United States Office of Water August 1998
Environmental Protection Washington, DC EPA 841-B-98-007
Agency (4503F)
&EPA
Lake and Reservoir
Bioassessment and
Biocriteria
Technical Guidance
Document
-------
Technical Guidance Document EPA 841-B-98-007
LAKE AND RESERVOIR BIOASSESSMENT
AND BIOCRITERIA
TECHNICAL GUIDANCE DOCUMENT
August 1998
Office of Wetlands, Oceans, and Watersheds (4503F)
Office of Science and Technology (4304)
Office of Water
U.S. Environmental Protection Agency
Washington, DC
-------
Technical Guidance Document EPA 841-B-98-007
ACKNOWLEDGMENTS
Principal authors of this document are Jeroen Gcrritsen (Tetra Tech, Inc.), Robert E.
Carlson (Kent State University), Donald L. Dycus (Tennessee Valley Authority). Chris
Faulkner (USEPA), George R. Gibson (USEPA), John Harcum (Tetra Tech, Inc.), and
S. Abby Markowitz (Tetra Tech, Inc.).
Additional scientific, technical, editorial, and production contributions were made by
Michael T. Barbour (Tetra Tech, Inc.), Allison M. Cauthorn (Tetra Tech, Inc.), Donald F.
Charles (Patrick Center for Environmental Research), Joyce Lathrop-Davis (Tetra Tech,
Inc.), Robert H. Kennedy (USACE), Edwin H. Liu (USEPA), Paula J. Proctor (Tetra Tech,
Inc.), Susan Ratcliffe (USEPA), Amanda L. Richardson (Tetra Tech, Inc.), Blaine D. Snyder
CTetra Tech, Inc.), James B. Stribling (Tetra Tech, Inc.), Jeffrey S. White (Tetra Tech, Inc.),
Sarah White (Tetra Tech, Inc.), Michel Schuster (Tetra Tech, Inc.), Emily Faalasli (Tetra
Tech, Inc.), and Kelly Gathers (Tetra Tech, Inc.). This document was prepared by Tetra
Tech, Inc.
Development of the protocols and this guidance document would not have been possible
without the substantial contributions of the Lake Bioassessment Workgroup. Members of
the Lake Bioassessment Workgroup included: Loren Bahls (Montana DHES), Michael T.
Barbour (Tetra Tech, Inc.}, Mary iJelefsIci fX.JSisPAJ, ^^[ilce tsii'ci (I h.IP^11 Don ^3oxmeaii (Io^va.
DNR), Roy Bouchard (Maine DEP), Dan Butler (Oklahoma Conservation Commission),
Robert E. Carlson (Kent State University), Neil E. Carriker (Tennessee Valley Authority),
Donald F. Charles (Patrick Center for Environmental Research), Dave Courtemanch (Maine
DEP), Donald L. Dycus (Tennessee Valley Authority), Chris Faulkner (USEPA), Jeroen
Gerritsen (Tetra Tech, Inc.), George R. Gibson (USEPA), David Hart (Patrick Center for
Environmental Research), Steven Heiskary (MPCA), James Hulbert (Florida DEP), Susan
Jackson (USEPA), Jack Jones (University of Missouri), Edwin Liu (USEPA), Ed Mills (Cornell
University), Steve Paulsen (USEPA), Spencer Peterson (USEPA), Susan Ratcliffe (USEPA),
Donna Sefton (USEPA), Tom Simon (USEPA), Kathy Stecker (South Carolina DHEC), and
Richard S. Stemberger (Dartmouth College).
Cover artwork by Blaine Snyder of Tetra Tech, Inc.
-------
CONTENTS
Preface i
PART 1: BIOLOGICAL APPROACHES TO WATER QUALITY MANAGEMENT
1. The Protection of Biological Integrity 1-1
1.1 Introduction 1-1
1.2 The Concept of Biocriteria 1-2
1.3 Uses of Bioassessments and Biocriteria 1-3
1.4 Other Biocriteria Reference Documents 1-5
2. Lake Biological Monitoring in EPA, Local, State, Tribal and Regional Protection and
Management Programs 2-1
2.1 Section 305(b)-Water Quality Assessment 2-3
2.2 Section 314-Clean Lakes Program 2-4
2.3 Section 319-Nonpoint Source Program 2-5
2.4 Watershed Protection Approach 2-5
2.5 Section 303(d)-The TMDL Program 2-6
2.6 NPDES Permits and Individual Control Strategies 2-7
2.7 Risk Assessment 2-8
2.8 USEPA Water Quality Criteria and Standards 2-8
2.9 Other Uses 2-9
3. Overview of Bioassessment and Biocriteria 3-1
3.1 Conceptual Framework 3-1
3.2 Application to Lakes 3-4
3.3 Biocriteria 3-8
PART 2: LAKE BIOLOGICAL MONITORING AND ASSESSMENT PROTOCOLS
4. Selection and Characterization of Reference Conditions 4-1
4.1 Regionalization and Preliminary Classification 4-1
4.2 Establishing Reference Conditions 4-5
5. Habitat Measurement 5-1
5.1 Watershed Habitat 5-2
5.2 In-lake Habitat 5-3
5.3 Shorezone and Littoral Habitat 5-3
6. Biological Assemblages 6-1
6.1 Primary Producers: Trophic State Assessment 6-2
6.2 Submerged Macrophytes 6-2
6.3 Sedimented Diatoms : 6-2
6.4 Benthic Macroinvertebrates 6-3
6.5 Fish 6-4
6.6 Phytoplankton Assemblage 6-5
6.7 Zooplankton 6-5
6.8 Periphyton 6-6
-------
7. Tiered Sampling 7-1
7.1 Desktop Screening 7-1
7.2 Tier 1: Trophic State and Macrophytes 7-9
7.3 Tier 2A: Biological Assemblage Assessment 7-14
7.4 Tier 2B: Short-term Indicators (Repeated Sampling) 7-22
7.5 Diagnostic Habitat Survey 7-23
8. Index Development 8-1
8.1 Overview 8-1
8.2 Characterization of Reference Condition 8-2
8.3 Index Development 8-4
8.4 Lake Tier Indices 8-8
9. Quality Assurance 9-1
9.1 Program Design 9-1
9.2 Sampling Design 9-4
9.3 Evaluation of Statistical Power 9-7
9.4 Management 9-14
9.5 Operational Quality Control 9-14
10. Biocriteria Implementation 10-1
10.1 Characteristics of Effective Biocriteria 10-1
10.2 Steps to Implementation 10-2
10.3 Technical Considerations 10-5
10.4 Program Resources 10-7
Appendix A Glossary of Terms
Appendix B Comparison of Existing Lakes Protocols
Appendix C Paleolimnological Sampling
Appendix D Biological Assemblages
Appendix E Statistical Analysis Methods for Biological Assessment
Appendix F Executive Summaries of State Pilot Studies
Appendix G Literature Cited
-------
LIST OF FIGURES
1-1 Interdependence of environmental monitoring and environmental criteria.
3-1 Effects of pollutants in lakes.
3-2 Biological assemblages used for lake assessments.
3-3 Distribution of TVA reservoirs in ecoregions.
4-1 Minnesota ecoregions and sampled lakes. From Heiskary 1989.
4-2 Chlorophyll a concentration of Minnesota reference lakes by ecoregion.
6-1 Distribution of responsive macroinvertebrate metrics in Florida lakes (from Gerritsen and White
1997).
6-2 Relationship of two highly correlated attributes, Shannon diversity and percent dominanace
(from Gerritsen and White 1997).
6-3 Distribution of nutrients, Secchi depth, and chlorophyll a in Florida reference and test lakes.
7-1 Tiered sampling structure.
7-2 Example of desktop screening questionnaire.
7-3 Sampling zones in large or complex lakes (large reservoirs, multi-basin lakes).
7-4 Integrated sampling, Tiers 1 and 2.
7-5 Example scoring sheet for shorezone habitat.
7-6 Florida lakes sampling scheme.
8-1 Basis of bioassessment scores—unimpaired reference sites; population distribution.
8-2 Species richness in TVA reservoirs.
8-3 Benthic macroinvertebrate taxa richness in littoral zone of Montana lakes and wetlands.
8-4 Assessing candidate metrics.
8-5 Responsiveness of metrics.
8-6 Total crustacean zooplankton taxa in North American lakes (redrawn after Dodson 1992).
8-7 Overall ecological condition of tributary reservoirs in the Tennessee Valley in 1994 using aggre-
gate of 5 indices.
8-8 1994 TVA ecological condition summary.
9-1 Illustration of the trade-offs in power and significance.
E-1 Graphical representation of bioassessment.
E-2 Box and whisker diagram (after Tukey 1977).
E-3 Ordination.
E-4 Canonical correspondence analysis of periphytic diatom communities from Rocky Mountain
Lakes.
E-5 Dendrogram from cluster analysis.
E-6 Illustration of discriminant function analysis.
E-7 Misclassification of test sites.
E-8 Assessing candidate metrics that have (a) high values under reference conditions, and (b) low
values under references conditions.
E-9 Illustration of alternative scoring methods, using an upper percentile, a lower percentile, or a
central tendency.
E-10 Total crustacean zooplankton taxa in North American lakes (redrawn after Dodson 1992).
E-ll Plot showing the regression line used to estimate salinity-adjusted species richness measures for
mean number of species per site.
E-12 Assessment by ordination.
-------
LIST OF TABLES
1-1 Biocriteria reference documents.
2-1 Applications of lake biological monitoring protocols and biocriteria.
3-1 Sampling tier summary.
4-1 Comparison of elements for characterizing reference conditions.
5-1 Watershed and basin habitat measurement and metrics.
5-2 Physical and chemical measurements and metrics.
5-3 Lakcshore habitat measurements and metrics.
6-1 Potential algal trophic state metrics.
6-2 Potential macrophyte metrics.
6-3 Potential sedimented diatom metrics.
6-4 Potential benthic macroinvertebrate metrics.
6-5 Potential fish metrics.
6-6 Potential phytoplankton metrics.
6-7 Potential zooplankton metrics.
6-8 Potential periphyton metrics.
7-1 Desktop screening assessment.
7-2 Tier 1: Trophic state and macrophyte sampling.
7-3 Sampling summary for chlorophyll, water quality, and phytoplankton.
7-4 Tier 2A: Routine biological sampling.
7-5 Sampling summary for submerged macrophytes.
7-6 Sampling summary for benthic macroinvertebrates.
7-7 Benthic macroinvertebrate sampling gear appropriate for major substrate types.
7-8 Sampling summary for fish assemblages.
7-9 Sampling summary for sedimented diatoms.
7-10 Tier 2B: Water column biological sampling.
7-11 Sampling summary for crustacean zooplankton.
7-12 Sampling summary for periphytic diatoms.
7-13 Supplemental components.
8-1 Example of TVA's computational method for evaluation of reservoirs: Wilson Reservoir 1994
(run-of-the-river reservoir).
9-1 Examples of sample units and populations.
9-2 Errors in hypothesis testing.
9-3 Common values of (Z + Z2Q)2 for estimating sample size for use with Equations 1 and 2
(Snedecor and Cochran 1980).
9-4 Number of taxa and individuals in 12 cumulative PONAR samples from 9 Florida lakes.
9-5 Comparison of two sample processing protocols, Florida lakes.
9-6 Minimum and maximum values, and standard deviations of repeated measures, of reservoir fish
metrics and the REAL
9-7 Example QC elements for field and laboratory activities
10-1 Sequential progression of the biocriteria process.
B-l Comparison of lakes protocols with EMAP, TVA Reservoirs, and Clean Lakes.
C-l Potential paleolimnological metrics.
C-2 EMAP analytical methods: sedimented diatoms indicator (from USEPA 1994b).
-------
D-l Advantages, disadvantages, and alternatives to using algal assemblages.
D-2 Potential algal metrics.
D-3 Advantages, disadvantages, and alternatives to using macrophyte assemblages.
D-4 Potential macrophyte metrics.
D-5 Advantages, disadvantages, and alternatives to using macroinvertebrate assemblages.
D-6 Potential benthic metrics.
D-7 Advantages, disadvantages, and alternatives to using zooplankton assemblages.
D-8 Potential zooplankton metrics.
D-9 Advantages, disadvantages, and alternatives to using fish assemblages.
D-10 Fish assemblage metrics under investigation by TVA. After Dycus and Meinert (1994) and
Hickman and McDonough (1996).
E-l Results of discriminant analysis test of alternative classification.
-------
This technical guidance document is based on
the concept that bioassessment and bioeriteria
programs for lakes and reservoirs are interre-
lated and critical components of comprehensive
water resource protection and management. The
United States has approximately 40 million
acres of lakes, ponds, and reservoirs. For the
decade following the passage of the Clean Water
Act in 1972, the Nation's lake acreage that
experienced a decline in water quality was four
times the acreage that experienced improvement
(Johnson 1989). Managing, protecting, and
restoring these waterbodies has been, and will
continue to be, a challenge requiring the balanc-
ing of human and environmental health con-
cerns with economic feasibility.
Our increased understanding of how lake sys-
tems function and respond to human activity has
led to the recognition that environmental protec-
tion requires a holistic approach to lake manage-
ment and protection. It has been necessary to
expand our thinking in regard to lake monitoring
approaches, incorporating biological assessments
into traditional chemical and physical evalua-
tions.
Section 101 of the Clean Water Act requires
federal and state governments to "restore and
maintain the chemical, physical and biological
integrity of the Nation's waters." Natural,
undisturbed aquatic ecosystems have high
biological integrity, which is defined as "the
condition of an aquatic community inhabiting
The goal of this guidance is to
assist in protecting the
ecological integrity of the
Nation's lake and reservoir
resources.
unimpaired waterbodies of a specified habitat as
measured by an evaluation of multiple attributes
of the aquatic biota. 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
characteristics such as
taxonomic richness and
composition, trophic
structure, and (3) is found in the study biogeo-
graphic region." (USEPA 1996a).
In 1992, the National Research Council of the
National Academy of Sciences, calling for
improved assessment programs to more effec-
tively target lake restoration efforts, recom-
mended the following (NRC 1992):
There is a great need for cost-effective, reliable
indicators of ecosystem Junction, including those
that will reflect long-term change and response
to stress. Research on indicators should include
traditional community and ecosystem measure-
ments, paleoecologiccd trend assessments, and
remote sensing.
Many natural resource agencies throughout the
country have begun the process of developing
and implementing biological assessment and
criteria programs primarily for rivers and
streams. This document is part of the effort to
i
-------
This document is intended to
provide managers and field
biologists with functional
methods and approaches
for bioassessment and
biocriteria.
advance the use of these strategies with regards
to lakes and reservoirs, thereby fostering the
development of credible and practical
bioassessment programs.
The goal of this guidance is to assist in protect-
ing the ecological integrity of the Nation's lake
and reservoir resources. It
does not address issues of
human health assessments
as these concerns are
widely discussed in other
technical documents and
regulations. This guidance
was developed through the
experience of existing state,
regional, and national lake
monitoring programs.
Several existing lake
programs are used as case studies and ex-
amples throughout the document illustrating
specific concepts or methods. It is important to
remember that circumstances vary throughout
the country and this document cannot specifi-
cally address every situation or experience.
The orientation of this document is toward
practical decision making rather than research
and its primary target audiences are state and
tribal natural resource agencies. It is intended
to provide managers and field biologists with
functional methods and approaches that will
facilitate the implementation of viable lake
bioassessment and biocriteria programs that
meet their needs and resources.
The methods, or protocols, presented here are
organized in a tiered framework, ranging from
trophic state surveys to more detailed
bioassessment, allowing users flexibility in
designing programs appropriate to their needs
and resources. Procedures for program design,
reference condition determination, field
biosurveys, biocriteria development and data
analysis are detailed. In addition, the document
provides information on the application and
effectiveness of lake bioassessment to existing
EPA and state/tribal programs such as the
Clean Lakes Program, 305(b) assessments,
NPDES permitting, risk assessment, and water-
shed management. The appendices of the
document include a glossary of terms, summa-
ries of existing programs and protocols, detailed
descriptions of biological assemblages, and
procedures for statistical analysis of biological
data.
The following is a summary of the information
contained in each chapter:
Chapter 1: The Protection of Biological
Integrity
This chapter introduces biological integrity,
bioassessment and biocriteria as fundamen-
tal considerations in developing and imple-
menting lake monitoring programs and
discusses the relationship between these
concepts and the Clean Water Act's goal of
restoring and protecting the Nation's water
resources. Chapter 1 provides a rationale for
biomonitoring as an integral component of
natural resource agency lake management
and protection programs.
Chapter 2: Lake Biological Monitoring in
USEPA, Local, State, Tribal, and Regional
Protection and Management Programs
Monitoring is a vital element in natural
resource protection programs. Chapter 2
summarizes the relationship of biological
surveys and biocriteria to various programs
in the Clean Water Act. The application of
lake biomonitoring and the development of
biocriteria in these programs play a critical
role and can have significant benefits for
natural resource agencies and their con-
stituents. This chapter addresses where and
how biomonitoring and biocriteria fit into
these programs. In addition, this chapter
explores some nonregulatoiy applications
and benefits of biomonitoring programs.
Chapter 3: Overview of Bioassessment
and Biocriteria
This chapter provides a sketch of the
conceptual framework, application and
approaches of bioassessment and biocriteria
that are detailed in the remaining chapters.
Chapter 4: Selection and Characterization
of Reference Conditions
Establishing reference conditions, which
represent the best attainable conditions for
lakes in a given region, lays the groundwork
for the development of biomonitoring and
biocriteria programs. The ecological health
of a lake, as measured through biosurveys,
is evaluated through comparison to the
reference conditions. This chapter recom-
mends and details an approach for designat-
ing and identifying reference conditions.
Chapter 5: Habitat Measurement
The evaluation of habitat provides essential
clues as to the status of a lake's biological
organisms. Chapter 5 discusses habitat,
including both watershed and in-lake
components, as an element of
bioassessment programs.
li
-------
Chapter 6: Biological Assemblages
This chapter describes the various biological
organisms that are surveyed in lake
bioassessment programs. Target assemblages
were chosen primarily based on their ability to
be sampled and analyzed in a cost-effective way
and their use in existing programs.
