects of environmental stress on species rich assemblages. Biol. J. Lin-
i>ean Soc. 37:19-32.
Gregg, W.W. and F.L. Rose. 1982. The effects of aquatic macrophytes on the stream mi-
croenvironment. Aquat. Bot. 14:309-24.
Gregg, W.W. and. F.L. Rose. 1985. Influences of aquatic macrophytes on invertebrate
community structure, guild structure, and microdistribution in streams. Hydrobi-
ologia 128:45-56.
Griffith, G.E. and J.M. Omernik. 1991. Alabama/Mississippi Project. Geographic Re-
search Component: Ecoregion/Subregion Delineation. ManTech Environ. Technol.
Inc., Environ. Res. Lab., U.S. Environ. Prot. Agency, Corvallis, OR.
Grimm, N.B. and S.G, Fisher. 1989. Stability of periphyton and macroinvertebrates to
disturbance by flash floods in a desert stream. J. N. Am. Benthol. Soc. 8:293-307.
Hamilton, P.B. and H.C. Duthie. 1984. Periphyton colonization of rock surfaces in a bo-
real forest stream studied by scanning electron microscopy and track autoradiogra-
phy. J. Phycol. 20:525-32.
Hammond, E.H. 1970. Classes of land-surface form, Map (scale 1:7,500,000). Pages 62-63
In The National Atlas of the United States of America. U.S. Geol. Surv., U.S. Dep.
Agric., Washington, DC.
Hart, D., ed. 1990. Proceedings: Third Annual Ecological Quality Assurance Workshop.
U.S. Environ. Prot. Agency, Can. Min. Environ., Burlington, Ontario.
Hart, D.D. 1992. Community organization in streams: the importance of species interac-
tions, physical factors, and chance. Oecologia 91:220-28. >
Hayslip, G.A. 1992. EPA Region 10 in-stream biological monitoring handbook (for wad-
able streams in the Pacific Northwest). Draft. EPA-910/9-92-013. Environ. Serv. Div.,
Reg. 10, U.S. Environ. Prot. Agency, Seattle, WA.
Herbertson, A.J. 1905. The major natural regions: An essay in systematic geography.
Geogr. J. 25:300-12. '
Hill, M.T., W.S. Platts, and R.L. Beschta. 1991. Ecological and geomorphoiogical con-
cepts for in-stream and out-of-channel flow requirements. Rivers 2(3):198-210.
Hilsenhoff, W.L. 1987. An improved biotic index of organic stream pollution. Great
Lakes Entomol. 20:31-39.
Holland, A.F., ed. 1990. Near Coastal Program Plan for 1990: Estuaries. EPA/600/4-40/033.
Environ. Res. Lab., Off. Res. Dev., U.S. Environ. Prot. Agency, Narragansett, RI.
Hsiao, T.H. 1988. Host specificity, seasonality, and bionomics of Leptinotarsa beetles.
Chapter 33 in P. Jolivet et al. eds. Biology of Chrysomelidae. Kluwer Acad. Publ.,
Boston, MA.
Huggins, D.G. and M. Moffett. 1988. A proposed biotic index and habitat development
index for use in streams in Kansas. Rep. 36. Kansas Biol. Surv., Lawrence, KS.
Hughes, R.M. 1989. Ecoregional biological criteria. In Proc. Water Qual. Stand, for the
21st Century. U.S. Environ. Prot. Agency, Dallas, TX.
Hughes, R.M. and J.M. Omernik. 1981. Use and misuse of the terms watershed and
stream order. Pages 320-36 in L. Krumholz, ed. Warm Water Streams Symp. Am.
Fish. Soc., Bethesda, MD. .
Hughes, R.M., D.P. Larsen, and J.M. Omernik. 1986. Regional reference sites: A method
for assessing stream potentials. Environ. Manage. 10:629-35.
154
-------
CHAPTER 10:
References
Hughes, R.M. and J.R. Gammon. 1987. Longitudinal changes in fish assemblages and
water quality in the Willamette River, Oregon. Trans. Am. Fish. Soc. 116(Z):196-209.
Hughes, R.M., E. Rexstad, and C.E. Bond. 1987. The relationship of aquatic ecoregions,
river basins, and physiographic provinces to the ichthyogeographic regions of Ore-
gon. Copeia 1987:423-32.
Hughes, R.M. and D.R Larsen. 1988. Ecoregions: an approach to surface water protec-
tion. J. Water Pollut. Control Fed. 60:486-93.
Hughes, R.M., C.M. Rohm, T.R. Whittier, and D.R Larson. 1990. A regional framework
for establishing recovery criteria. Environ. Manage. 14:673-84.