Chapter 7: Tiered Sampling
Chapter 7 details an additive tiered approach to
lake biosurveys that includes evaluation of
habitat and biological assemblages, or organ-
isms. The purpose of the tiered approach Is to
provide natural resource agencies a menu of
assessment and protocol options that take Into
consideration varying levels of familiarity with
biosurveys, regional needs, resource limitations,
and regulatory requirements.
Chapter 8: Index Development
The final step toward functional bioassessment
is the development of an index, comprised of the
sum of a series of metrics, or measurement
scores. The total index value of a test site is then
compared to the index value for the reference
condition. Chapter 8 provides an overview of
procedures involved In selecting appropriate
measurements and determining an index. The
Tennessee Valley Authority's experience In
developing metrics and Indices is highlighted in
this chapter as an example. (Appendix E pro-
vides more detailed discussions and examples of
statistical methods used in data analysis.)
Chapter 9: Quality Assurance
This chapter discusses the various factors to
consider in ensuring the reliability of monitoring
and measurement data. Chapter 9 addresses
quality assurance and control considerations for
each step of the process including sampling
design, field operations, laboratory operations,
data analysis, and data reporting.
Chapter 10: Biocriteria Implementation
Chapter 10 discusses the characteristics of
biocriteria and details the steps to implement a
biocriteria program. Biocriteria provide natural
resource agencies with a mechanism to protect
the biological integrity of lakes and to establish
aquatic life-use classifications. Issues of focus in
this chapter include technical and resource
considerations.
Appendix A: Glossary of Terms
Appendix B: Comparison of Existing Lakes
Protocols
Appendix C: Paleolimnologieal Sampling
Appendix D: Biological Assemblages
Appendix E: Statistical Analysis Methods for
Biological Assessment
Appendix F: Executive Summaries of State
Pilot Studies
Appendix G: Literature Cited
-------
In This Chapter...
> The Relationship Between Bioassessment
and Biocriteria
> Uses of Bioassessment and Biocriteria
Chapter 1
The Protection of Biological
Integrity
1.1 INTRODUCTION
Biological monitoring is integral to the measure-
ment of the total ecological health of a
waterbody and is becoming increasingly impor-
tant in water quality monitoring and assess-
ment. Historically, most natural resource
programs have measured individual pollutants
in the water column and sediments. Although
such programs have effectively monitored and
controlled point source discharges of nutrients
and contaminants, their efforts to assess total
ecological integrity, measured by combined
chemical, physical (including habitat), and
biological attributes, have been limited. Many
surface waters have continued to deteriorate
from nonpoint pollution, habitat modification,
and other impacts of human activities (Karr
1991). For example, in the United States, the
total lake acreage that deteriorated in quality
from 1972 to 1982 was four times the acreage
that improved (Johnson 1989).
This document describes a set of protocols for
biological assessment of lakes and reservoirs
relevant to issues of ecological integrity. It is not
intended to address human health concerns as
these issues have been addressed in previous
Around the country, various agencies use terms
differently. This can lead to confusion when •.
developing a guidance document intended for
national use. Therefore, for the purposes; of this
document, the following terms are defined:
A biological survey (blosurvey) is the process of
collecting and processing representative portions
of a resident aquatic community to determine the
community structure and function.
A biological assessment (bioassessment) is an
evaluation of the biological condition of a
waterbody that uses biosurveys and other direct
measurements of resident biota in surface
waters.
Biological monitoring (biomonitoring) is the use
of a biological entity as a detector, and its re-
sponse as a measure, to determineenvironmental
conditions. Toxicity tests and biosurveys are
common biomonitoring methods.
Biological criteria (biocriteria) are numeric :k:
values or narrative expressions that describe the
reference biological condition of aquatic communi-
ties inhabiting w&ter;s of a given designated
aquatic life use.
1-1
-------
Chapter 1
Biological monitoring is
guidance documents. The protocols in this
document are intended for use by local, state,
tribal, and regional natural resource monitoring
agencies, and they can be used in the imple-
mentation of biological criteria.
The document includes a general strategy for
biocriteria development, identifies steps in the
process, and provides technical guidance on
how to complete each step,
using the experience and
knowledge of existing state,
integral to the measure- regional, and national
surface water programs
ment of the total ecologi- where appropriate. The
protocols are tiered to allow
cai health of a waterbody flexibility in customizing
individual monitoring
programs according to the
ingly important in water user's own requirements
and available resources.
quality monitoring and
assessment.
and is becoming increas-
The multiple assemblage
and multimetric assess-
ment approach outlined
here is designed to address elements and
processes associated with community balance,
trophic structure, and richness. This guidance
is not intended to replace existing biological
assessment or biocriteria programs. Rather, it
can be used as a tool for developing new pro-
grams and/or enhancing current programs.
Although not designed to "push the envelope" of
lake bioassessment, this document was devel-
oped to provide methods that are technically
credible, practical, and geared toward the
genuine needs and resources of natural re-
source agencies.
1 .2 THE CONCEPT OF
BIOCRITERIA
Efforts to monitor human effects on waterbodies
have ranged from 19th century physical obser-
vations of sediment and debris movement
(Caper et al. 1983) to chemical metrics, cur-
rently the most commonly employed source of
water quality criteria (USEPA 1992e). Investiga-
tors and resource managers, however, have long
recognized that water column measurements
reflect conditions only at the time of sampling.
As an important supplement to chemical sam-
pling, biological measurements can reflect both
current conditions and temporal changes in
waterbodies, including the cumulative effects of
successive disturbances. However, the develop-
ment and widespread use of formal biological
criteria have lagged behind chemical-specific,
instream flow, and toxicity-based water quality
criteria in waterbody management (USEPA
1985b, USEPA 1985c). Recent recommendations
on monitoring strategies for aquatic resources
have emphasized the need to accelerate the
development of biological sampling as a regular
part of surface water programs (USEPA 1987b,
USEPA 1987c).
Biological criteria are benchmarks for water
resource protection and management decision
making. Expressed as numeric values or narra-
tive expressions, they measure attainment of
biological integrity. In turn, biological integrity
describes the most robust aquatic community to
be expected in a natural condition in a water
resource relatively unaffected by human activities.
The development of biocriteria by natural
resource agencies depends on the assessment of
conditions at reference sites. Reference sites are
not necessarily pristine, although they must
exhibit only minimal impairment relative to the
overall region of study. Based on biological
sampling, or surveys, a bioassessment of
multiple sites is done, resulting in values that
represent the biological potential for waters in
the region. The regional biological potential is
then used to establish biocriteria. Biocriteria
can then be used as a measuring stick for
determining the status of test sites. The sites
can be surveyed, scored, and compared to the
established biocriteria.
Biocriteria supported by bioassessment serves
several purposes in surface water programs. The
use of biocriteria expands and improves water
quality standards, helps identify impairment of
beneficial uses, and helps set program priorities.
The use of bioassessments to investigate impair-
ment, evaluate the severity of problems, ascertain
the causes of the problems, and determine
appropriate remedial action is a step-by-step
process. Decision criteria for ascertaining impair-
ment are part of the implementation plan and the
foundation for establishing biocriteria to deter-
mine beneficial use categories and assess subse-
quent management efforts. This should be
followed by continued monitoring, improving the
resource quality with each cycle. (See Figure 1-1.)
1-2
-------
The Protection of Biological Integrity
NITORING CRITERI
Figure 1-1. Interdependence of Environmental
Monitoring and Environmental Criteria.
1 .3 USES OF
BIOASSESSMENTS AND
BIOCRITERIA
By directly measuring the condition of the water
resource at a site, surveys and assessments of
resident biota are an important foundation in
the derivation and maintenance of biocriteria
and, thus, are a critical tool for natural resource
agencies in protecting the quality of water
resources. Biocriteria, in conjunction with
surveys of aquatic assemblages, are useful for a
variety of purposes including:
• Problem screening and identification,
through the early detection of problems that
other methods might fail to uncover or
might underestimate,
• Assessing the effectiveness of implemented
water resource management practices,
• Determining attainment of designated
aquatic life uses.
• Refining aquatic life uses categories.
• Identifying impact sources.
Applications of bioassessments and biocriteria
to specific USEPA, state, local, tribal, and
regional management programs (such as under
Clean Water Act sections 303, 305(b), 314, 319)
are discussed in Chapter 2 of this document.
1.3.1 Screening and
Identifying Problems
Monitoring of the resident biota can be used to
identify and rank problem areas for further
attention and dedication of resources. It can
also serve as an early warning system to identify
problems and to ensure against continued
degradation.
Biological assessments can be used to establish
priorities for remedial actions. Screening can be
done on an individual lake to establish manage-
ment priorities. Screening can also be used as a
tool on a regional or statewide basis to deter-
mine programmatic
priorities. For example,
regional screening could
determine whether nutri-
ent controls, sediment
controls, or toxic elimina-
tion should have the
highest priority for im-
proving regional surface
water quality.
Monitoring of the resident
biota can be used to identify
and rank problem areas for
further attention and
dedication of resources.
1.3.St Assessing Effectiveness
of Management
Practices
Bioassessments can be used to track the
effectiveness of remediation measures. In
managing nonpoint source pollution, the natu-
ral resource agency may initiate cooperative
land use programs in a given area or install best
management practices
(BMPs) to improve the
water resource. Both
Nonpoint Source (NPS)
and Clean Lakes Pro-
grams require monitoring
of BMPs. Before-and-after
bioassessments compared
to the biocriteria "bench-
mark" make it possible to
objectively evaluate the
relative success of man-
agement by assessing
actual biological commu-
nity changes.
Before-and-after bioas-
sessments compared to
the biocriteria "benchmark"
make it possible to objec-
tively evaluate the relative
success of management by
assessing actual biological
community changes.
While other management uses of biocriteria
include reviewing the adequacy of NPDES
permits, biocriteria are not recommended at this
time for inclusion as NPDES permit limits.
1-3
-------
Chapter 1
Rather, they are Ideal for assessing the ad-
equacy of the permit to protect the resident
biota. This can be done by comparing biosurvey
results at the test site to the criteria established
for that waterbody. Failure to meet the criteria
suggests that the waterbody is not meeting Its
aquatic life use. One possible explanation is that
the permit is not protective enough for the use
class.
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
relative improvement as a result of mitigation.
Several natural resource agencies have estab-
lished monitoring stations for conducting
periodic biosurveys in streams as part of their
biomonitoring programs. Very few natural
resource agencies have initiated biological
assessment for compliance monitoring in lakes.
1.3.3 Refining Aquatic Life
Uses
Both classification and definition of designated
uses of lakes and reservoirs are important in the
planning, development, and use of biocriteria.
Historical data from existing state efforts such
as surface water classifica-
tion and Clean Lakes
Programs, along with
to biological resource additional field efforts, aid
completion of these key
condition rather than planning steps. Informa-
tion obtained through
biological surveys can be
of particular pollutants. used to explicitly describe
each designated use.
Biocriteria relate directly
surrogate concentrations
A designated use is a classification designated
in state water quality standards for each
waterbody or segment that defines the optimal
purpose for that waterbody regardless of attain-
ment status. The designated uses for lakes and
reservoirs are usually defined by individual
natural resource agencies and include such
uses as drinking water, aquatic life, recreational
use, industrial use, and agricultural use.
Use attainability—The potential for a waterbody
to meet, reach, or develop to its optimal purpose
or designated use.
Aquatic life uses—Classifications specified in
state water quality standards for each
waterbody or segment relating to the level of
protection afforded to the resident biological
community by the state agency.
General information on use designation can be
found in Biological Criteria: National Program
Guidance for Surface Waters (USEPA 1990a).
Specific technical guidance for conducting
use-attainability analyses is provided in Techni-
cal Support Manual: Waterbody Surveys and
Assessments for Conducting Use Attainability
Analyses (USEPA 1984).
Designated uses of waterbodies are formulated
on, and in turn influence, the level of protection
afforded the aquatic resource. Natural resource
agencies establish standards appropriate to the
protection of specific designated uses. For
example, the designation outstanding waters is
sometimes assigned to waterways which are
located in undisturbed or minimally influenced
watersheds and are characterized by aquatic
communities that are deemed to be as naturally
occurs (USEPA 1990a). Alternatively, other use
designations may reflect preexisting land use
patterns that prevent attainment of the highest
quality waters. However, an observed downward
trend does not justify lowered use designation.
1.3.4 Determining Attainment
of Designated Use
Biological surveys and criteria are fundamental
tools for assessing aquatic life use Impairment.
State water quality standards exist to define and
protect designated uses conducive to overall
water resource enhancement and preservation.
Current biomonitoring tools used to judge
nonattainment are not well-formulated in many
Instances. Consequently, many natural resource
agencies rely exclusively or primarily on
chemical-specific criteria to evaluate use impair-
ment.
Biocriteria provide the only direct assessment of
resource condition, and they are sensitive to a
broader range of human influences on the
watershed than are chemical criteria alone (Karr
1991, USEPA 1991b). By including biocriteria, a
natural resource agency gains a much more
complete assessment of the condition of the
water resource. Biocriteria relate directly to
biological resource condition rather than surro-
gate concentrations of particular pollutants.
1-4
-------
The Protection of Biological Integrity
Cumulative impacts on the biota can be mea-
sured, revealing synergistic degradation that may
occur even though all specific permit conditions
may be met. Similarly, this measure of the biotic
community often reveals the sum total of effects
over the entire year, not just at one point in time.
1.3.5 Identifying Causes of
impairment
The concept of measuring the attributes of
aquatic communities in unimpacted areas for
biocriteria was first developed for stream sys-
tems (Index of Biotic Integrity [IBI], Karr et al.
1986; Invertebrate Community Index IICI], Ohio
EPA 1987; Rapid Bioassessment Protocol [RBPJ,
USEPA 1989b) Observed deviations from the
unimpacted conditions are assumed to be
indicative of impairment. Human-induced
alterations affect biological integrity through
their effects on five major classes of factors
important to the aquatic biota (adapted from
Karr et al. 1986):
• Energy base.
• Chemical constituents.
• Habitat structure.
• Hydrologic regimen.
• Biotic interactions.
These factors influence the aquatic biota and
can adversely affect elements and processes that
normally occur in a lake or reservoir. By specifi-
cally designing a survey to include all five of
these elements, it is possible to address causal-
ity when a lake fails to meet its biocriteria. Such
information will assist in diagnosing impaired
sites and determining management actions, for
example, distinguishing between impacts from
toxic substances and disruption of habitat.
1.4 OTHER BIOCRITERIA
REFERENCE DOCUMENTS
USEPA has developed technical guidance
documents for implementing biocriteria in
response to biocriteria development issues
including legislative authority, steps in develop-
ing biocriteria, and the application of biocriteria
to surface water management (USEPA 1990a). A
reference guide to the technical literature
pertaining to biocriteria has been developed to
provide support interest from natural resource
agencies (Table 1-1). This reference guide
contains cross-references to technical papers
that present concepts, approaches, and proce-
dures necessary to implement habitat evalua-
tions and biological surveys in the development
and use of biocriteria.
In December 1990, a symposium on biological
criteria provided a forum for discussing techni-
cal issues and guidance for the various
waterbody types of the Nation's surface waters.
The proceedings at this conference are pre-
sented in USEPA (1991b). The Agency has also
developed guidance to help natural resource
agencies initiate narrative biological criteria
(USEPA 1992e).
Recently, the Agency issued a technical guid-
ance documentor biocriteria use in streams
and small rivers (USEPA 1996a). Much of the
approach and many of the issues addressed by
the stream document serve as a template for
developing biocriteria for other waterbody types,
including lakes.
1-5
-------
Chapter 1
Table 1-1. Biocriteria reference documents.
Title
Document Citation
Biological Criteria: National Program Guidance for
Surface Waters.
USEPA 1990. EPA-440/5-90-004. U.S.
Environmental Protection Agency, Washington, DC.
Biological Assessment Methods, Biocriteria, and
Biological Indicators: Bibliography of Selected
Technical, Policy, and Regulatory Literature.
USEPA 1996. EPA-230-B-96-001. Office of Policy,
Planning, and Evaluation, U.S. Environmental
Protection Agency, Washington, DC.
Biological Criteria: Research and Regula tion.
Proceedings of a Symposium.
USEPA 1991. EPA-440/5-91 -005. U.S.
Environmental Protection Agency, Office of Water,
Health and Ecological Criteria Division,
Washington, DC.
Procedures for Initiating Narrative Biological
Criteria.
USEPA 1992. EPA-822-B-92-002. U.S.
Environmental Protection Agency, Office of Science
and Technology, Washington, DC.
Biological Criteria: Technical Guidance for Streams
and Small Rivers.
USEPA 1996. EPA-822-B-96-001. U.S.
Environmental Protection Agency, Office of Science
and Technology, Washington, DC.
Summary of State Biological Assessment Programs
for Streams and Rivers.
USEPA 1996. E PA-230-R-96-007. Office of Policy,
Planning and Evaluation, U.S. Environmental
Protection Agency, Washington, DC.
1-6
-------
/#? ThSs Chapter...
> Programmatic Applications of Biological Data
Chapter 2
Lake Biological Monitoring
in USEPA, Local, State, Tribal, and Regional Protection
and Management Programs
A state monitoring program is the source of data
for all other state resource management pro-
grams. It helps to identify water quality prob-
lems, identify waters needing total maximum
daily loads (TMDLs), quantify loads, verify
models, and evaluate the effectiveness of point
and nonpoint source water quality controls. A
state's monitoring program also serves as the
backbone of its water quality programs. The
biological monitoring protocols presented in this
guidance document will strengthen a state's
monitoring program. An effective and thorough
water quality program can help to improve
reporting (e.g., 305(b) reporting), increase the
effectiveness of pollution prevention efforts, and
document the progress of mitigation efforts.
Biological monitoring and the establishment of
biocriteria provide scientifically sound and
detailed descriptions of designated aquatic life
use for waterbodies. Biocriteria are biological
benchmarks for measuring the condition of
aquatic biota. They help determine whether
water quality goals are attained, set priorities,
and evaluate the effectiveness of implemented
controls and management actions. Developing
and implementing biocriteria for lakes and
reservoirs is complicated In some states because
of a high level of human intervention on a
significant percentage of lakes and reservoirs.