Hupp, C.R. 1992. Riparian vegetation recovery patterns following stream channeliza-
tion: a geomorphic perspective. Ecology 73(4):1209-26.
Hupp, C.R. and A. Simon. 1986. Vegetation and bank-slope development. Pages 83-92 in
Proc. 4th Fed. Interagency Sedimentation Conf.. Vol 2. Las Vegas, NV.
Hupp, C.R. and A. Simon. 1991. Bank accretion and the development of vegetated de-
positional surfaces along modified alluvial channels. Geomorphology 4:111-24.
Hurley, L.M. 1990. Field Guide to the Submerged Aquatic Vegetation of Chesapeake
Bay. Chesapeake Bay Estuary Prog., U.S. Fish Wildl. Serv., U.S. Dep. Inter., Annapo-
lis, MD.
Hynes, H.B.N. 1970. The Ecology of Running Waters. Univ. Toronto Press, Toronto, On-
tario, Canada.
Idaho Department of Health and Welfare. 1991. State funded 319 project: biological met-
ric development study plan. Pages 15-18 in Rapid Bioassessment Protocol Develop-
ment, Boise, ID.
- Iffrig, G.F. and M. Bowles. 1983. A compendium of ecological and natural subdivisions
of the U.S. Nat. Areas J. 3:3-11.
Joerg, W.L.G. 1914. The subdivision of North America into natural regions: a prelimi-
nary inquiry. Ann. Ass. Am. Geogr. 4:55-83.
Jones, R.C. and C.C. Clark. 1987. Impact of watershed urbanization on stream insect
communities. Water Resour. Bull. 23(6):1047-55.
Jongman, R.H.G., C.J.F. ter Braak, and O.F.R. van Tongeren, eds. 1987. Data Analysis in
Community and Landscape Ecology. Pudoc, Wageningen.
Karr, J.R. 1981. Assessment of biotic integrity using fish communities. Fisheries 6(6):21-
27.
. 1990. Biological integrity and the goal of environmental legislation: lessons for
conservation biology. Conserv. Biol. 4:244-50.
. 1991. Biological integrity: A long-neglected aspect of water resource manage-
ment. Ecol. Appl. 1:66-84.
Karr, J.R. and M. Dionne. 1991. Designing surveys to assess biological integrity in lakes
and reservoirs. Pages 62-72 in Biological Criteria: Research and Regulation.
EPA/440/5-91-005. Off. Water, U.S. Environ. Prot. Agency, Washington, DC.
Karr, J.R. and D.R. Dudley. 1981. Ecological perspective on water quality goals. Environ.
Manage. 5:55-68.
Karr, J.R. et al. 1986. Assessing Biological Integrity in Running Waters: A Method and Its
Rationale. Spec. Publ. 5. Illinois Nat. History Surv., Urbana, IL.
Karr, J.R. and B.L. Kerans. 1992. Components of biological integrity: Their definition
and use in development of an invertebrate IBI. Pages 1-16 in T.P. Simon and W.S.
Davis, eds. Environmental Indicators: Measurement and Assessment Endpoints.
EPA-905/R-92/003. Environ. Sci. Div., Reg. 5, U.S. Environ. Prot. Agency, Chicago,
IL.
Kay, J.J. 1990. A non-equilibrium thermodynamic framework for discussing ecosystem
integrity. Pages 209-37 in C.J. Edwards and H.A. Regier, eds. An Ecosystem Ap-
155
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
proach' to the Integrity of the Great Lakes in Turbulent Times. Spec. Publ. 90-4.
Great Lakes Fish. Comm., Ann Arbor, MI.
Kay, J.J. and E.D. Schneider. In press. Thermodynamics and measures of ecological in-
tegrity. In S. Paulsen et al. eds. Ecological Indicators. Elsevier Publ.
Kent, R.T. and K.E. Payne. 1988. Sampling groundwater monitoring wells: special qual-
ity assurance and quality control considerations. Pages 231-46 in L.H. Keith, ed.
Principles of Environmental Sampling. ACS Prof. Ref. Book. Am. Chem. Soc.,
Washington, DC.
Kerans, B.L. and J.R. Karr. In Press. Development and testing of a benthic index of biotic
integrity (B-IBI) for rivers of the Tennessee Valley. Ecol. Appl. 4.
Kerans, B.L., J.R. Karr, and S.A. Ahlstedt. 1992. Aquatic invertebrate assemblages: spa-
tial and temporal differences among sampling protocols. J. N. Am. Benthol. Soc.
ll(4):377-90.
Klemm, D.J., P.A. Lewis, E Fulk, and J.M. Lazorchak. 1990. Macroinvertebrate Field and
Laboratory Methods for Evaluating the Biological Integrity of Surface Waters.
EPA/600/4-90-030. Off. Res. Develop., U.S. Environ. Prot. Agency Washington, DC.
Kolkwitz, R, and M. Marsson. 1908. Okologie der pflanzlichen saprobien. Berich.
Deutsch. Bot. Gesellsch. 26:505-19.
Korte, V.L. and D.W. Blinn. 1983. Diatom colonization on artificial substrata in pool and
riffle zones studied by light and scanning electron microscopy. J. Phycol. 19:332-41.
Kuchler, A.W. 1964. Potential natural vegetation of the coterminous United States. Spec.
Publ. 36. Am. Geogr. Soc., A. Hoen & Co. Baltimore, MD.
Lamberti, G.A. and J.W. Moore. 1984. Aquatic insects as primary producers. Pages 164-
95 in V.H. Resh and D.M. Rosenburg, eds. The Ecology of Aquatic Insects. Praeger
Scientific, New York, NY.
Lamberti, G.A., L.R. Ashkenas, S.V. Gregory; and A.D. Steinman. 1987. Effects of three
herbivores on periphyton communities in laboratory streams, J. N. Am. Benthol.
Soc. 6:92-104.
LaPoint, T.W. and J.F. Fairchild. 1989. Aquatic surveys. In W. Warren-Hicks et al. eds.
Ecological Assessment of Hazardous Waste Sites: A Field and Laboratory Refer-
ence. EPA/600/3-89/013. Environ. Res. Lab., U.S. Environ. Prot. Agency, Corvallis,
OR.
Larsen, E.W., D.L. Johnson, and W.F. Lynch, Jr. 1986. A buoyant pop net for accurately
sampling fish at artificial habitat structures. Trans. Am. Fish. Soc. 115(2):351-55,
Lenat, D.R. 1988. Water quality assessment of streams using a qualitative collection
method for benthic macroinvertebrates. J. N. Am. Benthol. Soc. 7:222-33.
Leonard, P.M. and D.J. Orth. 1986. Application and testing of an index of biotic integrity
in small, coolwater streams. Trans. Am. Fish. Soc. 115:401-14.
Lincoln, R.J., G.A. Boxshall, and P.F. Clark. 1982. A Dictionary of Ecology, Evolution, and
Systematics. Cambridge, University Press, Cambridge, England.
Ludwig, J.A. and J.F. Reynolds. 1988. Statistical Ecology. John Wiley and Sons, Inc., New
York.
Lyons, J. 1989. Correspondence between the distribution of fish assemblages in Wiscon-
sin streams and Omernik's ecoregions. Am. Midi. Nat. 122:163-82.
. 1992. Using the index of biotic integrity (IBI) to measure environmental quality
in warmwater streams of Wisconsin. Gen. Tech, Rep. NC-149, Forest Serv., U.S.
Dep. Agric, St. Paul, MN.
MacArthur, R,H. 1972. Geographical Ecology. Princeton University Press, Princeton, NJ,
McCarthy, J.F. 1990. Concluding remarks: implementation of a biomarker-based envi-
ronmental monitoring program. Pages 429-39 in J.F. McCarthy and L.R. Shugart,
eds. Biomarkers of Environmental Contamination. Lewis Publ., Boca Raton, FL.
156
-------
CHAPTER 10:
References
McDermid, K.J. and R.J. Naiman. 1983. Macrophytes: Freshwater forest of lakes and riv-
ers. Am. Biol. Teacher 45:144-50.
MacDonald, L.H., A.W. Smart, and R.C. Wissmar. 1991. Monitoring guidelines to evalu-
ate effects of forestry activities on streams in the Pacific Northwest and Alaska.
Rep. EPA 910/9-91-001. U.S. Environ. Prot. Agency, Region 10, Seattle, WA.
Maine Department of Environmental Protection. 1992. Briefing prep, by S.P. Davies.
Augusta, ME.
. 1992. Long-term Response of Aquatic Communities to Differing Water Quality
Conditions in Three Rivers in Maine. Rep. prep, by S.P. Davies and Leon Tsomides.
Augusta, ME.
Mid-Atlantic Coastal Streams Workgroup. 1993. Standard Operating Procedures and
Technical Basis: Macroinvertebrate Collection and Habitat Assessment for Low-gra-
dient Nontidal Streams. Draft Rep. Delaware Dep. Nat. Res. Environ. Conserv.,
Dover.