Many lakes and reservoirs are managed by the
states for different uses
(e.g., drinking water,
recreation, fishing). This section provides
Several lake management
practices mask natural suggestions for the
conditions; for example, application of biological
stocking of fish and
periodic lowering of lake monitoring and biological
levels. In addition, entire
regions of the country criteria to lakes and
have no natural lakes but . _
reservoirs through existing
state programs.
have abundant reservoirs,
which do not have the
same attributes as most
natural lakes.
Despite the variability in lake conditions,
performing biological monitoring and developing
biocriteria for lakes have important benefits.
This section provides suggestions for the appli-
cation of biological monitoring and biological
criteria to lakes and reservoirs through existing
state programs (Table 2-1).
2-1
-------
Chapter 2
Tablo 2-1. Applications of lakes biological monitoring protocols and biocriteria.
Program
Biological Monitoring and Assessment
Biological Criteria
Section 305(b)
Reporting
• Improving data for beneficial use
assessment.
• Improving water quality reporting.
• Identifying waters that are
not achieving their aquatic
life use support.
• Defining an understandable
endpoint in terms of
"biological health" or
"biological integrity" of
waterbodies."
Section 314/Clean
Lakes Program
• Assessing status of biological components
of lake systems.
• Measuring effects of ongoing restoration
projects.
• Measuring success of lake clean-up efforts
and other mitigation activities.
• Assessing lake trophic status and trends
assessing biological trends.
[Monitoring and sampling needs vary for
each lake]
[Clean Lakes Program Regulations
monitoring components: algal pigments,
aigal genera, cell densities, algal cell
volumes, limiting nutrients, macrophyte
coverage, bacteria, and fish flesh analysis]
« Identifying lakes that are not
attaining designated use
(including aquatic life use)
support.
• Defining lake biological
integrity based on a
reference condition.
• Identify impairments due to
toxic substances.
Section 319/Nonpoint
Source Program
• Evaluating nonpoint source impacts and
sources.
• Measuring site-specific ecosystem
response to remediation or mitigation
activities.
• Assessing biological resource trends within
watersheds.
• Determining effectiveness of
nonpoint source controls.
Watershed Protection
Approach
* Assessing biological resource trends within
watersheds.
• Setting goals for watershed
and regional planning.
TMDLs
• Identifying biological assemblage and
habitat impairments that indicate
nonattainment of water quality standards.
• Documenting ecological/water quality
response as a result of TMDL
implementation.
• Priority ranking waterbodies.
• Identifying water
quality-limited waters that
require TMDLs.
• Establishing endpoints for
TMDL development, i.e.,
measuring success.
NPDES Permitting
• Measuring improvement or lack of
improvement of mitigation efforts.
* Developing protocols that demonstrate
relationship of biological metrics to effluent
characteristics.
• Performing aquatic life use
compliance monitoring.
« Helping to verify that
NPDES permit limits are
resulting in achievement of
state water quality standard.
2-2
-------
Lake Bio/ogfcaf Monitoring
Table 2-1. Applications of lakes biological monitoring protocols and biocriteria (continued).
Program
Biological Monitoring and Assessment
Biological Criteria
State Monitoring
Programs
• Improving water quality reporting.
• Documenting improvement or lack of
improvement of mitigation efforts including
lake clean-up efforts, TMDL application,
NPDES efforts, nonpoint source pollution
controls, etc,
• Problem identification and trend
assessment,
• Prioritizing waterbodies.
• Measuring effectiveness of
controls.
• Performing watershed
planning.
• Performing regional
planning.
Risk Assessment
• Providing data needed to estimate
ecological risk to assessment endpoints.
• Development of an
assessment or
measurement endpoint.
Water Quality Criteria
and Standards
• Developing data bases for lake
phytopiankton, macroinvertebrates, fish
plants, and other assemblages.
• Developing indices that assess lake biota
compared to reference.
* Identifying waterbodies that
are not attaining aquatic life
use support.
* Refining aquatic life use
classifications.
* Developing site-specific
standards.
2.1 SECTION 305(B) WATER
QUALITY ASSESSMENT
Section 305(b) establishes a process for report-
ing information about the quality of the Nation's
water resources (USEPA 1993c, USEPA 19941).
States, the District of Columbia, territories, and
certain River Basin Commissions have devel-
oped programs to monitor surface and ground
waters and to report the current status of water
quality biennially to USEPA. Special grants are
available for Native American groups to provide
similar assessments of water quality on tribal
lands. This information is compiled into a
biennial National Water Quality Inventory report
to Congress. The 305(b) reports are a major data
source helping USEPA to:
• Determine the status of water quality. (Are
the designated/beneficial uses being met?)
• Evaluate the causes of poor water quality
and the relative contributions of pollution
sources.
• Report on the activities under way to assess
and restore water quality.
• Determine the effectiveness of control
programs.
Section 305(b) establishes
• Determine the workload remaining in
restoring waters with poor quality and
protecting threatened waters.
Use of biological assessment in 305(b) reports
helps to define an understandable endpoint of
relevance to society—the
biological health and
integrity of waterbodies.
Many of the better known a process for reporting
and widely reported
pollution cleanup success information about the
stories have involved the
recovery or reappearance quality of the Nation s
of valued sport fish and
other pollution intolerant
species to systems from
which they had disappeared (USEPA 1980b,
USEPA 1985a). Improved coverage of biological
integrity issues, based on monitoring protocols
with clear bioassessment endpoints, will make
the 305(b) reports more accessible and mean-
ingful to many segments of the public.
The 305(b) process encourages monitoring and
assessment for all lakes. The Clean Water Act
Section 314 Clean Lakes Program outlines
specific assessment or classification information
for significant publicly owned lakes. Section
water resources.
2-3
-------
Chapter 2
314(a)(2) of the CWA, as amended by the Water
Quality Act of 1987, requires the states to
submit a biennial assessment of their lake water
quality as part of their 305(b) reports (USEPA
1993c). The specific elements of the assessment,
as outlined in section 314(a)(10)(A-F), constitute
the minimal requirements for approval and for
subsequent grant assistance as required by
section 314(a)(4). Each state report should
reflect the status of lake water quality in the
state, restoration/protection efforts, and trends
in lake water qualify. Each state should report
the total number of significant publicly owned
lakes and their acreage, the trophic status of
each lake, control methods, restoration and
rehabilitation efforts, the number of impaired
and threatened lakes, acid effects on lakes, toxic
effects on lakes, trends in lake water quality,
and a description of the state's water quality
standards that are applicable to lakes.
Biological monitoring can provide data that
could augment several of the 305(b) reporting
requirements. In particular, the following lake
assessment activities and reporting require-
ments could be enhanced through the use of the
biological monitoring information:
• Measuring the success of restoration and
rehabilitation efforts when measured against
reference conditions.
• Measuring the success of Clean Lakes
Program projects.
• Developing and using lake water quality
standards or, if water quality standards
have not been developed for lakes, develop-
ing and using other biological measures to
determine impaired or threatened status of
lakes.
States are encouraged to
develop Integrated water
quality strategies that
Include lake and reservoir
management, restoration,
and protection activities.
Identifying lakes and
lake acres affected by
acidify or toxics and
those with elevated
levels of toxics.
Identifying sources of
acidify and toxic
pollutants in lakes
and estimating the
number of affected
lake acres attributed
to each source.
• Identifying lake water quality trends,
including trends in acidity, toxic pollutants,
and their effects.
The Waterbody System (WBS) can generate
many of the tables needed to report the required
305(b) summary data (USEPA 1994f). The
Waterbody System can record general informa-
tion on the types of monitoring protocols used in
making assessments for specific lakes. Since
WBS is intended as a data base of assessments,
it does not have facilities for storing actual
monitoring data or bioassessment metrics.
Bioassessment information could, however, be
entered in WBS comment fields.
2.2 SECTION 314 CLEAN
LAKES PROGRAM
Historically, the Clean Lakes Program has been
active in awarding grants for the study and
restoration of publicly owned lakes. Under this
program, states are encouraged to develop
integrated water quality strategies that include
lake and reservoir management, restoration,
and protection activities.
The Clean Lakes Program regulations (40 CFR
part 35, subpart H) list the primary components
that could be monitored to characterize the
biological component of a lake system, including
algal pigments, algal genera, cell densities, algal
cell volumes, limiting nutrients, macrophyte
coverage (by species), bacteriological compo-
nents, and fish flesh analysis. The regulations
do not specifically require monitoring for
macroinvertebrates. Whether a complete limno-
logical investigation or some more focused set of
investigations should be undertaken depends on
the status of available baseline data and the
problems affecting a particular lake.
Monitoring and sampling needs vaiy from lake
to lake. For example, a lake program might do a
more detailed benthic macroinvertebrate survey
if dredging or restoration work involving the
disturbance of sediments is planned. Even if
this survey work is being done for dredging
purposes only, it can aid in the formulation of
an on-site reference. The use of a reference
condition, whether it is developed by historical
data or through a regional approach, can
improve Clean Lakes projects by identifying
biological impairments that were previously
2-4
-------
Lake Biological Monitoring
unknown or not adequately documented based
on chemical and physical monitoring data alone.
In particular, biological monitoring will provide
data to help accomplish the following:
• Determine the success of restoration and
rehabilitation efforts when measured against
reference conditions,
• Better characterize the biological component
of the lake system.
• Measure aquatic life use support.
• Develop and use lake water quality stan-
dards, or develop and use other biological
measurements to determine impairment or
threatened status of lakes.
• Develop and update lake management
plans.
All of the activities listed above can be partially
achieved through the use of biological monitor-
ing protocols in lake programs. They will lead to
improved data for assessing beneficial uses and
for improving both 305{b} and other grant
reporting requirements.
2.3 SECTION 319 NONPOINT
SOURCE PROGRAM
The 1887 Water Quality Act Amendments to the
Clean Water Act added section 319, which
established a national program to control
nonpoint source (NPS) pollution. States assess
their NPS pollution problems and submit these
assessments to USEPA. The assessments
include a list of "navigable waters within the
state which, without additional action to control
nonpoint sources of pollution, cannot reason-
ably be expected to attain or maintain appli-
cable water quality standards or the goals and
requirements of this Act." Other activities under
the section 319 process require the identifica-
tion of categories and subcategories of NPS
pollution that contribute to the identification of
impaired waters, descriptions of the procedures
for identifying and implementing BMPs, control
measures for reducing NPS pollution, and
descriptions of state and local programs used to
abate NPS pollution. Based on the assessments,
states have prepared nonpoint source manage-
ment programs, and USEPA grants are now
available to assist in the implementation of
approved state programs.
Biological assessment techniques can improve
evaluations of nonpoint source pollution con-
trols (or the combined effectiveness of current
point and nonpoint source
controls) by comparing States assess their NPS
biological integrity indica-
tors before and after pollution problems and
implementation of con-
trols. Likewise, biocriteria
can be used to measure to USEPA The assgssmgnts
site-specific ecosystem
response to remediation or include a list of "navigable
mitigation activities aimed
at reducing nonpoint wafers within the state
source pollution impacts
or response to pollution
prevention activities. acf/on to control nonpoint
submit these assessments
which, without additional
sources of pollution, cannot
reasonably be expected to
attain or maintain applicable
water quality standards or
the goals and requirements
of [section 319 of the Clean
Water Act]"
Several section 319
projects involve lake
restoration (USEPA
1994f). Currently,
biocriteria have not been
developed for these lakes,
but their use would
greatly improve the ability
of lake managers to focus
their efforts. By providing
a measuring tool,
biocriteria can be key in
Identifying the most
significant sources of a lake's pollutants.
Minimum lake monitoring guidance for nonpoint
source pollution assessment is being developed
and will include biological protocols for lakes.
2.4 WATERSHED PROTECTION
APPROACH
Since 1991, USEPA has been promoting the
Watershed Protection Approach as a framework
for meeting the Nation's remaining water re-
source challenges (USEPA 1994k). The agency's
Office of Water has taken steps to reorient and
coordinate point source, nonpoint source, lakes,
wetlands, coastal, ground water, and drinking
water programs in support of the watershed
approach, USEPA has also promoted multi-
organizational, multi-objective watershed
management projects across the Nation.
2-5
-------
Chapter 2
The watershed approach is an integrated,
holistic strategy for more effectively protecting
and managing surface water and ground water
resources and achieving broader environmental
protection objectives using the naturally defined
hydrologic unit (the watershed) as the integrat-
ing management unit. Thus, for a given water-
shed, the approach encompasses not only the
water resource, such as a stream, river, lake,
estuary, or aquifer, but all the land from which
water drains to the resource. The watershed
approach places emphasis on all aspects of
water resource quality: physical (e.g., tempera-
ture, flow, mixing, habitat); chemical (e.g.,
conventional and toxic pollutants such as
nutrients and pesticides); and biological (e.g.,
health and integrity of bio tic communities,
biodiversity).
The Clean Lakes Program (CLP) has been an
important model for the Watershed Protection
Approach and ecosystem
The watershed approach is management (USEPA
1994k). The CLP has
an Integrated, holistic been referred to as the
quintessential watershed
strategy for more effec- program because it has
taken a holistic, place-
based approach that uses
managing surface water sound science, involves
stock holders, and forms
and ground water re- partnerships for compre-
hensive, integrated action
sources and achieving to protect ^ restore
lake resources in the
Nation. A newly devel-
protection objectives using oped Clean Lakes Pro-
gram framework calls for
the naturally defined better integration of the
CLP with nonpoint
hydrologic unit (the source water quality
management, permitting,
and other ecosystem
ing management unit. protection activities.
tively protecting and
broader environmental
watershed) as the integrat-
es SECTION 303(D)
THE TMDL PROGRAM
The technical backbone of the Watershed
Protection Approach is the process for total
maximum daily loads (TMDL). TMDLs is a tool
used to achieve applicable water quality stan-
dards. The TMDL process quantifies the loading
capacity of a waterbody for a given stressor and
ultimately provides a quantitative scheme for
allocating loadings (or external inputs) among
pollutant sources (USEPA 1994c). In doing so,
the TMDL quantifies the relationships among
sources, stressors, recommended controls, and
water quality conditions. For example, a TMDL
might mathematically show how a specified
percent reduction of a pollutant is necessaiy to
reach the pollutant concentration reflected in a
water quality standard.
Section 303(d) of the CWA requires each state to
establish, in accordance with its priority
rankings, the total maximum daily load for each
waterbody or reach identified by the state as
failing to meet or not expected to meet water
quality standards after imposition of technology-
based controls.
In addition, TMDLs are vital elements of a
growing number of state programs. For example,
as more permits incorporate water quality-based
effluent limits, TMDLs are becoming an increas-
ingly important component of the point source
control program.
TMDLs are suitable for nonchemical as well as
chemical stressors (USEPA 1994c). These
include all stressors that contribute to the
failure to meet water quality standards, as well
as any stressor that presently threatens but
does not yet impair water quality. TMDLs are
applicable to waterbodies impacted by both
point and nonpoint sources. Some stressors,
such as sediment deposition or physical alter-
ation of instream habitat, might not clearly fit
traditional concepts associated with chemical
stressors and loadings. For these nonchemical
stressors, it might sometimes be difficult to
develop TMDLs because of limitations in the
data or in the technical methods for analysis
and modeling. In the case of nonpoint source
TMDLs, another difficulty arises in that the CWA
does not provide well-defined support for
regulatory control actions as it does for point
source controls, and controls based on another
statutory authority might be necessary.
Because they directly measure the aquatic
community's response to pollutants or stressors,
biological surveys can provide compelling
evidence of water quality impairment. Biological
assessments and criteria address the cumula-
tive impacts of all stressors, especially habitat
degradation, loss of biological diversity, and
nonpoint source pollution. Biological informa-
2-6
-------
Lake Biological Monitoring
tion can help provide an ecologically based
assessment of the status of a waterbody and
thus can be used to decide which waterbodies
need TMDLs (USEPA 1993c).
Incorporation of bioasscssment data aids in the
ranking process to target waters for TMDL
development by allowing more accurate
prioritization because of the direct link between
bioassessment and ecological integrity (i.e., the
condition of an unimpaired ecosystem as
measured by combined chemical, physical, and
biological attributes of surface waters (Barbour
etal. 1992).
Finally, the TMDL process is a geographically
based approach to preparing load and wasteload
allocations for sources of stress that might
impact waterbody integrity. The geographic
nature of this process will be complemented and
enhanced if ecological regionalization is applied
as part of the bioassessment activities. Specifi-
cally, similarities among ecosystems can be
grouped into ecoregions. The ecoregion concept
provides a geographic framework for more
efficient aquatic resource management.
2.6 SECTION 402
NRDES PERMITS AND
INDIVIDUAL CONTROL
STRATEGIES
All discrete sources of wastewater must obtain a
National Pollutant Discharge Elimination
System (NPDES) permit, which regulates the
facility's discharge of pollutants. The approach
to controlling and eliminating water pollution is
focused on the pollutants determined to be
harmful to receiving waters and on the sources
of such pollutants. Authority for issuing NPDES
permits is established under section 402 of the
CWA (USEPA 1989a).
Point sources are generally divided into two
types, industrial and municipal. Nationwide,
there are approximately 50,000 industrial
sources, which include commercial and manu-
facturing facilities. Municipal sources, also
known as publicly owned treatment works
(POTWs), number about 15,700 nationwide.
Wastewater from municipal sources results from
domestic wastewater discharged to POTWs, as
well as the "indirect" discharge of industrial
wastes to sewers.
USEPA does not recom-
mend the use of biological
criteria as the basis for
deriving an effluent limit
for an NPDES permit
(USEPA 1994e). Unlike
chemical-specific water
quality criteria, biological
criteria do not measure the
concentrations or levels of
chemical stressors. In-
stead, they directly mea-
sure the impacts of any
and all stressors on the
resident aquatic biota. Because of this, biologi-
cal criteria do not definitively establish the
causal relationship between a biological impact
and its source. This is not to say that biological
criteria have no role in the permitting process,
now or in the future. Where appropriate, biologi-
cal criteria can be used for assessment purposes
within the NPDES process (USEPA 1996a). The
criteria can provide information on the status of
a waterbody where point sources might cause,
or contribute to, a water quality problem. In
conjunction with chemical water quality and
whole-effluent toxicity data, biological criteria
can be used to detect previously unmeasured
chemical water quality problems and to evaluate
the effectiveness of implemented controls.
Some states have already demonstrated the
usefulness of biological criteria under certain
circumstances to indicate the need for addi-
tional or more stringent permit limits (e.g., sole-
source discharge into a stream where there is no
significant nonpoint source discharge, habitat
degradation, or atmospheric deposition) (USEPA
1996a). In these situations, the biological
findings triggered additional investigations to
establish the cause-and-effect relationship and
to determine the appropriate limits. In this
manner, biological criteria support regulatory
evaluations and decision making. Biological
criteria can also be useful in monitoring highly
variable or diffuse sources of pollution that are
treated as point sources such as wet-weather
discharges and stormwater runoff (USEPA
1996a). Traditional chemical water quality
monitoring is not usually appropriate for these
types of point source pollution, and a biological
survey of their impact might be critical to
evaluate these discharges and treatment mea-
sures effectively.
Biological information can
help provide an ecologi-
cally based assessment of
the status of a waterbody
and thus can be used to
decide which waterbodies
need TMDLs.
2-7
-------
Chapter 2
2.7 RISK ASSESSMENT
Ecological risk assessment is defined as "The
process that evaluates the likelihood that
adverse ecological effects may occur or are
occurring as a result of exposure to one or more
stressors" (USEPA 1992c). Risk management is
a decision-making process that involves all the
human-health and ecological assessment
results, considered with political, legal, eco-
nomic, and ethical values, to develop and
enforce environmental standards, criteria, and
regulations (Maughan 1993). Ecological risk
assessment can be performed on an on-site
basis or can be geographically based (i.e.,
watershed scale) to assess risks to ecologically
valuable endpoints (USEPA 1996d).
Results of regional bioassessment studies can
be used in watershed ecological risk assess-
ments to develop regional empirical models of
biological responses to stressors. Such models
can then be used in a predictive mode, together
with predicted exposure information, to predict
risk due to stressors or to alternative manage-
ment actions. Risks to biological resources are
characterized, and sources of stress can be
prioritized. Watershed risk managers can use
such results for critical management decisions.
2.8 SECTION 303(C)
USEPA WATER QUALITY
CRITERIA AND
STANDARDS
The water quality standards program, as envi-
sioned in section 303(c) of the CWA, is a joint
effort between the states
and USEPA. The states
have primary responsibil-
bloassessment studies ity for setting, reviewing,
revising, and enforcing
can be used in watershed water quality standards.
USEPA develops regula-
tions, policies, and
ments to develop regional guidance to help states
implement the program
empirical modes of and oversees states'
activities to ensure that
biological responses to state-adopted standards
are consistent with the
requirements of the CWA
and that water quality
Results of regional
ecological risk assess-
stressors.
standards regulations (40 CFR Part 131) are
met. USEPA has authority to review and approve
or disapprove state standards and, where
necessary, to promulgate federal water quality
standards. A water quality standard defines the
water quality goals of a waterbody, or a portion
thereof, by designating the use or uses to be
made of the water, setting criteria necessary to
protect those uses, and preventing degradation
of water quality through antidegradation provi-
sions. States adopt water quality standards to
protect public health or welfare, enhance the
quality of water, and protect biological integrity.
Environmental stressors can be chemical,
physical, or biological in nature, and likewise can
impact the chemical, physical, and biological
characteristics of an aquatic ecosystem. For
example, the impact of a chemical stressor might
be observed in impaired functioning or loss of a
sensitive species and a change in community
structure. The impact of a biological stressor,
such as an introduced species, can result in a
change in community structure through competi-
tion, predation, etc. Ultimately, the number or
intensity of all stressors within an ecosystem will
be evidenced by a change in the condition and
function of the biotic community. The interac-
tions among chemical, physical, and biological
stressors and their compounding impacts
emphasize the need to directly detect and assess
actual water quality impairments of the biota.
Sections 303 and 304 of the CWA require states
to protect biological integrity as part of their
water quality standards. This can be accom-
plished, in part, through the development and
use of biological criteria. As part of a state or
tribal water quality standards program, biologi-
cal criteria can provide scientifically sound and
detailed descriptions of the designated aquatic
life use for a specific waterbody or segment.
They fulfill an important assessment function in
water quality-based programs by establishing
the biological benchmarks for (1) directly
measuring the condition of the aquatic biota,
(2) determining water quality goals and setting
priorities, and (3) evaluating the effectiveness of
implemented controls and management actions.
The challenge of evaluating effects from ecological
stressors will best be met when the condition of
the biota within an ecosystem can be assessed
directly. Biological criteria for aquatic life will
2-8
-------
Lake Biological Monitoring
help meet this need by allowing direct assess-
ment of the condition of the biota that live either
part or all of their lives in aquatic systems. These
criteria (narrative or numeric) describe the
expected biological condition of an aquatic
community. They can be used as benchmarks to
Identify biological impairments and to help define
ecosystem goals and endpoints. Biological criteria
supplement traditional measurements (for
example, as backup for hard-to-detect chemical
problems) and will be particularly useful in
assessing impairment due to nonpoint source
pollution and nonchemical (e.g., physical and
biological) stressors. Thus, biological criteria
fulfill a function missing from USEPA's tradition-
ally chemical-oriented approach to pollution
control and abatement (USEPA 1996a).
Biological criteria can also be used to refine the
aquatic life use classifications for a state. Each
state develops its own designated use classifica-
tion system based on the generic uses cited in
the CWA, including protection and propagation of
fish, shellfish, and wildlife. States frequently
develop subcategories to refine and clarify
designated use classes when several surface
waters with distinct characteristics fit within the
same use class or when waters do not fit well into
any category; for example, cold-water versus
warm-water habitat. As data are collected from
biosurveys to develop a biological criteria pro-
gram, analysis may reveal unique and consistent
differences between aquatic communities that
inhabit different waters with the same designated
use. Therefore, measurable biological attributes
can be used to refine aquatic life use or to
separate one class into two or more subclasses.
2.9 OTHER USES
Although biological criteria and monitoring
might be perceived in a regulatory context as
one form of water quality
management, they serve
many other equally impor-
tant functions, including
the following:
• Evaluating the effec-
tiveness of manage-
ment practices.
• Regional planning.
• Watershed planning.
• Determining manage-
ment priorities for
multiple waterbodies.
Sections 303 and 304 of
the CWA require states to
protect biological integrity
as part of their water
quality standards. This
can be accomplished, in
part, through the develop-
ment and use of biological
criteria.
• Further classifying and
qualifying relative
water quality in a waterbody.
• Characterizing aquatic life that is at risk
from various hazards.
• Providing a means to evaluate impacts that
might not be protected by traditional risk
assessment methods.
2-9
-------
Chapter 2
2-10
-------
In This Chapter...
> Outline of the Biological Assessment and
Criteria Process
> Application to Lakes
Chapter 3
Overview of Bioassessment and
Biocriteria
3.1 CONCEPTUAL
FRAMEWORK
The impact of human activities on lakes has
been recognized for many centuries, and in the
past 50 years, there has been more focus on the
biological measurement of this impact. By 1950,
the first index of aquatic species' tolerance to
organic pollution, the "Saprobic System," was in
use (see Hynes 1994 for review). More recently,
indices such as the Hilsenhoff Biotic Index
(HBI), which takes into account both organic
pollution tolerance and the relative abundance
of species (e.g., Hilsenhoff 1987), have been
developed.
As modern ecologists recognized that human
influences were reducing local and global
biological diversity, the measurement of commu-
nity structure (including species diversity and
ecological roles) assumed increasing importance
in evaluation of polluted sites. Indices to mea-
sure species diversity and distribution in a
community (Pielou 1977) were developed, but
achieved only limited use because their one-
dimensional focus leads to high levels of uncer-
tainty in assessment.
3.1.1 Multimetric Biological
Assessment
The multiple attribute (or multimetric) approach,
incorporating pollution tolerance, diversity, and
ecological functions, was developed to more fully
characterize the human impact on aquatic
organisms. Karr (1981) and Karr et al. (1986)
developed the fish Index of
Biotic Integrity (IBI) and The multimetric approach
demonstrated that combi-
nations of these at- defines an array of mea-
tributes, or measure-
ments. forming an index, surements, each of which
provide valuable assess-
ments of water resources.
The multimetric approach
defines an array of mea-
surements, each of which
represents a measurable
characteristic of the
biological assemblage that
changes in a predictable
way with increased or
decreased environmental
stressors (USEPA 1996a,
USEPA 1997d). When integrated, a multimetric
index functions as an overall indicator of
biological condition. Each assemblage in the
represents a measurable
characteristic of the
biological assemblage that
changes in a predictable
way with increased or
decreased environmental
stressors.
3-1
-------
Chapter 3
aquatic community (for example, fish or algae)
might have differing responses to pollution or
degraded conditions. Thus, assessment methods
that target multiple species and assemblages
are capable of detecting a
broad range of stresses
and reflect the condition of
a large segment of the
ecosystem. However, there
is not yet a complete
understanding of how
measurements respond,
either quantitatively or
qualitatively, to perturba-
tion in general and to
particular stresses.
Biological assessment of
waterbodies depends on
our ability to define,
measure, and compare
biological condition
among similar systems.
To provide for an effective assessment, the
variables selected to determine biological
integrity should:
Be relevant to societal concerns—Biological
measurements must be related to the proper-
ties of biotic systems that are of concern to
society, such as native species, fish produc-
tion, and biological diversity.
Be responsive to environmental stresses—Biologi-
cal measurements and the measurements
developed from them must be sensitive to
environmental stress, and the response must
be interpretable.
Have low uncertainty—Variability should be
understood and measurement error should be
controllable.
Be cost-effective—The cost incurred in measure-
ment should be proportional to the value of the
information obtained.
Be environmentally benign to measure—Sampling
methods that disturb or alter habitats and
organisms should be avoided.
Assessment of biological integrity typically
focuses on a few broad but integral classes of
ecological properties (e.g., Barbour et al. 1992,
Karr 1991) that respond to anthropogenic
impacts (e.g., Schindler 1988, Schindler et al.
1989), including:
Health—Individuals or populations.
Species structure and composition—The number
and kinds of species in an assemblage. Species
structure includes both diversity and the
presence of pollution-tolerant species.
Trophic structure—The relative proportion of
different feeding levels, such as filter feeders,
scavengers, or predators.
System junction—The productivity and material
cycling of the system.
Multimetric assessment typically includes
several measurements of at least three proper-
ties (species structure, trophic structure, and
system function). Individual and population
health measurements are used less often
because they are not yet well developed for
invertebrates and plants.
Biological assessment of waterbodies depends
on our ability to define, measure, and compare
biological condition among similar systems.
Impairment of the waterbody is judged by its
departure from the expected condition. This
ability requires a functional definition of
biological integrity as the condition of the
aquatic community inhabiting unimpaired
waterbodies of a specified habitat as mea-
sured by community structure and function
(USEPA 1990a).
This definition of biological integrity makes the
explicit assumption that natural, undisturbed
systems are healthier than those changed by
human activities. Because biological integrity is
defined relative to unimpaired conditions, it
must also be measured relative to those condi-
tions. The four classes of ecological properties
listed above are measurable relative to natural
or unimpaired conditions.
Few waterbodies, however, are unimpacted.
Minimally impaired waterbodies typically form
the basis for defining reference conditions for
biological assessment. Artificial lakes, such as
reservoirs and impoundments, have no natural
or "least disturbed" condition. Nevertheless, it is
possible to define "most desirable" and "least
desirable" conditions for artificial lakes.
3.1.2 Biological Assessment
Process
The information of the biological variables is
transformed to numeric scores, or rankings from
good to poor. Such scores reduce the complexity
3-2
-------
Overview of Bioassessment and Biocriteria
and uncertainty of multidimensional data for
purposes of assessment, remediation, and
communication of results to the public and
decision makers. For example, managers might
need to know whether a lake is in good condi-
tion, whether it needs to be watched more
closely, or whether more intensive studies
should be made to determine a course of action
for restoration or remediation. Data analysis
streamlines the information from the data to two
or three dimensions that can be used in deci-
sion-making.
Mullimetric biological indices are similar in
concept to the common economic indices such
as the Index of Leading Economic Indicators
(Lahiri and Moore 1991). Both economic and
biological indices are based on comparison to an
operationally defined and measurable reference
standard. In the economic indices, individual
attributes are first standardized as a percentage
of a baseline value, usually an annual average
from a decade before (Green and Beckman
1992). The attribute scores are summed, and
the sum is likewise expressed as a percentage of
the index baseline. Standardization weights
indicators equally and allows the use of indica-
tors with different units (hours worked, persons
unemployed, billions of dollars, etc.). In
multimetric biological indices the metrics are
standardized as a score compared to a reference
standard. The basic procedural steps for biologi-
cal assessment are as follows:
1. Sample the biological groups (assem-
blages) selected by the program, record-
ing the relative abundance and other
characteristics of each species.
2. Calculate chosen metrics using relative
abundance and other measurements:
for example, number of species, number
of intolerant species, percent abundance
of filter feeders.
3. Compare each to its expected value
under reference conditions and assign a
numeric score corresponding to good
(similar to reference), fair (different from
reference), or poor (substantially differ-
ent from reference).
4. Sum the scores of all metrics of an
assemblage to derive a total score for
the assemblage.
5. Compare the total score to the biologi-
cal criterion based in part on the
expected total score under reference
conditions.
In biological assessment, reference conditions
are established by identifying least impaired
reference sites, characterizing the biological
condition of the reference sites, and setting
thresholds for scoring the measurements. For
reservoirs or in other instances where "best-
quality" lakes are too few or not definable, an
alternative is to select the
highest quality conditions
from among all lakes (TVA
1994). bioassessment Is most
Multimetric
effective when it is
modified to specific
regional conditions.
Multimetric bioassess-
ment is most effective
when it is modified to
specific regional condi-
tions. Bioassessment of
streams has been suc-
cessful when modified and calibrated regionally
(e.g., Barbour et al. 1996a, Miller et al. 1988,
Ohio EPA 1990). Success requires region-
specific selection and calibration of measure-
ments, as well as regional characterization of
reference conditions. For example, submerged
macrophytes are rare in rocky, high-elevation
or high-latitude lakes and may be an inappro-
priate assemblage in such a region.
3.1.3 Biological Assessment in
Ecological Risk
Assessment
Ecological risk assessment "evaluates the
likelihood that adverse ecological effects may
occur or are occurring as a result of exposure to
one or more stressors" (USEPA 1992c). Risk
assessment is a process for organizing and
analyzing data, information, assumptions, and
uncertainties in order to examine the likelihood
of adverse effects (USEPA 1996d). This process
provides risk managers with a framework for
explicitly considering available scientific infor-
mation in conjunction with social, political, and
economic factors when planning a course of
action with environmental consequences.
Problem formulation is the foundation of risk
assessment and depends on identification of
assessment endpoints, development of concep-
tual models, and creation of an analysis plan.
3-3
-------
Chapter 3
Assessment endpoints are "explicit expres-
sions of the actual environmental value that is
to be protected" (USEPA 1992c). Assessment
endpoints include both a valued ecological
entity and an attribute of that entity that is
potentially at risk (USEPA 1996d). For ex-
ample, the fish community of a lake is an
entity, and its overall similarity to native fish
communities in undisturbed lakes could be the
attribute for ecological risk assessment.
Biological assemblages and their attributes, as
discussed in this and other biocriteria docu-
ments (e.g., USEPA 1996a), are clearly potential
assessment endpoints for ecological risk
assessments. Following
risk assessment, a
decision may be made to
proceed with a manage-
ment action. Monitoring
can help determine if the
desired result of the
management action is achieved. Again, moni-
toring must include assessment endpoints,
and established biocriteria can provide unam-
biguous ecological assessment endpoints.
Biological assessment
emphasizes evaluation of
both habitat and biota.
3.2 APPLICATION TO LAKES
Biological assessment emphasizes evaluation
of both habitat and biota. As integrators of
processes in their watersheds, lakes receive
and retain matter and energy released through-
out the watershed. Human activities are part of
these processes and can affect a lake's habitat
and biological community. The impact of human
activities directly affects lake habitat and can
alter the lake's physical-chemical environment
(Figure 3-1). For example, contaminant dis-
charges can affect the chemistry of both the
water and the sediment. Agricultural and urban
land uses in the watershed contribute sediment
that affects the physical habitat. Humans can
affect biological communities either directly by
such activities as stocking and harvesting, or
indirectly through impacts to the physical and
chemical habitat of the biota.
Previous multimetric indices of lake quality have
focused on lake condition compared to water
quality standards, rather than on the actual
biological condition of a lake compared to its
regional potential. A multimetric index for
NUTRIENTS
SEDIMENTS
TOXICANTS
5
HUMAN HEALTH
CONCERNS
WEEDY PLANT
GROWTH
LOW DISSOLVED
OXYGEN
FOOD CHAIN
POOR WATER CLARITY
AQUATIC PLANT
GROWTH INHIBITED
WM
BIOLOGICAL
ASSEMBLAGES ALTERED
A Healthy Laka has clean water, balanced algal growth, adequate oxygen levels, and abundance and
diversity of fish and bottom-dwelling Invertebrates. Natural aquatic plants flourish in appropriate habitats,
and bottom habitat Is uncontamlnated.
Figure 3-1. Effects of pollutants In lakes.
3-4
-------
Overview of Bioassessment and Biocriteria
environmental quality of the Great Lakes used
physical, chemical, biological, and toxicity
variables (Steinhart et al. 1982). The Ohio EPA
developed a multimetric assessment for inland
lakes and reservoirs, the Ohio Lake Condition
Index (LCI) (Davie and DeShon 1989), which was
used to report lake condition for more than 300
public lakes in Ohio. The Ohio LCI consists of
14 metrics which represent biological, chemical,
physical, and public perception of lake condition.
Biological components in the Ohio LCI include
fish IBI, macrophytes, phytoplankton chloro-
phyll, fecal coliform bacteria, and fish tissue
contamination. Data are compared against
water quality standards or general criteria to
determine good, fair, or poor condition.