Miller, D.L. et al. 1988. Regional applications of an index of biotic integrity for use in
water resource management. Fisheries 13(5):12-20.
Miller, A.C., D.C. Beckett, C.M. Way, and E.J. Bacon. 1989. The Habitat Value of Aquatic
Macrophytes for Macroinvertebrates. Tech. Rep. A-89-3. Waterways Exp. Sta., U.S.
Army Corps Eng., Vicksburg, MS.
Minshall, G.W. 1978: Autotrophy in stream ecosystems. Bioscience 28:767-71.
Molles, M.C. Jr. and C.N. Dahm. 1990. A perspective on El Nino and La Nina: global im-
plications for stream ecology. J. Natl. Benthol. Soc. 9(l):68-76.
Montana Department of Health and Environmental Sciences. 1990. Montana's Ap-
proach to Nonpoint Assessment and Monitoring. Outline. Water Qual. Bur.,
Helena, MT.
Moss, D., M.T. Furse, J.F. Wright, and P.D. Armitage. 1987. The prediction of the
macroinvertebrate fauna of unpolluted running-water sites in Great Britain using
environmental data. Freshw. Biol. 17:41-52.
Mount, D.I. and T.J. Norberg-King, eds. 1985. Validity of Effluent and Ambient Toxicity
Tests for Predicting Biological Impact, Scippo Creek, Circleville, Ohio. EPA 600/3-
85/044. U.S. Environ. Prot. Agency, Washington, DC.
Mount, D.I. and T.J. Norberg-King, eds. 1986. Validity of Effluent and Ambient Toxicity
Tests for Predicting Biological Impact, Kanawha River, Charleston, West Virginia.
EPA 600/3-86/006. U.S. Environ. Prot. Agency, Washington, DC.
Mount, D.I., T.J. Norberg-King, and A.E. Steen, eds. 1986. Validity of Effluent and Ambi-
ent Toxicity Tests for Predicting Biological Impact, Naugatuck River, Waterbury,
Connecticut. EPA 600/8-86/005. U.S. Environ. Prot. Agency, Washington, DC.
Mount, D.I., A.E. Steen, and T.J. Norberg-King, eds. 1986. Validity of Effluent and Ambi-
ent Toxicity Tests for Predicting Biological Impact, Back River, Baltimore Harbor,
Maryland. EPA 600/8-86/001. U.S. Environ. Prot. Agency, Washington, DC.
Mount, D.I., N. Thomas, M. Barbour, T. Norberg, T. Roush, and R. Brandes. 1984. Efflu-
ent and Ambient Toxicity Testing and In-stream Community Response on the
Ottawa River, Lima, Ohio. EPA 600/3-84-080. Permits Division, Washington, DC,
and Off Res. Dev., Duluth, MN.
Mount, D.I. et al., eds. 1985. Validity of Effluent and Ambient Toxicity Tests for Predict-
• ~ ' "• '• ' Mabama. EPA 600/8-85/015.
National Kesearcn council. restoration or Aquatic Ecosystems. Nat. Academy
Press, Washington, DC.
Newbury, R.W. 1984. Hydrologic determinants of aquatic insect habitats. Pages 232-57
in V.H. Resh and D.M. Rosenberg, eds. The Ecology of Aquatic Insects. Praeger Sci-
entific, New York, NY.
f, Washington, DC.I
U.S. Environ. Prot. Agency,
157
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Norberg-King, T.J. and D.I. Mount eds. 1986. Validity of Effluent and Ambient Toxicity
Tests for Predicting Biological Impact, Skeleton Creek, Enid, Oklahoma. EPA 600/8-
86/002. U.S. Environ. Prot. Agency, Washington, DC.
North Carolina Department of Health and Natural Resources. 1990. Standard Operating
Procedures, Biological Monitoring. Environ. Sci. Branch, Ecosystems Analysis Unit,
Biol. Assess. Group, Water Qual. Sec., Div. Environ. Manage., Raleigh, NC.
North Carolina Division of Environmental Management. 1978. 208 Phase I Results.
Raleigh, NC.
Noss, R.R 1990. Indicators for monitoring biodiversity: a hierarchical approach. Con-
serv. Biol. 4:355-64.
Oberdorff, T. and R.M. Hughes. 1992. Modification of an index of biotic integrity based
on fish assemblages to characterize rivers of the Seine-Normandie Basin, France.
Hydrobiologia 228:117-30.
Ohio Environmental Protection Agency, 1987. Biological Criteria for the Protection of
Aquatic Life. Vol. 1-3. Monitor. Assess. Prog., Surface Water Sec., Div. Water Qual.,
Columbus, OH.