The Tennessee Valley Authority (TVA) developed
biological assessment for its reservoirs that used
a similar approach to the multimetric indices
developed for stream assessment (Dycus and
Meinert 1992. TVA 1994). TVA's assessment uses
five indices based on benthic macroinvertebrates,
fish, chlorophyll a sediment quality, and
dissolved oxygen. The macroinvertebrate and
fish indices are multimetric.
The USEPA lake biological assessment procedure
developed in this document may include up to
seven biological assemblages; planktonic algae,
attached algae, sedimented diatoms, aquatic
plants, bottom-dwelling invertebrates, fish, and
planktonic animals (Figure 3-2). Habitat scoring
components include the watershed, nearshore
zone, water chemistry, and sediment.
The proposed assessment of lake condition is
accomplished with additive indices that inte-
grate the habitat and biological scores. The
process produces up to three habitat scores,
and three or more biological index scores. The
scores reduce the complexity of a lake to an
understandable level for guiding appropriate
remediation or other management actions.
3.2.1 Tiers for Sampling
Biological assessment of lakes is implemented
in tiers corresponding to the level of effort
required. Each suggested tier includes both
biological and habitat components. The tiered
approach for lake bioassessment developed here
allows customization of the methodology to the
Bottom-Dwelling
(banthlc)
Invertebrates
Diatoms
Surface
Growing
(perlphyton]
.Sediment Diatoms
\ (planktonic and
\ surface)
Submerged
Aquatic Plants
Fish,
Recent
Diatoms
Fossil Diatoms
Figure 3-2. Biological assemblages used for lake assessments,
-------
Chapter 3
user's needs, questions, and resources avail-
able. Her 1 focuses on sampling trophic state
indicators, and Tier 2 focuses on sampling
biological assemblages for composition and
structure indicators (Figure 3-3 Table 3-1). Each
tier is further divided into single- and multiple-
visit sampling, A and B, respectively. Tier 1A and
IB are the same except that Tier IB requires
several samples during the growing season to
obtain seasonal averages
of chlorophyll a and nutri-
ent concentrations.
Relevant lake classes
must be determined by
existing Information and
the professional judgment
of scientists familiar with
lakes of the region.
Tier 2A consists of biologi-
cal assemblages that
integrate lake conditions
and are sampled during an
index period. Tier 2B
consists of assemblages
with individuals that are
short-lived, and hence do
not integrate over time. Tier 2B assemblages
are sampled repeatedly during the growing
season to obtain seasonal averages.
Because chlorophyll and nutrient concentra-
tions are highly variable. Tier 1A, which is
sampled only during an index period, may fail
to characterize an individual lake. Tier 1A is
appropriate for characterizing a region or a
class of lakes, especially if many lakes are to
be sampled. For characterizing the trophic
state of individual lakes with confidence, Tier
IB is preferred.
Both Tier 2A and 2B sample biological assem-
blages to estimate indicators of species struc-
ture, trophic structure, and function. Tier 2B
requires multiple visits and analysis, but does
not necessarily obtain better or more precise
information than Tier 2A.
3.2.2 Classification of Lakes
Because there is tremendous variation in the
physical, chemical, and biological characteris-
tics of lakes nationwide, the first step in defining
reference conditions is to classify lakes so that
comparisons can be made within, not across,
classes. Classification of natural lakes should
reflect the inherent properties of lakes indepen-
dent of human influence and therefore must be
made on the basis of measurements that do not
change as a result of human activities.
A second requirement of classification is that it
should reflect differences in the biota of the
classes. A deep lake might have a fish assem-
blage different from that of a shallow lake, and
classification should distinguish between the
two types of systems. Several lake classifications
have been proposed (e.g., Hutchinson 1957,
Leach and Herron 1992); however, only a
handful of lake classes would be present in a
single region. Relevant lake classes must be
determined by existing information and the
professional judgment of scientists familiar with
lakes of the region.
3.2.3 Characterization of
Reference Conditions
Five elements, detailed in Section 4.2, may be
used to establish reference conditions for lake
biological assessment:
• Biological survey of sites.
• Paleolimnology.
Table 3-1. Sampling tier summary.
Tier 1A
Trophic State Indices and macrophyte cover. Sampled once during index period.
Inference limited to regional assessment.
Tier IB
Trophic state indices and macrophyte cover.
Sampled repeatedly during growing season.
Tier 2A
Tier 1 (1A or 1B) plus two or more integrating biological assemblages;
macrophytes, macroinvertebrates, sedimented diatoms, fish. Sampled once
during index period.
Tier 2B
Tier 1B plus two or more short-term biological assemblages: phytoplankton,
zooplankton, periphyton. Sampled repeatedly during growing season.
3-6
-------
Overview of Bioassessment and Bioer/tesSa
• Evaluation of historical data.
• Prediction of expected conditions using
models.
• Expert consensus.
Expert consensus is required for developing
reference conditions. Reference conditions
developed from empirical data are preferred:
such as biosurveys, sites, paleolimnology, or
historical data.
A biological survey provides the best current
information about the biota for the system of
concern as a real world reflection of biological
integrity. This information is essential to deter-
mining the reference condition and subsequent
biological criteria. There are two approaches for
characterizing reference conditions from a
biological survey:
Site based—Selection of minimally impaired or
most natural sites in a region; or
Condition based—Setting reference conditions
as the best available ambient biological condi-
tions.
Paleolimnology is the microscopic examination
of sediment cores to provide an accurate record
of the relative abundance of certain organisms
(primarily diatoms} over the history of natural
lakes. The advantage of paleolimnology is that
any lake with an accurate sedimentary record
can be a reference site regardless of the severity
of present-day pollution. Thus, a truly represen-
tative sample of lake reference sites can be
drawn. With some exceptions, paleolimnology is
generally not applicable to impoundments.
A panel of diverse regional experts involved in
the determination of the reference condition and
the derivation of the biocriteria is the best
approach to thoroughly and objectively assimi-
late the above information. With a carefully
selected and balanced panel, all of the nuances
of the local ecology as well as the best interests
of the jurisdiction can be equated to the desig-
nated uses of the waterbody in designing the
most protective criteria possible. This approach
also reduces the risk of making insufficiently
informed decisions inherent in data interpreta-
tion by just one or a few like-minded people.
.2.4
Reference Condition In
Reservoirs
Throughout this document where
differences between lakes and reser-
voirs dictate alternative methods,
strategies, etc., an icon appears,
directing the reader to reservoir-specific
information.
The methodology described in this document is
intended for both reservoirs and natural lakes.
Because reservoirs are entirely artificial
environments, "natural reference condition"
has no meaning. Reservoirs, created by the
damming of a stream, have characteristics of
both rivers and lakes (Thornton 1990a). Reser-
voirs are divided into three zones (riverine,
transitional, and lacustrine), which correspond
to flowing, river-like conditions: transition to
lake conditions; and nonflowing, lake-like
conditions near the dam, respectively. With
expected life spans ranging from one to several
decades, reservoirs are more ephemeral than
most natural lakes and have several physical
characteristics not shared
with natural lakes. The
lakes most like reservoirs Because reservoirs are
are those formed by
natural dams in stream entirely artificial environ-
valleys (e.g., beaver dams,
terminal moraines, land-
slides). condition" has no mean-
ments, "natural reference
ing. Reservoirs, created by
the damming of a stream,
have characteristics of
both rivers and lakes.
Reservoirs vaiy widely in
physical characteristics of
shape, size, and hydrology.
They can range from small
shallow impoundments, to
deep storage reservoirs, to
"run of the river" flow-
through reservoirs on large
rivers. They are built and managed for widely
different purposes, including flood control,
navigation, water storage, hydroelectric genera-
tion, gamefish production, and others. The
management practices in turn affect both
physical characteristics (water level variability,
stratification) and biota (stocking of fish).
Although no "natural" reservoir reference
conditions can exist, the operational determi-
nation of reference conditions for reservoirs is
3-7
-------
Chapter 3
the same as that for natural lakes. Reservoirs
can be classified according to hydrology,
morphometry, management objectives, and
other factors. Age of the reservoir will be
Important in determining the assessment
expectations of the reservoir.
Historical data are important because they
provide insight to past conditions essential to
knowing what may be achievable, especially for
degraded or significantly altered systems.
Comparison of the historical record to present
reference site data greatly expands the
manager's perspective of the system. However,
care must be exercised in making these com-
parisons when the objectives and survey meth-
ods have changed over time.
Ecological models may be used to identify water
chemistry reference conditions for reservoirs or
for other significantly altered waterbodies. Most
reservoirs are less than 50 years old, and there
is insufficient empirical evidence to document
the expected condition of basins for all regions.
Where documentation is available (historical
data), extrapolation and model development help
qualify the reference condition and may be the
best way to derive and calibrate the bioeriteria.
3.2.85 Mefrle Determination
Metrics are evaluated for relevance to biological
assessment and for response to stress. Expected
measurement values vary as a function of
regional species pools, regional characteristics
{climate, geology, soils, land use, regional scale
barriers to colonization),
A regional approach and local site characteris-
tics (habitat factors.
Involving collaboration including local barriers). A
regional approach involving
of neighboring jurisdic- collaboration of neighbor-
ing jurisdictions will
enhance characterization
characterization of of reference conditions.
Cross-state comparisons
reference conditions. can be made more easily if
common methods and
measurements can be established among
states.
Metrics are typically calculated from data
collected on single assemblages of lake biota,
such as planktonic algae, zooplankton, fish.
tori will enhance
aquatic plants, and benthic invertebrates. The
metrics might include counts, species identifi-
cations, ratios, and indices combining several
data variables depending on the level of effort,
or tier, of the survey.
3.2.6 Data Analysis
When performing bioassessment of lakes,
individual metrics are assigned scores, usually a
number corresponding to good, fair, or poor
relative to the values of the measurements in
reference conditions (Karr 1991, Karr et al. 1986).
This serves to standardize the metrics on the
same scale so they can be combined into an
additive index. Measurement scores are summed
to obtain an index score for each assemblage,
such as an IBI or macroinvertebrate community
score. Currently, each measurement is weighted
equally in the summed index score.
Additive biological indices collapse a great deal
of information into a single number. Yet they
have been shown to be reliable in detecting
impairment of aquatic systems (Fore et al. 1994.
Fore et al. 1996, Wallace et al. 19S6); they are
simple to compute once criteria are established,
and they are easily communicated to managers
and the public (Gerritsen 1995).
Habitat component scores may give clues to the
causes of impairments reflected in biological
indices rated fair or poor. Habitat variables that
are significantly different from reference condi-
tions are identified as probable causes of
impairment, warranting further investigation or
remediation. This sort of bioassessment cannot
establish cause of impairment; it can only
separate probable from improbable causes of
impairment. In any given bioassessment, several
probable causes might be identified.
3.3 BIOCRITEPIIA
Biological data are used to help set biological
criteria based on management needs and
defined management classes. States may draft
general narrative bioeriteria early in their
program—even before they have designated
reference sites or refined their approach to
biological surveys. This does not mean that
having reference sites and a refined system for
conducting surveys is unimportant; it means
3-8
-------
Overview of Bioassessment and Blocriteria
that a biocriteria program begins with writing
into law a statement of intent to protect and
manage water resources predicated on an
objective or benchmark, for example, "aquatic
life shall be as naturally occurs."
When the objective to restore and protect the
biological integrity of the water resources has
been formally mandated, then the operational
meaning 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 judgement and management.
Although based on the same concept as narra-
tive 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.
The key distinction between narrative biocrite-
ria supported by a quantitative database and
numeric biocriteria is the direct inclusion of a
specific 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 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, help establish
the normal range of the biological community as
it would exist in unimpaired systems.
3-9
-------
Chapter 3
Case Study: Biological Assessment of Reservoirs by TVA
The Tennessee Valley Authority is currently using
a multimetric biological assessment methodology
on its reservoirs. The Tennessee River watershed
drains portions of four ecoregions: Blue Ridge, Cen-
tral Appalachian Ridge and Valley, Southwestern
Appalachians, and Interior Plateau (Omernik 1987)
(Figure 3-3), The Tennessee River begins at the
confluence of the Holston and French Broad Rivers
mid receives drainage from the Ridge and Valley
and Blue Ridge ecoregions. Downstream, the river
drains a small portion of the Southwestern Appala-
chians and a large part of the Interior Plateau. The
main stream carries water from two to four
ecoregions. Therefore, dividing die main stream res-
ervoirs by ecoregion does not contribute to a mean-
ingful classification. Figure 3-3 illustrates that the
tributary reservoirs can be easily divided by
ecoregion. There are several reservoirs with water-
sheds entirely within the Blue Ridge and Ridge and
Valley ecoregions. There is a third, and more dis-
persed, group of tributary reservoirs in the Interior
Plateau.
Physical, chemical, and biological indicators were
selected to provide information on the health or
condition of habitats or ecological compartments.
The open water or pelagic area was represented by
physical and chemical characteristics of water (in-
cluding chlorophyll) in mldchannel. The shoreline
or littoral area was evaluated by sampling the fish
community. The bottom or benthic compartment was
evaluated using two indicators: quality of surface
sediments in midchannel (determined by chemical
analysis of sediments) and examination of benthic
macroinvertebrates from a transect across the full
width of the sample area (including overbanks if
present).
Three areas were selected for monitoring: the in-
flow area, generally riverine in nature, the transition
zone or mid-reservoir area where water velocity
decreases due to increased cross-sectional area,
suspended materials begin to settle, and algal pro-
ductivity increases due to increase water clarify; and
the forebay, the lacustrine area near the darn.
Overbanks, basically the floodplain which was in-
undated when the dam was built, were included in
transition zone and forebay areas. Four large
embayments (all with drainage areas greater than
500 square miles and surface areas greater than
4500 acres) were included in the Vital Signs Moni-
toring program. Ecosystem interactions within an
embayment are mostly controlled by physical char-
acteristics of the embayment and by activities and
characteristics within the embayment watershed,
usually with little influence from the main body of
the reservoir (Meinert et al. 1992).
Sampling frequencies and index periods take Into
account the expected temporal variation for each
indicator. Physical and chemical components vary
significantly in the short term so they are monitored
monthly from spring to fall. Biological Indicators bet-
ter integrate long-term variations and are sampled
once each year. Fish assemblage sampling Is con-
ducted in autumn (September-November).
Initially, benthic macroinvertebrate sampling was
conducted in early spring (February-April) to avoid
aquatic insect emergence. The TVA experience
showed that a late winter/early spring sampling pe-
riod is not acceptable for benthic macroinvertebrates
because results reflected conditions which occurred
the previous year. This causes results from this in-
dicator to be out of synch with the other four indica-
tors. A late fall/early winter collection avoids prob-
lems resulting from early spring sampling.
The TVA case study is continued in subsequent
chapters.
3-10
-------
Overview of Bioassessment and Biocrtteria
Case Study: Biological Assessment of Reservoirs by TV A (continued)
X) \s
V&tts
CNck
V
Norma rx^L^c
tahda Lake
eTeke
Tims Ford Lake
Rckwsck Lake
Wheeler Lake
K Qnlersvile
68
AL
Eooregions
ffi Southeastern Plains 70 Western Allegheny Rateau
66 Hue Ridge Mountains 71 Interior Plateau
67 Ridge and Valley 74 Mississippi Valley Loess Plains
68 Southwestern Appalachians 72 Interior River Lowland
69 Central Appalachians
Figure 3-3. Distribution of TVA reservoirs in ecoreglons
3-11
-------
In This Chapter...
> Preliminary Classification of Lakes
> Methods for Establishing Reference
Conditions
Chapter 4
Selection and Characterization of
Reference Conditions
Establishment of reference conditions is key to
biological assessment and biocriteria programs.
Reference conditions are a representation of the
biotic potential for lakes in the absence of
human activity or pollution. The attainment of
aquatic life use is evaluated against the expecta-
tions of the reference condition as expressed in
the biocriteria. Reference conditions are expec-
tations on the status of biological communities
under minimal anthropogenic disturbances and
pollution. The expectations are usually based on
the status of reference sites, which might be
subject to anthropogenic influences. Ideally,
reference sites are minimally impacted by
human pollution and disturbance. The care that
states use in selecting reference sites and
developing reference condition parameters,
together with the survey techniques employed,
will bear directly on their ability to defensibly
assess a waterbody. At a minimum, reference
conditions should be identified for each of the
lake classification categories developed for a
state. As pointed out in Section 3.2.4, the
definition of reference condition differs between
natural lakes and artificial reservoirs.
The general sequence of reference condition
characterization is to first assemble a panel of
experts and make a preliminary classification of
lake resources within a region. Following classi-
At a minimum, reference
fication, sampling sites are selected, and habitat
and biological data are obtained from those sites
(either from existing data
bases or from a survey).
The preliminary classifica-
tion is reconciled with the conditions should be
biological data to ensure
that the final classification identified for each of the
is biologically meaningful,
and the reference condi- lake classification catego-
tions are characterized as .
part of the biocriteria nes developed for a state.
development process.
4. 1 REGIONALIZATION AND
PRELIMINARY
CLASSIFICATION
The regional differences in biological communi-
ties across the United States must be accounted
for in the development of biological monitoring
programs. This is done by comparing the biology
of lakes to a regional reference condition. As
biological conditions change across the country,
the reference conditions will change also. To
account for the regional differences in biological
communities, and also for the differences that
result from structural differences in biological
4-1
-------
Chapter 4
single reference condition
habitat (either natural or caused by human
activities), USEPA recommends that states
classify lakes into categories and that a refer-
ence condition should be developed for each of
the lake categories. Biotic
Lakes vary widely In size, index comparisons can
then be made within each
shape, and ecological category, and inappropri-
ate biological comparisons
characteristics, and a between different classes
will be avoided. Moreover,
the aquatic life expecta-
that applies to all lakes tions of waterbodies are
tempered by realistic
would be misleading. regional expectations; there
is no attempt to set a single
numeric aquatic life designated use standard for
the entire nation.
Lakes vary widely in size, shape, and ecological
characteristics, and a single reference condition
that applies to all lakes would be misleading.
The purpose of classification is to group similar
lakes together; i.e., to prevent comparison of
apples and oranges. By classifying lakes the
variability of biological measures within classes
is reduced and the variability among classes is
maximized. Classification invariably involves
professional judgment to arrive at a workable
system that separates clearly different ecosys-
tems, yet does not consider each lake a special
case. The intent of classification is to identify
groups of lakes that, under ideal conditions,
would have comparable biological communities.