. 1990. The Use of Biocriteria in the Ohio EPA Surface Water Monitoring and As-
sessment Program. Columbus, OH.
. 1991. Biological and Water Quality Study of the Hocking River Mainstem and
Selected Tributaries: Fairfield, Hocking, and Athens County, Ohio. Rep., Columbus.
. 1992. Ohio Water Resource Inventory. Executive Summary and Vol 1: Summary,
Status, and Trends, Columbus.
Omernik, J.M. 1987. Ecoregions of the conterminous United States. Annu. Ass. Am.
Geogr. 77(l):118-25.
Omernik, J.M. and G.E. Griffith. 1991. Ecological regions versus hydrologic units:
Frameworks for managing water quality. J. Soil Water Conserv. 46(5):334-40.
Omernik, J.M., M.A. Shirazi, and R.M. Hughes. 1982. A synoptic approach for regional-
izing aquatic ecosystems. Pages 199-218 in In-place Resource Inventories: Principles
and Practices. Proc. Soc. Am. Foresters, Univ. Maine, Orono.
Omernik, J.M. and A.J. Kinney. 1985. Total alkalinity of surface waters: a map of the
New England and New York region. EPA-600/D-84-216. Environ. Res. Lab., U.S.
Environ. Prot. Agency, Corvallis, OR.
Omernik, J.M. and G.E. Griffith. 1986. Total alkalinity of surface waters: a map of the
Upper Midwest region. Map (scale 1:2,500,000). Environ. Manage. 10:829-39.
. 1991. Ecological regions versus hydrologic units: frameworks for managing
water quality. J. Soil Water Conserv. 46(5):334-40.
Omernik, J.M., C.M. Rohm, S.E. Clarke, and D.P. Larsen. 1988. Summer total phospho-
rus in lakes: a map of Minnesota, Wisconsin, and Michigan. Environ. Manage.
12:815-25.
Omernik, J.M. and A.L. Gallant. 1990. Defining regions for evaluating environmental re-
sources. In Global Natural Resource Monitoring and Assessments. Proc. Int. Conf.,
Venice, Italy.
Oregon Department of Environmental Quality. 1991. Biological Criteria-Implementation
Plan. Draft. Portland, OR.
Osborne, L.L. et al. 1991. Stream habitat assessment programs in states of the AFS north
central division. Fisheries 16(3):28-35.
Overton, J. 1991. Utilization of biological information in North Carolina's Water Quality
Regulatory Program. In Biological Criteria: Research and Regulation. Off. Sci. Tech-
nol., U.S. Environ, Prot. Agency, Washington, DC.
15i
-------
CHAPTER 10:
References
Padgett, W.G., A.P. Youngblood, and A.H. Winwood. 1989. Riparian Community Type
Classification of Utah and Southeastern Idaho. R4-Ecol-89-01. Intermountain Res.
Sta., Forest Serv., U.S. Dep. Agric.'Ogden, UT.
Pandit, A.K. 1984. Role of macrophytes in aquatic ecosystems and management of fresh-
water resources. J. Environ. Manage. 18:73-88.
Patrick, R. 1949. A proposed biological measure of stream conditions, based on a survey
of the Conestoga Basin, Lancaster County, Pennsylvania. Notulae Naturae of the
Nat. Acad. Sci. CI 277, Philadelphia, CI, 277.
Paulsen, S.G. et al. 1991. EMAP-Surface Water Monitoring and Assessment Program —
Fiscal Year 1991. EPA/600/3-91/022. Off. Res. Dev., U.S. Environ. Prot. Agency,
Washington, DC.
Plafkin, J.L. 1989. Water Quality-based controls and ecosystem recovery. Pages 87-96 in
J.Cairns Jr., ed. Rehabilitating Damaged Ecosystems. Vol. 2. CRC Press, Boca Raton,
FL-
Plafkin, J.L. et al. 1989. Rapid Bioassessment Protocols for Use in Streams and Rivers:
Benthic Macroinvertebrates and Fish. EPA/444/4-89-001. Off. Water, U.S. Environ.
Prot. Agency, Washington, DC.
Platts, W.S., W.F. Megahan, and G.W. Minshall. 1983. Methods for Evaluating Stream,
Riparian, and Biotic Conditions. Gen. Tech. Rep. INT-138. Intermountain Res. Sta.,
Forest Serv., U.S. Dep. Agric., Ogden, UT.
Platts, W.S. et al. 1987. Methods for Evaluating Riparian Habitats with Applications to
Management. Rep. Int-221. Intermountain Res. Sta., Forest Serv., U.S. Dep. Agric.,
Ogden, UT.
Poff, N.L., N.J. Voetz, and J.V. Ward. 1990. Algal colonizationmnder four experimentally
controlled current regimes in a high mountain stream. J. N. Am. Benthol. Soc. 9:303-
18.
Pollard, J.E. 1981. Investigator differences associated with a kicking method for sam-
pling macroinvertebrates. J. Freshw. Ecol. 1:215-24.
Power, M.E.' 1990. Effects of fish in river food webs. Science 250:811-14.
Price, P.W. 1975. Insect Ecology. John Wiley, Sons, New York, NY.
Primrose, N.L. 1989. Routine Benthic Biomonitoring Protocol: A Proposal. Maryland
Dep. Environ., Annapolis, MD.
Primrose, N.L., W.L. Butler, and E.S. Friedman. 1991. Assessing biological integrity us-
ing EPA Rapid Bioassessment Protocol II — The Maryland Experience. Pages 131-
32 in Biological Criteria: Research and Regulation. EPA/440/5-91-005. Off. Water,
U.S. Environ. Prot. Agency, Washington, DC.
Pringle, C.M. et al. 1988. Patch dynamics in lotic systems: the stream as a mosaic. J. N.
Am. Benthol. Soc. 7:503-24.
Rankin, E.T. 1989. The Qualitative Habitat Evaluation Index (QHEI): Rationale, Meth-
ods, and Application. Div. Water Qual. Plann. Assess., Ohio Environ. Prot. Agency,
Columbus, OH.
Rankin, E.T. 1991. The use of the qualitative habitat evaluation index for use attainabil-
ity studies in streams and rivers in Ohio. Page 133 in Biological Criteria: Research
and Regulation. EPA-440/5-91-005. Off. Water, U.S. Environ. Prot. Agency, Wash-
ington, DC.
Rankin, E.T. and C. Yoder. 1991. Calculation and uses of the area of degradation value
(ADV). In Ohio Water Resource Inventory. Exec. Sum. and Vol. 1. Ohio Environ.
Prot. Agency, Columbus, OH.
.1992. Summary, Status and Trends. In Ohio Water Resource Inventory, Vol. 1.
Ohio Environ. Prot. Agency, Columbus, OH.
159
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
Reckhow, K, In review. Biological Criteria: Technical Guidance for Survey Design and
Statistical Evaluation for Biosurvey Data. Off. Res. Dev., Off. Sd. Technol., U.S. En-
viron. Prot. Agency, Washington, DC.
Resh, V.H. and J.K. Jackson. 1993. Rapid assessment approaches in benthic macroinver-
tebrate biomonitoring studies. In D.M. Rosenberg and V.H. Resh, eds. Freshwater
Biomonitoring and Benthic Macroinvertebrates. Chapman and Hall, New York, NY.
Rohm, C.M., J.W. Glese, and C.C. Bennett. 1987. Evaluation of an aquatic ecoregion clas-
« sification of streams in Arkansas. Freshw. Ecol. 4:127-40,
Rosgen, D.L. 1994. A classification of natural rivers. Catena 22:169-99.
Schelske, C.L. 1984. In situ and natural phytoplankton assemblage bioassays. Pages 15-
47 in L.E. Shubert, ed. Algae as Ecological Indicators. Academic Press, London,
England.
Science Advisory Board. 1993. Evaluation of Draft Technical Guidance on Biological Cri-
teria for Streams and Small Rivers (prep, by the biological criteria subcommittee of
the ecological processes and effects committee). An SAB Rep. EPA-SAB-EPEC-94-
003. U.S. Environ. Prot. Agency, Washington, DC.
Schuster, E.G. and H.R. Zuuring. 1986. Quantifying the unquantifiable: or have you
stopped abusing measurement scale? J. Forest. 1986:25-30.
Seddon, B. 1972. Aquatic macrophytes as limnological indicators. Freshw. Biol. 2:107-30.
Shackleford, B. 1988. Rapid Bioassessments of Lotic Macroinvertebrate Communities:
Biocriteria Development. Arkansas Dep. Pollut. Control Ecol., Little Rock, AR.
Simon, A. and C.R. Hupp. 1987. Geomorphic and vegetative recovery processes along
modified Tennessee streams: an interdisciplinary approach to disturbed fluvial sys-
tems. Int. Ass. Hydrol. Sri. 167:251-262.
Smith, R.L. 1974. Ecology and Field Biology. 2nd ed. Harper & Row, New York, NY.