As far as possible, classification should be
restricted to those characteristics of lakes that
are intrinsic, or natural, and not the result of
human activities.
-a.. 1.1 Definition of the
Resource
Most large reservoirs, and some natural lakes,
are on rivers and might be considered large
pools in the rivers rather than lakes. At what
point does a pool become a lake? For the
purpose of lake bioassessment, it is when
distinctly lake-like flora and fauna occur (i.e.,
phytoplankton and zooplankton). Phytoplankton
require a water retention time of 3 days or more
(Uhlmann 1971). Microzooplankton (e.g.,
rotifers) have generation times roughly twice
that of phytoplankton cells; therefore, the
minimum retention time for zooplankton to
develop may be approximately 1 week.
For the purposes of bioassessment described
here, a lake is any inland body of open water
with some minimum surface area free of rooted
vegetation and with an average hydraulic
retention time of more than 7 days.
These characteristics distinguish lakes from
small ponds and wetlands, and from riverine
pools (natural or artificial) that retain their lotic
character. The distinction between lake and
small pond is arbitrary, and the minimum size
for a waterbody to be considered a lake must be
set by resource agencies. For practical reasons,
this document does not explicitly consider
emergent wetlands at the margins of lakes.
Bioassessment methods for wetlands are being
developed separately by USEPA and other
agencies.
The unit of assessment and sampling (the
sampling unit) is, most commonly, a definable,
relatively self-contained basin of a lake. Most
lakes have a single basin and thus will consist
of a single sampling unit. Larger lakes, and
especially reservoirs, have embayments, arms,
and basins that are hydrologically isolated from
the main body of the lake. Each isolated basin
can be considered a separate sampling unit
because of restricted water flow between basins.
Large lakes can thus comprise several sampling
units. Alternatively, a state may wish to define
the sampling unit as an area or point in space
(e.g., lm2).
Most reservoirs are also divided into three
zones—riverine, transitional, and lacustrine—to
reflect differences among these zones (Thornton
1990b). Each zone is a separate sampling unit;
in large reservoirs, zones might be represented
in each major arm (TVA 1994),
4.1.2 Basic Rules
There is no single "best" classification, nor are
resources available to determine all possible
differences between all lakes in a region. The
key to classification is practicality within the
region or state in which it will be applied; local
conditions determine the appropriate classes.
Classification will depend on regional experts
4-2
-------
Selection and Characterization of Reference Conditions
familiar with the range of lake conditions in a
region, as well as biological similarities and
differences between the lakes. Ultimately,
classification can be used to develop a predictive
model of lake characteristics that affect the
values of the biological metrics and indices in
reference sites.
There are two fundamental approaches to
classification, a priori and a posteriori (Conquest
et al. 1994). The a priori approach consists of
developing logical rules for classification based
on observed patterns in characteristics of the
objects. Thus, classifying lakes on ecoregion,
surface area, and maximum depth would be an
a priori, rule-based classification. The a poste-
riori approach develops groups from a data base
of observations from the sites. The classification
is restricted to those sites and variables in the
data base and typically involves cluster analysis
to develop the groups. The a posteriori approach
is useful for exploratory analysis of a substan-
tial data set, but it is not appropriate for opera-
tional assessment and management, where a
site's class must be established from prior
information (e.g., maps) before intensive data
are collected. A few general rules for the devel-
opment of a priori lake classification include:
• In a priori classification, lake characteristics
that are readily affected by human activities
or are a biological response to physical or
chemical conditions should not be used as
classification variables. Such responses
might include trophic state, chlorophyll, or
nutrient concentrations. For example, in the
Northern Lakes and Forests ecoregion of
Minnesota, lake trophic state is characteris-
tically low whereas in the nearby Northern
Glaciated Plains ecoregion, trophic state is
relatively high (Heiskaiy 1989). The classifi-
cation variable in this case is ecoregion, and
trophic state is a response to ecoregion. A
eutrophic lake in the Northern Lakes and
Forests is considered impaired, but a
eutrophic lake in the Northern Glaciated
Plains is not considered impaired. Using
trophic state as a classification variable
could lead to misclassifications and inap-
propriate assessments.
• As shown in the example above, the best
classification variables are those which are
readily obtained from maps, bathymetric
charts, or regional water characteristics,
such as alkalinity or hardness.
4.1.3 Considerations for
Reservoirs
Several differences between reser-
voirs and natural lakes affect the
classification and interpretation of
biological data (Thornton 1990a,
Wetzel 1990):
RESERVOIR
Most of the differences
between reservoirs and
natural lakes are resolved
in classification of the lake
Distribution—Reservoirs are most numerous in
regions with few natural lakes: the
nonglaciated parts of North America (except
Florida) have the largest numbers of reservoirs
(Thornton 1990a).
Form—The form or shape of the basin and
watershed may be the most important distinc-
tion between natural and
artificial lakes. Shape
substantially influences
the hydrology and water
quality of reservoirs. Large
reservoirs are drowned
river valleys and tend to
be long and deep with
numerous embayments
from tributaries. The resource.
watersheds of reservoirs
are typically much larger
than those of natural lakes and contribute
greater sediment loads.
Longitudinal gradient—Reservoirs have charac-
teristics typical of both lakes and streams
within the same basin. They are more like
streams at the head where major tributaries
enter and are more like lakes near the dam
(Thornton 1990b).
Turbidity and loading—Reservoirs are typically
more turbid, and they receive more nutrients
and organic matter from their tributary streams
than do most natural lakes.
Management—Reservoirs were built and are
managed for specific purposes: hydro-power,
irrigation, flood control, fisheries, and multiple
uses. Management might include extreme
water level fluctuations, fish stocking, and
other effects not present in natural lakes.
4-3
-------
Chapter 4
Most of the differences between reservoirs and
natural lakes are resolved in classification of
the lake resource. The needs for which reser-
voirs were designed dictate many attributes of
these waterbodies. Operational strategies can
influence reservoir characteristics and resultant
water quality (Kennedy and Walker 1990,
Kennedy et al. 1985). The release of water from
deep in the water column increases heat gain
and the dissipation of materials accumulated in
bottom waters (Martin and Ameson 1978,
Wright 1967). Surface releases dissipate heat
and retain materials. These and other opera-
tional differences can provide a basis for group-
ing reservoirs because reservoirs operated
similarly can be expected to exhibit similar
limnological responses, even when compared
across large, heterogeneous regions.
4.1.4 Hierarchical
Framework
This protocol is not intended to develop a
classification scheme applicable to the entire
United States. Overviews of global lake classifi-
cation systems are in Hutchinson (1957) and in
Leach and Herron (1992). Classification must be
regional, and regional expertise must be used to
determine those classification variables which
are useful in a region.
A useful classification scheme is hierarchical,
beginning at the highest (regional) level and
stratifying as far as necessary (Conquest et al.
1994). The procedure is to classify lakes at the
highest level (usually geographic), and then to
continue stratification in the classification
hierarchy to a reasonable point. Although
several possible classification levels are outlined
below, in practice, only one, or at most two,
relevant levels would typically be used. Classifi-
cation should be parsimonious to avoid prolif-
eration of classes that do not contribute to
assessment. One or two relevant levels of the
hierarchy will yield the best classification scheme.
The proposed hierarchical scheme below applies
to both natural lakes and reservoirs.
Geographic Region—The geographic region (e.g.,
ecoregion, physiographic province) determines
landscape-level features such as climate,
topography, regional geology and soils, biogeog-
raphy, and broad land use patterns. Ecoregions
are based on geology, soils, geomorphology,
dominant land uses, and natural vegetation
(Hughes and Larsen 1988, Omernik 1987) and
have been shown to account for variability of
water quality and aquatic biota in several areas
of the United States (e.g., Barbour et al. 1996a,
Barbour et al. 1996b, Heiskaiy et al. 1987,
Hughes et al. 1994, Ohio EPA 1987).
Because of the importance of geography in
determining aquatic biota, the National Re-
search Council's Aquatic Restoration Committee
made the following recommendation (NRC
1992):
The committee believes that goals for
restoration of lakes need to be realistic and
should be based on the concept of ex-
pected conditions for individual ecoregions.
Further development of project selection
and evaluation techniques based on
ecoregion concepts and refinement of
ecoregion definitions and descriptions
should be encouraged and supported by
the U.S. Environmental Protection Agency.
Many of the characteristics below that can be
used as classification variables are often sub-
sumed by ecoregion. For example, watersheds
are often similar within ecoregions, having been
formed by the regional geomorphology, and
water quality characteristics such as alkalinity
are determined by regional bedrock and soils.
Within ecoregions, it might be sufficient to
classify using only lake basin morphology (e.g.,
depth, area, development ratio): anthropogenic
or natural origin; or management objective.
Anthropogenic Origin Reservoirs and
Ljl other artificial lakes cannot have
IHHSjVllllfl "natural" reference conditions.
Therefore, reservoirs and natural
lakes should be separated in develop-
ing reference expectations.
Watershed Characteristics—Watershed charac-
teristics affect lake hydrology, sediment and
nutrient loads, alkalinity, and dissolved solids.
As noted above, many watershed characteristics
are relatively uniform within an ecoregion and
may not be necessary if ecoregions were the
primary classification variable. Watershed
characteristics that may be used as classifica-
tion variables include:
4-4
-------
Selection and Characterization of Reference Conditions
• Lake drainage type (e.g., flowage, drainage,
, seepage, reservoir type).
• Land use.
• Watershed-to-lake area ratio (especially for
reservoirs).
• Slope (especially for reservoirs).
• Soils and geology (erosiveness of soils).
Lake Basin Characteristics—Lake basin mor-
phology influences lake hydrodynamics and lake
responses to pollution. Characteristics of some
reservoirs change with age, particularly regional
shoaling and silting of aged reservoirs subject to
high sediment loads (O'Brien 1990). Morphologi-
cal metrics include:
• Depth (mean, maximum).
• Surface area.
• Bottom type and sediments.
• Shoreline development ratio (shoreline
length: circumference of equal area circle).
• Age (of reservoirs).
• Epilimnetic/hypolimnetic discharge (reser-
voirs).
Lake Hydrology—Lake hydrology forms a basis
for water quality. Mixing and circulation pat-
terns influence nutrient retention and the
development of hypoxia. Hydrological factors
include:
• Retention time.
• Stratification and mixing.
• Circulation.
• - Water level fluctuation and drawdown.
Characteristic Water Quality—Lakes can be
classified by characteristic water types into
categories, such as marl lakes, alkali lakes,
ombrotrophic bog lakes, and others. Many water
quality characteristics are relatively uniform
within an ecoregion and as the result of re-
gional, watershed, basin, and hydrologic charac-
teristics. Water types are determined by the
following water quality variables:
Alkalinity.
Salinity.
Conductivity.
Turbidity (Secchi depth, clarity, etc.).
Color.
Dissolved organic carbon (DOC)-
Dissolved inorganic
carbon (DIC).
Classification must be
regional, and regional
expertise must be used to
determine those classifi-
cation variables which are
useful in a region.
Human actions (e.g.,
discharges, land use) alter
water quality, especially
sediment and nutrient
concentrations, but they
can also affect alkalinity,
salinity, conductivity,
color, and DOC. Care must
be taken that classifica-
tion according to characteristic water types
reflects natural conditions and not anthropo-
genic impacts. For example, if a lake is highly
turbid due to poor land management practices,
it should not be classified as highly turbid.
Rather, it should be classified as it would have
been in the absence of poor land use.
4.2 ESTABLISHING
REFERENCE CONDITIONS
Five elements are used to establish lake refer-
ence conditions for biological monitoring and
biological criteria: (1) expert consensus,
(2) biological survey of sites, (3) paleolimnology,
(4) evaluation of historical data, and (5) predic-
tion of expected conditions using ecological
models (Table 4-1).
4.2.1 Expert Consensus
Expert consensus is essential in supporting the
information and data interpretation derived from
the other approaches. It provides a balanced
and comprehensive assessment of all of the
information and promotes the optimum criteria
when properly done. A panel of experts is
assembled before any other steps are imple-
mented, to guide the process and to select the
best methods appropriate to the region for
4-5
-------
Chapter 4
Case Study: Selection of Candidate Reference Lakes
Florida has nearly 8,000 natural lakes larger than
10 acres. Owing to Florida's wet climate, flat topog-
raphy, and abundant karst-dominated geomorphol-
ogy, depressions are abundant and filled with water.
In the process of developing bioassessment and
biocriteria for Florida takes, the Florida Department
of Environmental Protection enlisted the help of
USEPA geographers and academic limnologists to
delineate lake ecoregions for the state. Forty-seven
lake regions were identified (USEPA 1997c). These
included regions with no natural lakes (only im-
poundments), regions with abundant lakes of a
single type, heterogeneous regions with several lake
types, and regions with ephemeral marsh lakes.
Several lake types were also identified including:
sand ridge lakes, solution lakes, swamp lakes, riv-
erine flowage lakes, marsh lakes, and others. After
the lake regions had been identified, candidate ref-
erence lakes were selected in each region. Candi-
date reference lakes are representative and rela-
tively least impacted within the lake region. In re-
gions where all lakes are impacted (for example,
the rapidly urbanizing area around Orlando, Florida),
candidate reference lakes are those that are least
impacted relative to the regional norm. Biologists
and limnologists with regional and local expertise
selected the candidate reference lakes. Following
selection, candidate lakes were surveyed to deter-
mine lake type and to confirm that they were rela-
tively least impacted. Reference sites were selected
from the candidate sites and a full biological survey
of the reference sites was conducted during Florida's
lake index period (late summer/fall).
characterizing reference conditions. The panel
should consist of skilled aquatic biologists,
physical scientists, fisheries biologists, and
natural resources managers.
In significantly disrupted areas where no
candidate reference sites are acceptable, a form
of this expert consensus is a workable alterna-
tive to establish reference expectations. Three
or four biologists can be
| convened for each assem-
The recommended bIage to be used in the
empirical approach is to assessment. Each expert
should be familiar with the
lakes of the region. Based
on their collective exper-
tise, they are asked to
develop a description of
the assemblage to be
will be used to Identify expected if the lakes were
relatively unimpacted. This
and calibrate metrics. description, developed by
consensus, will necessar-
ily be more qualitative
than quantitative, but will allow development
of metrics and metric scoring.
4.2.2 Biological Survey
The recommended empirical approach is to use
a population of reference lakes to establish
conditions that will be used to Identify and
use a population of
reference lakes to
establish conditions that
calibrate metrics. Pairwise comparison of two
lakes leads to the trivial conclusion that they are
different (Hurlbert 1984). All monitoring sites,
reference or impaired, can vary over time and
space for natural reasons. A central measure
from a composite of several reference sites is
used to base expectations to account for natural
variability and uncertainty. Statistically, this
means that the status of a lake is judged by
comparing the lake (the "test site") to a popula-
tion of reference sites. In hypothesis-testing
terminology, the null hypothesis examines
whether the test lake is a member of the popula-
tion of reference sites.
A critical requirement for the use of reference
conditions in biocriteria is the USEPA
antidegradation policy, which protects against
incremental deterioration of waterbodies and
reference conditions. An observed downward
trend in reference sites cannot be used to justify
relaxing reference expectations, reference
conditions, and the associated biological crite-
ria. Once established, biocriteria may only be
refined in a positive direction in response to
improved conditions.
To characterize reference conditions, surveys of
both reference sites and known impaired sites
are made for both biota and physical habitat.
These data are needed to determine gradients of
conditions (from best to impaired) for the
purpose of measurement calibration and dis-
4-6
-------
Selection and Characterisation of Reference Conditions
Table 4-1. Comparison of elements for characterizing reference conditions.
Expert Consensus
Biological Survey
Paleolimnology
Historical Data
Predictive
Models
Strengths
Guides and reviews
other procedures
May be used alone.
Relatively
inexpensive.
Common sense
and experience can
be incorporated.
Yields obtainable,
best current status.
Any assemblages
deemed important
can be used.
Two methods:
- selected reference
sites
- best of abmient
conditions
Yields historical
time series for
assemblages of
diatoms,
chrysophytes, and,
to a lesser extent,
some crustaceans
and some insects.
Can infer water
quality.
Yields actual
historical
information on
status.
Inexpensive to
obtain.
When data are
insufficient.
Works well for
water quality.
Weaknesses
Qualitative
descriptions of
"ideal"
assemblages.
Might be unrealistic
and not
representative of a
best attainable
potential.
Experts might have
strong biases.
Even best sites
subject to human
impacts.
Degraded sites might
lower subsequent
biocriteria.
Preservation of fish,
invertebrates,
macrophytes, and
non-diatom algae is
poor.
Studies may require
complex data
analysis and
interpretation by
experts.
Adequate sediment
record may not
exist in reservoirs.
Data might be
limited.
Studies likely were
designed for
different purposes;
data might be
inappropriate.
Human impacts
present in
historical times
were sometimes
severe.
Extrapolation
beyond known
data and
relationships is
risky.
Can be expensive.
crimination. The raw data must be evaluated
within the ecological context (waterbody type
and size, season, geographic location, and other
elements) that defines what is expected for
similar waterbodies.
Candidate metrics are developed from the key
biological attributes, and the effects of stressors
on specific metrics must be understood (USEPA
1996a). Those measurements that have a
monotonia response to a gradient of conditions
There am two primary approaches for selecting or mapped Information such as land use and roads,
determining reference conditions using data from sur- and other existing data bases,
veyed sites. The first approach uses selected best- ¦"
quality sites as the basis for determining reference * Determination of reference conditions based pn
conditions. The second approach does not use refer- best conditions found in a representative
ence sites, but draws its reference conditions directly sample of lake^ within a class- This approach^
from those found in a sample of many lakes of vary- is used when few appropriate reference sites exist
ing quality °rwh8n they cannot be suitably defined. A num-
'.'•"-.v. witbin the ds^s are surveyed, 'and'
1. Selection of^reference sites basedon a priordefi- v the best conditions for eachmeasurement ate
nition of reference site criteria—-This approach determined from the entire sample of lakes,
is used when a sufficient number of lakes exist These best conditions are then us&d as the; ref-
that are minimally impacted. Since nearly all lakes erence for biological assessment within thatlake
are affected by human activities to some degree, class. This is the preferred approach for many
the lakes need not be pristine or unimpacted, but large reservoirs and some exceptionally large or
the level of impact must be minimal relative to unusual lakes, where there are few other lakes
lakes in the region. Reference sites are selected ofthaiclass.
using local expert knowledge on candidate sites, .
4-7
-------
Chapter 4
(from unimpaired to heavily impaired) will be the
best candidates for assessing biological impair-
ment. Therefore, ambient sites other than refer-
ence sites should be surveyed as part of the data
base. Selection and confirmation of the measure-
ments must address the ability to differentiate
between impaired and unimpaired sites.
Minimally Impaired Reference Sites
Reference sites must be carefully selected
because they will be used as a benchmark
against which test sites will be compared. The
conditions at reference sites should represent
the best range of minimally impaired conditions
that can be achieved by similar lakes within the
region. The reference sites must be representa-
tive of the region, and relatively least impacted
compared to other lakes of the regions.
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. 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
regions where minimally impaired sites are
significantly degraded, the search for suitable
sites should be extended over a wider area.
Stringent criteria might require using park or
preserve areas for reference lakes. Criteria for
reference lakes will also pertain to the condition
of the watershed, as well as the lake itself. If
relatively unimpaired conditions do not occur in
the region, the selection process could be
modified to be more realistic and reflect attain-
able goals, such as the following:
Land use and natural vegetation—Natural vegeta-
tion has a positive effect on water quality and
hydrological response of streams. Reference
lakes should have at least some percentage of
the watershed in natural vegetation.
Riparian zones—Zones of natural vegetation
alongside the Iakeshore and streams stabilize
shorelines from erosion and contribute to the
aquatic food source through allochthonous
input. They also reduce nonpoint pollution by
absorbing and neutralizing nutrients and
contaminants. Watersheds of reference lakes
should have at least some natural riparian
zones regardless of land use.
Best management practices—Urban, industrial,
suburban, and agricultural nonpoint source
pollution can be reduced with successful best
management practices (BMPs). Watersheds of
reference lakes should have BMPs in place
provided that the efficacy of the BMPs has been
demonstrated.
Discharges—Absence or minimal level of
permitted discharges (NPDES) into surface
waters.
Management—Management actions, such as
extreme water level fluctuations for hydropower
or flood control, can significantly influence lake
biota. Reference lakes should be only mini-
mally impacted by management activities.
Predefined reference conditions for lakes have
been used in Minnesota to determine ambient
phosphorus criteria (Heiskary 1989). Maine
uses a similar approach in regulating the water
quality of streams and uses a reference stan-
dard of aquatic life as naturally occurs (Davies
etal. 1993).
If a fixed definition of reference condition is
deemed to be overly restrictive or an impractical
ideal, then an empirical working definition is an
alternative. For example, because natural condi-
tions for reservoirs cannot be defined, the best
existing conditions are used instead. This ap-
proach is also useful in ecoregions with little or
no contiguous stands of natural vegetation
If all lakes In a region are significantly altered, It
might not be possible to characterize reference
conditions from ecoreglonal data. In this case,
an alternative would be to use lakes from
neighboring regions as reference sites If those
lakes are deemed acceptable, by professional
judgment, with respect to impact and overall
comparability to the lakes of the affected region.
This Is one of the reasons why USEPA encour-
ages Interstate cooperation in monitoring arid
biocriteria development. If lakes from nearby
regions cannot reasonably be considered
reference sites, then reference conditions must
be predicted or inferred from other information,
including models and historical data. In design-
ing such an approach, the consensus of a panel
of regional experts helps ensure an objective
and rational design.
4-8
-------
Selection and Characterization of Reference Conditions
remaining, such as In the agricultural Midwest.
Choosing the best sites requires at least a
representative survey (or better, a census) of lake
watershed variables in the ecoregion. Individual
lakes with the best conditions, such as the
greatest percentage of forest or natural vegetation,
the lowest percentages of agricultural and urban
land use, etc., are chosen as reference sites.
Without antidegradation safeguards, the best
available approach might allow continual
deterioration. For example, construction and
development in a lake watershed that is one of
the "best" in a region might cause biological
degradation of the lake. If the set of "best" lakes
in the ecoregion have suffered similar degrada-
tion, they might still be the reference sites, but
the new reference condition will be degraded
relative to its earlier state. For example, Maine
has a antidegradation policy that requires that
lakes remain stable or improve in trophic state
(Courtemanch et al. 1989, NALMS 1992). An
effective antidegradation policy can promote
continually improving conditions.
The selected reference lakes should be represen-
tative of each of the classes, and a sufficient
number of lakes are then sampled to enable
characterization of each class. A general "rule of
thumb" for optimal sample size is 10-30 lakes
per class, and each lake is a sampling unit (see
Chapter 9 for estimating power and sample
size). In regions where all lakes are impacted,
the 10 to 30 relatively least impacted lakes of
each class (e.g., ecoregion) are sampled, where
"best" is determined by least anthropogenic
disturbance or impacts, but not by most
desirable biota. In regions where the popula-
tion of unimpaired reference lakes is large, a
stratified random sampling scheme (lakes in
each class selected randomly) will yield an
unbiased estimation of reference conditions.
"Stressed Reference Sites"—Effective metrics
respond to environmental degradation and
allow discrimination of impaired sites from the
reference expectations. Metrics that do not
respond are not useful in bioassessment.
Response is determined by sampling a set of
stressed sites in the same way as the refer-
ence sites—in effect, sampling a set of
"stressed reference" sites. Lakes with known
problems, such as nutrient loading, thermal
pollution, toxic sediments, or urban land use,
are good candidates for "stressed reference"
sites. There should be several in each class or
lake ecoregion for adequate tests of metric
responses. Because impaired lakes are fre-
quently objects of monitoring by natural
resource agencies, data might already exist to
test the biological metrics. However, the
sampling methods for reference and impaired
lakes should be comparable.
Sampling and Data Analysis—One or more of the
recommended tiers of biological assemblages
are sampled and identified. It is imperative
that reference sampling include all assem-
blages that will be used in operational sam-
pling and assessment. Sampling methods are
described in Chapters 4 and 5; data analysis is
described in Chapter 6.
Reference Conditions from Distributions of
Biological Metrics
If sufficient minimally impaired reference sites
do not exist or cannot be found, reference
conditions can be selected from an entire
population of sites. This approach is especially
relevant for human-made impoundments and
reservoirs, where no least-impaired systems
exist, as well as for resources subject to strong
and relatively uniform human impacts, such as
lakes in large urbanized areas or in heavily
agricultural regions. The approach was devel-
oped by Karr et al. (1986) for the Index of Biotic
Integrity (IBI). It has since been applied to
estuary assessment (Engle et al. 1994,
Ranasinghe et al. 1994) and reservoir assess-
ment (TVA 1994).
A representative sample of lakes is taken from
the entire population. Sites thai are known to be
severely impaired may be excluded from the
sample, if desired. The population distribution
of each biological metric (Chapter 5) is deter-
mined, and the 95th percentile of each metric is
taken as its reference value. The range from the
minimum possible value (usually 0) to the
reference value is trisected, and values in the
top third of the trisected range are taken to be
similar to reference conditions. Scoring of
metrics is explained more fully in Chapter 6.
A central assumption of the population ap-
proach is that at least some sites in the
population of lakes are in good condition,
which will be reflected in the highest scores of
the individual metrics. Because there is no
independent definition of reference (indepen-
dent of biological status), reference conditions
defined in this way must be taken as interim
4-9
-------
Chapter 4
and subject to future reinterpretation. Again,
antidegradation safeguards must be in place to
prevent deterioration of the reference standard.
Periodic examination of the reference stan-
dards for trends can detect deterioration or
improvement. Strictly speaking, the distribu-
tional approach is circular because the refer-
ence biological conditions are characterized as
the best of existing biological conditions,
without consideration of impacts. This is
necessary when reference criteria cannot be
defined a priori, or when all lakes under consid-
eration are equally impaired. The object of the
method is to develop a measurement standard
for assessment of lakes. Its validity must then
rest on external confirmation of the response
of metrics to stressors, usually from published
or other independent studies.
Following the initial classification of the lakes
in a region, biota are surveyed to determine
those aspects of the classification that are
relevant in explaining biological variability
among lakes. The objective of the survey is to
determine the final classification and to
characterize the biota of each of the lake
classes. Analysis of biological data includes
testing classes developed in the initial classifi-
cation, as well as aggregating classes as
necessary to obtain a parsimonious classifica-
Case Study: Ecoreglonal Classification of Minnesota Lakes
Minnesota has over 12,000 takes spread across di-
verse geographic areas. Previous studies had
shown distinct regional patterns in lake productivity
associated with regional differences in geology, veg-
etation, hydrology and land use (Heiskary and Wil-
son 1889). Four of the seven ecoregions in Minnesota
(Omemik 1987) contain 98 percent of the lakes. These
am the Northern Lakes and Forest (NLF), North Cen-
tral Hardwood Forest (NCHF), Northern Glaciated
Plains (NGP), and Western Com Belt Plains (WCBP)
(Figure 4-1). Minnesota has used environmental dif-
ferences along with regional differences In lake uses
to develop ecoregion-based frameworks for data
analysis, developing monitoring strategies, assess-
ing use patterns, and developing phosphorus goals
7 and criteria (Heiskary 1989).
The Minnesota Pollution Control Agency (MPCA)
and several other groups collected data on chloro-
phyll a concentrations and several water quality
parameters (total phosphorus, total nitrogen, and
- Secchl transparency) in 90 reference lakes between
1985and 1987. Secchl transparency data were col-
lected mostly by volunteer participants in the Citi-
zen Lake Monitoring Program. Reference lakes were
chosen to represent minimally impacted sites within
each ecoregion. Criteria used in selecting reference
lakes included maximum depth, surface are, fish-
ery classification, and recommendations from Min-
nesota Department of Natural Resources (DNR)
(Heiskary and Wilson 1989). Lake morphometry had
previously been examined. In addition to the refer-
ence lake data base, MPCA examined a statewide
data base containing data collected by these same
groups on approximately 1,400 lakes from 1977 to
1987.
Differences in morphology, chlorophyll a concentra-
tions, total phosphorus, total nitrogen, and Secchi
transparency were found among the 4 ecoregions
in both studies. Lakes in the 2 forested ecoregions
(NLF and NCHF) are deeper (median maximum
depth 11 m) with slightly smaller surface areas (40
to280ha) than those in the plains ecoregions (NGP
and WCBP). Lakes in the 2 plains ecoregions were
typically shallow (median maximum depth 3 m) with
larger surface areas (60 to 300 ha).
Box-and-whisker plots for chlorophyll a and water
quality measurements in the reference lake study
paralleled the morphological differences seen
among the ecoregions (Heiskary and Wilson 1989).
The 2 plains ecoregions had significantly higher
chlorophyll a levels than either of the 2 forested
ecoregions (Figure 4-2). Another biological param-
eter, ecological classification, also differs among the
ecoregions. Ecological classification refers to the
type of fish assemblage likely to bo present if no
fisheries management occurred, in the forested
ecoregions, 37 percent to 48 percent of the lakes
are classified as "basspanfish walleye" (Heiskary et
ai. 1987). Additionally, only the 2 forest ecoregions
support any lakes classified as *walleye." Results
of the statewide data base analysis showed these
same trends. The results of these 2 data base analy-
ses support the use of ecoregions in developing
frameworks for data analysis, monitoring strategies,
assessing use patterns, and developing phospho-
rus goals and criteria.
4-10
-------
Selection and Characterization of Reference Conditions
Case Study: Eeomgloiial Classification
of Minnesota Lakes (continued)
IQKTH1RN
M* FORESTS
NORTHefcNj" **
QUtCtATeg __
plain*
WHTERN CORN MLT PLAINS
Figure 4-1. Minnesota ecoregions and sampled
lakes. From Heiskary 1989.
182
CHLOROPHYLL a (no/1)
msari
LEGEND
120
100
10
80
40
NLF
NCHF
WCBP
NOP
Figure 4-2. Chlorophyll a concentration of Minnesota reference lakes byecoregion. Notches In
box pilots represent 95 percent confidence intervals of the medians. NLF = Northren Lakes-arid
Forests; NCHF = North Central Hardwood Forests: WCBP = Western Com Belt Plains; NGP =
Northern Glaciated Plains.(From Heiskary 1989.) , "
4-11
-------
Chapter 4
tion that accounts for the greatest amount of
biological variability. The survey may use
existing data, although a new survey allows
careful selection of reference sites representative
of each of the classes of lakes.
4.2.3 Paleollmnology
An alternative to characterizing present-day
reference conditions is to estimate historic or
prehistoric pristine conditions. In many lakes,
presettlement conditions can be inferred from
fossil diatoms, chrysophytes, midge head
capsules, cladoceran carapaces, and other
remains preserved in lake sediments (e.g.,
Charles et al. 1994, Dixit et al. 1992). Fossil
diatoms are established indicators of historical
lake alkalinity, salinity, and trophic state (e.g.,
Hall and Smol 1992). Diatom frustules, com-
posed of silica, are typically well preserved in
lake sediments and easy to identify. However,
remains of other organisms are problematic
because of incomplete preservation.
Paleolimnological investigations can be per-
formed in lakes in which identifiable remains
are preserved, and the
sediments can be dated to
the period of interest
Identify presettlement (Charles et al. 1994). In
some lakes, sediments are
conditions (reference subject to scouring,
resuspension, or periodic
conditions) for an Indi- drying and are not suitable
for coring. Most lakes have
a quiescent depositional
lakes within a region (e.g., area in the deepest
profundal waters, and
Cummlng et al. 1992). these sediments receive
material from both pelagic
Paleollmnology can
vldual lake or for many
and littoral zones, as well as from the sur-
rounding watershed. Reservoirs meeting the
depositional criteria can also be analyzed in
this way, yielding a history of the reservoir.
However, historical conditions in a reservoir
might or might not be a desired reference
condition.
Design of paleolimnological studies to deter-
mine reference conditions can range from basic
to complex. The simplest procedure is to
analyze only the top and bottom of a sediment
core, and to make a comparison of assemblages
to determine if there has been a significant
shift in taxa composition. If there is little
difference, then there has probably been rela-
tively little change in major ecological character-
istics in the lake. If there are significant
differences, then further investigation may be
warranted, including quantitative inference of
past water chemistry conditions (Charles and
Smol 1994). The more informative approach is to
analyze several sediment intervals from a
sediment core that has been dated (usually Pb-
210), and infer specific past conditions. This
design leads to understanding of the magnitude,
rate, and timing of change and can be related to
specific watershed or in-lake events.
Using paleolimnology to characterize lake
reference conditions requires selection of a time
period for the reference. In general, the time
period should be as close to the present as
possible when anthropogenic impacts on the
lakes were minimal. If there is concern that
background conditions may have varied sub-
stantially, a few to several presettlement time
periods could be analyzed to determine natural
variability. In most cases this variability is
relatively small compared with changes following
European settlement.
The greatest advantage of paleolimnology is that
a sample of reference sites can be selected
without regard to present conditions in the lakes.
Thus, there is usually no need to select "least-
impaired" lakes because nearly all lakes in the
selected reference period are least-impaired by
definition. Reference sites are selected such that
each lake class has at least 5 to 10 representa-
tive lakes. Reference sites should be representa-
tive of their respective class. Transitional,
exceptional, or uncertain lakes should not be
included in the reference sample.
The population approach to defining reference
conditions means that a single site is never
taken as a representative reference for an entire
class. Similarly, the condition at only 1 time
period of a single lake may not represent a
reference for its present condition. Ecosystems
are not constant in time, even in the absence of
disturbance, and the condition of a single lake
is likely to change in the course of a century.
Therefore, samples of past conditions at several
points in time are more likely to characterize
reference conditions than a single sample.
Sampling and Data Analysis—Sediment diatoms
are the recommended assemblage for
paleolimnological determination of reference
4-12
-------
Selection and Characterization of Reference Conditions
conditions because preservation of frustules is
excellent and identification is based solely on
the frustules. Other assemblages (e.g., cladocer-
ans, midges) are not recommended at this time
because preservation is incomplete and identifi-
cation of fragments is problematic. Cores are
taken from the representative lakes and ana-
lyzed as described in Appendix C.
4.2.4 Historical Data
Some lakes have extensive historical data bases
from the early to mid-20th century, typically on
water quality, diatoms, zooplankton, or fish.
However, historical data may not represent
undisturbed conditions, and the biological data
and auxiliary historical information should be
examined carefully to ensure that the data
actually represent conditions better than at
present. Cultural eutrophication has occurred
since neolithic peoples first settled on
lakeshores, and in many American waterbodies
cultural eutrophication was most pronounced in
the 1950s and 1960s.
Historical data might not always be representa-
tive of lakes in a region because the lakes were
selected for special reasons (e.g., unique lakes,
near laboratory, site of water intake, etc.).
Universities, municipal water supply depart-
ments, and other agencies are often good
sources of long-term lake water quality data. It
might be possible to augment present-day
reference site data with historical data.
The greatest advantage of
4.2.5 Modeling Approaches
Several modeling approaches can be used,
including mathematical models (logical con-
structs following from first principles and
assumptions), statistical models (built from
observed relationships between variables), or a
combination of the 2. The degree of complexity
of mathematical models to predict reference
conditions is potentially unlimited, with
attendant increased costs
and loss of predictive
ability as complexity
increases (Peters 1991). paleolimnology is that a
Mathematical models are
complex and untestable sample of reference sites
hypotheses (Oreskes et al.
1994, Peters 1991). Never- can be selected without
theless, models to predict
water quality in rivers and
reservoirs from first tions In the lakes.
principles of physics and
chemistry have been quite
successful (e.g., Kennedy and Walker 1990).