Smith, G.R., J.N. Taylor, and T.W. Grimshaw. 1981. Ecological survey of fishes in the Rai-
sin River drainage, Michigan. Mich. Acad. 13:275-305.
Smith, F., S. Kulkarni, L.E. Myers, and M.J. Messner. 1988. Evaluating and presenting
quality "assurance data. Pages 157-68 in L.H. Keith, ed. Principles of Environmental
Sampling. ACS Prof. Ref. Book. Am. Chem. Soc., Washington, DC,
Standard Methods for the Examination of Water and Wastewater. 1989. 17th ed. Joint
Editorial Board, Am. Pub. Health Ass., Am. Water Works Ass., and Water Pollut.
Control Fed., Washington, DC.
Steedman, R.J. 1988. Modification and assessment of an index of biotic integrity to
quantify stream quality in southern Ontario. Can. J. Fish. Aquat. Sci. 45:492-501.
Steinman, A.D. and C.D. Mclntire. 1986. Effects of current velocity and light energy on
the structure of periphyton assemblages in laboratory streams. J. Phycol. 22:352-61.
. 1987. Effects of irradiance on the community structure and biomass of algal as-
semblages in laboratory streams. Can. J. Fish. Aquat. Sci. 44:1640-48.
Steinman, A.D. et al. 1987. Effects of herbivore type and density on taxonomic structure
and physiognomy of algal assemblages in laboratory streams. J. N. Am, Benthol.
Soc. 6:175-88.
Steinman, A.D. and A.F. Parker. 1990. Influence of substrate conditioning on periphytic
growth in a heterotrophic woodland stream. J. N. Am. Benthol. Soc. 9:170-79.
Stevenson, R.J. 1990. Benthic algal community dynamics in a stream during and after a
spate. J. N. Am. Benthol. Soc. 9:277-88.
Stribling, J.B. and M.T. Barbour. 1991. Application of QA concepts from the third eco-
logical workshop to a long-term aquatic field assessment pilot study. Proc. 4th Eco-
logical QA Workshop. U.S. Environ. Prot. Agency, Environ. Can., Cincinnati, OH.
Tauber, M.J., C.A. Tauber, and S. Masaki, 1986. Seasonal Adaptations of Insects. Oxford
Univ. Press, New York, NY.
160
-------
CHAPTER 10:
References
Tetra Tech, Inc. 1993. Evaluation of combined sewer overflows (CSOs) in Ohio using
rapid bioassessment protocols (RBPs). Draft Rep. Off. Sri. Technol. and Off. Wet-
lands, Oceans and Watersheds. U.S. Environ. Prot. Agency, Washington, DC.
Townsend, C.R. 1989. The patch dynamics concept of stream community ecology. J. N.
Am. Benthol. Soc. 8:36-50.
Ulanowicz, R.E. 1990. Ecosystem integrity and network theory. Pages 69-77 in C.J. Ed-
wards and H.A. Regier, eds. An Ecosystem Approach to the Integrity of the Great
Lakes in Turbulent Times. Spec. Pub. 90-4. Great Lakes Fish. Comm., Ann Arbor,
MI.
U.S. Department of Agriculture. 1981. Land resource regions and major land resource
areas of the United States. Agricultural Handbook 296. U.S. Gov. Print. Off., Wash-
ington, DC.
U.S. Environmental Protection Agency, 1980a. Guidelines and Specifications for. Prepar-
ing Quality Assurance Program Plans. QAM5-004/80. Washington, DC.
— . 1980b. Interim Guidelines and Specifications for Preparing Quality Assurance
Project Plans. QAMS-005/80. Washington, DC.
—:—. 1983. Technical Support Manual: Waterbody Surveys and Assessments for Con-
ducting Use Attainability Analyses. Vol. 1-3. Off. Water Reg. Stand., Washington,
DC.
. 1984a. Policy and Program Requirements to Implement the Quality Assurance
Program. EPA Order 5360.1. Washington, DC.
. 1984b. The Development of Data Quality Objectives. Qual. Assur. Manage. Staff,
DQO Workgroup, Washington, DC.,
. 1984c. Guidance for Preparation of Combined Work/Quality Assurance Project
Plans for Environmental Monitoring. Rep. OWRS QA-1. Washington, DC.
. 1985a. Technical Support Document for Water Quality-based Toxics Control.
EPA/440/4-85/03. Washington DC.
. 1985b. Methods for Measuring the Acute Toxicity of Effluents to Freshwater and
Marine Organisms. EPA/600/4-85/013. Cincinnati, OH.
. 1986. Development of Data Quality Objectives — Descriptions of Stages I and II.
Qual. Assur. Manage. Staff, Washington, DC.
. 1987a. Report of the National Workshop on In-stream Biological Monitoring
and Criteria. In-stream Biol. Criteria Comm., Off. Water Reg. Stand., Environ. Res.
Lab., Reg. 5, Corvallis, OR.
. 1987b. Surface Water Monitoring: A Framework for Change. Off. Water, Off. Pol-
icy Plann. F.val., Washington, DC.
. 1988. Proceedings of the First National Workshop on Biological Criteria, Lin-
colnwood, Illinois, December 2-4,1987. Rep. 905/9-89/003. Chicago, IL.
. 1989. Short-term Methods for Estimating the Chronic Toxicity of Effluents and
Receiving Water to Freshwater Organisms. EPA/600/4-89/001. Environ. Monitor.
Systems Lab., Cincinnati, OH.
. 1990. Biological Criteria: National Program Guidance for Surface Waters. EPA-
440/5-90-004. Qff. Water, Washington, DC.
. 1991a. Biological Criteria: State Development and Implementation Efforts. EPA-
440/5-91-003. Off. Water, Washington, DC.
— . 1991b. Biological Criteria Guide to Technical Literature. EPA-440/5-91-004. Off.
Water, Washington, DC,
. 1991c. Technical Support Document for Water Quality-based Toxics Control.
EPA-505/2-90-001. Off. Water, Washington, DC.
—. 1991d. Biological Criteria: Research and Regulation Proceedings of the Sympo-
sium. EPA-440/5-91-005. Off. Water, Washington, DC.
-------
BIOLOGICAL CRITERIA:
Technical Guidance for Streams and Small Rivers
. 1991e. Report of the ecoregions subcommittee of the ecological processes and ef-
fects committee. Evaluation of the Ecoregion Concept, EPA-SAB-EPEC-91-003.
Washington, DC,
. 1991f. The Watershed Protection approach: An Overview, EPA 503/9-92/002.
Off. Water, Washington, DC.
. 1993. The Watershed Protection approach. Annual Report 1992. EPA 840-S-
93/001. Off. Water, Washington, DC.
U.S. Government Printing Office. 1988. The Clean Water Act as amended by the Water
Quality Act of 1987. Public Law 100-4. Washington, DC.
Vannote, R.L. et al. 1980. The river continuum concept. Can. J. Fish. Aquat. Sd. 37:130-37.
Weber, C.I. 1973. Biological monitoring of its aquatic environment by the Environmental
Protection Agency. In Biological Methods for the Assessment of Water Quality.
ASTM STP 528. Am. Soc. Test. Mater., Philadelphia, PA.
Wallace, A.R, 1869. The Malay Archipelago. MacMillan, London, England.
Westlake, D.F. 1975. Macrophytes. Pages 106-28 in B.A. Whitton, ed. River Ecology. Univ.
California Press, Berkeley,
WhitHer, T.R., R.M. Hughes, and D.P. Larsen. 1988. Correspondence between ecoregions
and spatial patterns in stream ecosystems in Oregon. Can. J. Fish. Aquat. Sci.
45:1264-78.
Wilhelm, G. and D. Ladd. 1988. Natural Area Assessment in the Chicago Region. Pages
361-75 in Trans. 53rd N. Am. Wildl. Nat. Resour. Conf.
Wilhm, J.L. and T.C. Dorris. 1968. Biological parameters of water quality criteria. Bios-
cience 18:477-81.
Williams, C.E. and A.L. Caskey. 1965. Soil fertility and cottontail fecundity in southeast-
ern Missouri. Am. Midi. Nat. 74:211-24.
Wright, J.F., D. Moss, P.D. Armitage, and M.T. Furse, 1984. A preliminary classification of
running-water sites in Great Britain based on macroinvertebrate species and the
prediction of community type using environmental data. Freshw. Biol. 14:221-56.
Yankey, R. et al. 1991. Rock Creek: Rural Clean Water Program. Final Report
1981 — April 1991. Agric. Stabil. Conserv. Serv,, Soil Conserv. Serv., Agric. Res.
Serv., U.S. Dep. Agric., Idaho Div. Environ. Qual., Twin Falls and Snake River Soil
Conserv. Dist., Twin Falls, ID.
Yoder, C.0.1991. The integrated biosurvey as a tool for evaluation of aquatic life use at-
tainment and impairment in Ohio Surface Waters. In Biological Criteria: Research
and Regulation. EPA-440/5-91-005. Off. Water, U.S. Environ. Prot. Agency, Wash-
ington, DC.
162
------- |