Statistical models can be fairly simple in
formulation, such as the Vollenweider model,
the Morphoedaphic Index, and others (Vighi
and Chiaudani 1985, Vollenweider 1975,
Mazumder 1994), to predict trophic status, but
they require a sufficiently large data base to
develop predictive relationships. If enough data
exist to construct a statistical model, it is
likely that there are lakes that can serve as
reference sites.
regard to present condi-
4-13
-------
Chapter 4
Case Study: Reference Conditions - TVA Reservoirs
RESERVOIR
(For TVA's reservoir bioassessment, see Chapter
3-)
It was not possible to use the well-accepted ap-
proach of using least-Impacted reference sites to
determine characteristics or expectations of a res-
ervoir since they are artificial systems. Other ap-
proaches must be used such as historical or
prelmpoundment conditions, predictive models, best
observed conditions, or professional judgment.
Prelmpoundment conditions are clearly inappropri-
ate. For the most part, models are of limited value
for a large variety of indicators because of such great
spatial and temporal variations within arid between
reservoirs. This leaves best observed conditions or
professional Judgement as the most viable alterna-
tives for establishing appropriate reference condi-
tions or expectations for reservoirs. TVA's experi-
ence has found use of best observed conditions
using professional judgement as the best approach.
In using best observed conditions one assumes that,
for the group of reservoirs to be compared, the range
of observed values represents the range of expected
conditions from good to poor for each community
characteristic or metric included in the evaluation.
Separation of reservoirs into appropriate classes
was a critical step In developing reference condi-
tions.
For dissolved oxygen (DO) and sediment quality,
best observed conditions were not used; instead,
Ideal conditions were expected. That is, poor DO is
unacceptable regardless of type of reservoir or dam
operation. Sediments should not have high concen-
trations of metals, should have ho or very low con-
centrations of pesticides, and should not pose atoxic
threat to biota. In this situation, there Is no need for
classification because the same conditions are de-
sired for all reservoirs.
For chlorophyll, benthos, and fish, the best observed
conditions approach was used. For these, reservoirs
were categorized because the same conditions do
not exist for all reservoirs. The classification scheme
that evolved for chlorophyll is actually a combina-
tion of two approaches: examination of the "natural"
nutrient level in the watershed; and a conceptual/
subjective decision as to the concentrations Indica-
tive of good, fair, and poor conditions. Two classes
of reservoirs were developed: reservoirs draining
nutrient-poor watersheds, primarily those In the Blue
Ridge Ecoregion; and the mainstream reservoirs
with their remaining tributary reservoirs.
For the benthic macroinvertebrate and fish assem-
blages, reservoirs were divided into four classes:
• Reservoirs on the Tennessee River plus two
navigable reservoirs on tributaries to the Ten-
nessee River, tills group of reservoirs has rela-
tively short retention times and little winter draw-
down.
• Reservoirs in the Blue Ridge Ecoregion.
• Reservoirs in the Ridge and Valley Ecoregion.
• Reservoirs in the Interior Plateau Ecoregion.
4-14
-------
in This Chapter..*
> Watershed Activities
> In-Lake Water Quality
> Shorezone and Littoral Characteristics
Chapter 5
Habitat Measurement
Habitat measurement is used to assess the
impacts of habitat on biota, and hence on the
interpretation of changes in biota. Habitat must
be taken Into account to make accurate com-
parisons between ambient and reference condi-
tions and to determine whether habitat might be
a cause of impaired biota.
Human activities modify the watershed, with
consequent effects on lake physicochemical and
biological processes. Agricultural and urban
land use affect nutrient, contaminant, and
sediment loadings; and shorezone housing
development can have a disproportionate
influence on nutrient loadings compared with
more distant parts of a lake watershed (Dillon et
al. 1994). Shorezone development can also
extend into the lake littoral zone with construc-
tion of docks, revetments, riprap, often leading to
destruction of littoral wetlands and macrophytes.
The habitat experienced by aquatic organisms
consists of the water and the substrate, includ-
ing structure and constituent chemicals. For the
purposes of this protocol, water quality is a
component of habitat. In-lake habitat includes
both the physical and chemical environment
experienced by the biota, and is, in turn,
influenced by the watershed through runoff and
loadings. Habitat measurement seeks to identify
the physical and chemical characteristics of the
lake habitat—both natural
and anthropogenic—that
affect the biota of the lake.
Habitat measurement is
used to assess the impacts
of habitat on biota, and
hence on the interpretation
of changes in biota.
Habitat measurement,
consisting of both water-
shed and in-lake observa-
tions, has two purposes.
First, it helps in placing a
lake into a category
determined by a classifica-
tion scheme. Second, it can help identify an-
thropogenic disturbances and exposure that
might be responsible for biological degradation.
Habitat measurement thus comprises two kinds
of variables:
Classification variables—Those attributes
intrinsic to the system and relatively unaffected
by human activities (e.g., geology, soils, lake and
watershed morphology).
Assessment variables—Those attributes which
either are direct measures of human activity
(e.g., land use, discharges) or are influenced by
human activity (e.g., most water quality vari-
ables).
5-1
-------
Chapters
The classification variables are those which are
not affected by human Influence, and are
primarily measures of the morphology and
geology of the lake and watershed. The classifi-
cation variables assist in placing the lake into
one of the categories for
which reference conditions
The purpose of examining have been determined. It is
then possible to determine
the deviation of conditions
In the test lake from
reference conditions, for
lake. both habitat and biological
indicators.
watershed parameters is
to assist in classifying a
Several habitat parameters are obtained or
estimated from existing sources of information
such as maps and Geographic Information
Systems (GIS). The parameters include lake
area, depth, shoreline length, watershed area,
watershed slope, soil types, geology, and water-
shed land use.
The habitat measurement component of the field
sampling program consists of in-lake physical
and chemical measurements, as well as a
shorezone habitat survey. The shorezone survey
Is based on the Environmental Monitoring and
Assessment Program (EMAP) lake habitat
assessment (USEPA 1994a, USEPA 1994b,
USEPA 1993a).
6.1 WATERSHED HABITAT
6.1.1 Measurements
The purpose of examining watershed parameters
is to assist in classifying a lake and to determine
whether watershed conditions might account for
observed biological status. A number of human
practices in lake watersheds affect lake habitat
through sediment loading, nutrient loading,
contaminant loading, hydrologlc changes, and
direct habitat alteration (e.g., removal of wet-
lands). Any one human activity can influence
several loading rates. For example, livestock
management practices can affect both nutrient
and sediment loads. Watershed parameters
include both classification and assessment
variables fTable 5-1). Most measures of mor-
phology and land use can be obtained from
USGS, state, or county data bases.
5.1.2 Watershed Metrics
Discharges—Data from permitted discharges can
be used to develop direct estimates of point-
source loadings into receiving waters, and they
take into account the effects of sewage diver-
sions and implemented control technologies.
However, discharges cannot account for
nonpoint sources.
Watershed Area—The quantity of runoff entering
a lake is directly affected by the lakes watershed
area. The ratio of lake watershed area to lake
surface area affects sediment and nutrient
loadings and retention time. Reservoirs with a
small ratio are better able to support sport fish
populations (Hill 1986). The ratio is especially
important for reservoirs and flowage lakes,
where its value can vary widely.
Land Use—Water quality, especially nutrient
concentrations and turbidity, is strongly associ-
ated with land use. The most important land
use variables are urban, agricultural, and forest
land use, as percent of the watershed area. Also
important is watershed road density (length per
area), which can be an excellent predictor of
trophic variables and chloride concentration
(USEPA 1993a). More detailed breakdowns of
land use classes (e.g., high-density urban,
transportation, pasture, row crops, etc.) can be
estimated for diagnostic investigation.
A detailed nonpoint source evaluation might be
called for if more than one land use type ap-
pears to be a probable cause for impairment. A
standard screening procedure (Schueler 1987)
can be applied to estimate sediments, nutrients,
and contaminants from both urban and
nonurban sources. The screening procedure
allows identification of the primary likely
sources of impairment and hence a preliminary
ranking of potential sources.
The land use variables are tabulated on a
watershed-wide basis. This approach does not
take into account the effects of distance from
the receiving waters, riparian buffers, or best
management practices (BMPs). Runoff and
pollution of surface waters from agricultural
land are highly variable, depending on slope,
soil erosivity, tillage practices, distribution of
rainfall, and the presence of riparian buffers
and hedgerows (Schueler 1987). Taking into
account riparian buffers and BMPs, together
5-2
-------
Habitat Measurement
with other watershed influences, would require
a comprehensive runoff and loading model, and
is beyond the scope of this guidance.
Population Density arid Related Measurements—
Nonagricultural pollution is the product of
people and their activities; hence, population
density is an excellent predictor of pollutant
loadings. Population density is also strongly
correlated with urban land use and discharges;
therefore, simultaneous assessment with these
collinear variables should be done with caution.
Population density might be a more accurate
indicator of total human activity than is land
use, because population estimates are updated
more frequently than land use data. The
variables that most directly affect lake quality
are discharges and the watershed impervious
area. Nevertheless, population density may be a
better single measurement.
>.2 IN-LAKE HABITAT
6.2.1 Measurements
Physical-chemical habitat measurement com-
prises several common measures of lake water
quality and can point to water quality problems
that are not observable at the coarser resolution
of the entire watershed. It can also provide
additional evidence for potential causes identi-
fied from the watershed or shoreline assess-
ment. Physical and chemical parameters and
the measurements derived from them are listed
in Table 5-2,
Secchi Depth—Secchi depth, which has a long
history as a lake assessment variable, is a
simple and reliable measure of light transmit-
tance and turbidity. It is used in various trophic
indices, including Carlson's Trophic State Index
(TSI) (Carlson 1977).
Nutrients—Water quality measurements can
form the basis of several measurements (Table
5-2). Total phosphorus concentration forms part
of Carlson's TSI (Carlson 1977) and is an
important predictor of lake productivity in north
temperate lakes (Vollenweider 1975). The
nitrogen-to-phosphorus (N:P) ratio is used to
predict the likelihood of cyanobacteria blooms
(e.g., Smith 1983). Calculation of trophic state
indices is given in Section 7.2.3.
Population density is an
excellent predictor of
pollutant loadings.
Dissolved Oxygen—Dissolved oxygen (DO) is
necessary for aquatic life,
and most state water
quality regulations include
a standard for dissolved
oxygen, usually expressed
as the maximum amount
of time that DO is allowed
to fall below a critical value (typically 4 or 5mg/
L), Several measurements have been developed
for DO, including:
• Index period DO measurement near bottom
of lake.
• Depth from the surface at which DO falls
below a threshold value (oxycline) (Scott et
al. 1991).
• Annual or seasonal minimum value in
hypolimnion or epilimnion.
• Annual or seasonal mean value in hypolim-
nion or epilimnion.
• Annual or seasonal percent time below a
threshold DO value at the bottom of the lake
(USEPA 1993a).
• Annual or seasonal mean water volume or
percent of total volume below a threshold
DO value (Dycus and Meinert 1992),
The first 2 measurements require only a single
DO profile. Depth of the oxycline might be the
most useful single-point DO measurement of a
waterbody (Scott et al. 1991), provided that the
observation is made when hypoxia is at its
maximum annual extent (usually late summer).
The remaining 4 measurements all require
regular observations during a year or an index
season. In general, estimates of time or volume
below a threshold value are more precise and
accurate than estimates of minimum values
(USEPA 1993a).
5.3 SHOREZONE AND
LITTORAL HABITAT
5.3.1 Measurements
The shorezone habitat assessment is important
for identifying potential causes of impairment
because many lakes are impacted by develop-
5-3
-------
Chapters
merit and land use on the shore. Because the
lakeshore is the part of the watershed closest to
the lake, shorezone land use has the largest
potential impact on lake biological integrity. The
shorezone assessment procedure is the same as
that for watershed evaluation: shorezone habitat
variables are compared to reference conditions
and, if significantly different, are identified as
probable causes of biological impairment.
EMAP Surface Waters has developed an exten-
sive shorezone and littoral survey methodology
to characterize riparian and littoral habitat
(USEPA 1994a, USEPA 1993a, USEPA 1991e).
The index period is late summer when vegeta-
tion is at its annual maximum. The riparian
characterization consists of estimates of domi-
nance of vegetation in canopy, understory, and
groundcover; substrate type; bank angle; and
dominance of human features (buildings, lawns,
cultivation, etc.). Littoral characterization is
done at a 10m distance from shore and includes
depth, surface film, substrate, macrophyte
cover, fish cover, and a summary habitat
classification (USEPA 1993a). The shore of each
lake is surveyed at 10 sites, and the frequency
of disturbance is estimated for each lake from
the survey data.
The shorezone and littoral assessment for lake
biological surveys presented here is a modifica-
tion of the EMAP shorezone assessment (Table
5-3) (USEPA 1994a).
Measurements
Additional Metrics
Calculation
Indicator
Lake and Basin Morphology
Watershed
drainage area.
Estimated from map
contours.
Hydrology
Lake surface area.
Map
Watershed: Lake
area ratio.
Watershed area/lake area.
Sediment, nutrients.
Shoreline length.
Shoreline
development ratio.
Effect of riparian zone.
Lake volume.
Estimated from Basin
contours.
Maximum depth.
Measurement
Stratification potential.
Mean depth.
Volume/surface area.
Mean basin slope.
Lake outflow.
Retention time.
Volume/outflow.
Eutrophication potential.
P
3
"*3
C
3
% forest or natural
vegetation.
Sediment, nutrients,
hydrology.
% agriculture.
GIS data base.
Sediment, nutrients,
contaminants.
% urban and
residential.
Sediment, nutrients,
contaminants, hydrology.
Watershed
impervious surface.
Estimate from land use.
Sediment, contaminants,
hydrology.
Population
density.
U.S. Census, state or
county.
Sediment, nutrients,
contaminants, hydrology.
Discharges
USEPA NPDES data base.
Nutrients, contaminants.
Road density.
Maps, GIS.
Sediment, contaminants,
hydrology.
Table 5-1, Watershed and basin habitat measurement and metrics.
5-4
-------
Habitat Measurement
Table 5-2. Physical and chemical measurements and metrics.
Measurements
Metrics
Calculation
Indicator
T Profile
Epilimnion temperature.
Mean from temperature profile.
Hypolimnion temperature.
Mean from temperature profile.
Metalimnion depth.
Inflection point of temperature
profile.
DO Profile
Epilimnion DO.
Mean from DO profile.
Hypolimnion DO.
Mean from DO profile.
Oxycline depth.
Depth at which DO falls below
2 mg/L.
DO problems.
Hypoxic volume.
Volume of water with DO <2
mg/L; annual or seasonal mean.
DO problems.
Secchi Depth
(SD)
TSI (SD) = 60 -14.41 In (SD)
Transparency
Total N
TSI (N) = 54.45 + 14.43 In (TN)
N enrichment.
Total P
TSI (P) = 4.15+ 14.42 in (TP)
P enrichment.
N:P ratio.
N concentration/P concentration
(molar).
Enrichment
Silica
¦¦ .V,;~
Depletion
Acid neutralizing
capacity (ANC)
ANC
Sensitivity to
acidifcation.
PH
pH
Acidity
Total Dissolved
Solids (TDS)
TDS
Dissolved
minerals.
5.3.2 Shorezone Metrics
Most shorezone measurements are means of the
littoral and shorezone habitat metric values. The
shorezone and littoral cover measurements are
expressed as the mean of the values of all
transects. The human influence measurements
are different because they are based on pres-
ence or absence observations within the
transects. These measurements are weighted,
with each present observation receiving a score
of 1 and each "adjacent" observation receiving a
score of 1/2. The human influence score in each
category is the mean of all transects. It is in the
range of 0-1, with 0 reflecting no influence and
1 indicating that the influence (e.g., buildings)
was found in every transect.
5-5
-------
Chapters
Table 5-3. Lakeshore habitat measurements and metrics (USEPA 1994a, EMAP Internal Report).
Habitat Measurement
Mean % Cover
Indicator
Bank Measurement
- Rocky (%)
- Soil (%)
- Vegetation (%)
- Other (%)
Mean % cover from shorezone habitat
transects.
Bank Stability.
Bank Erosion (0-4)
0=none
4=severe erosion
Riparian Vegetation Measurements
- Canopy - (% cover)
- Understory - (% cover)
- Ground Cover - (% cover)
% cover of vegetation.
Disturbance
Human Influence Measurements
Buildings
- In-lake structures
- Roads, railroads
- Agriculture
-Lawn
- Dump or landfill
Influence score (mean score of transects).
Presence/absence.
Human Influence.
5-6
-------
In This Chapter...
> Description of Assemblages
> Response to Stress
> Discussion of Assemblage Analysis
> Level of Sampling Effort
Chapter 6
f IJiOt^PS
The proposed biological sampling methodology
is tiered, ranging from a trophic state assessment
to detailed biosurveys. Many of the methods are
based on those used in USEPA's Clean Lakes
Program and Environmental Monitoring and
Assessment Program (EMAP) lakes component.
Lake surveys require sampling of biological
assemblages and habitat in one or more field
visits. Several of the proposed lake biosurvey
measurements are made from transects extend-
ing from the shore to the sublittoral habitat, and
several other measurements are made from one
or more stations in the pelagic region of the lake
(Chapter 7). The integrated sampling scheme
combines all sampling activities carried out on
the transects and includes mid-lake sampling
sites for pelagic samples. The number of
transects, the number of sampling sites, the
assemblages sampled, and the frequency of
sampling vary among the survey tiers.
The study of any group of organisms will yield
information on the status of their environment.
The objective in selecting assemblages for lake
bioassessment was to find assemblages that:
* Are unambiguously useful for biological
assessment.
• Can be sampled and interpreted in a cost-
effective way.
• Are consistent with the current mix of
expertise in natural resource agencies.
• Can be easily converted to a multimetric
index of the assemblage.
The recommended assemblages are phy-
toplankton, sedimented diatoms, submerged
and floating aquatic macrophytes, crustacean
zooplankton, benthic macroinvertebrates, fish,
and periphyton. The discussion of each
assemblage includes some estimates on the
level of effort required for sampling. These are
intended as general guidelines. Actual time
and effort involved will depend on the specific
expertise and resources available to individual
agencies.
Emergent vegetation is not included as an
assemblage in this document because meth-
ods for emergent plants are under develop-
ment by USEPA and other agencies as part of
the development of wetlands bioassessment
methods. Several other potential assem-
blages were not considered because there was
little information on their utility as environ-
mental indicators for lakes. They included
benthic meiofauna, protozoa, and bacteria.
Background and rationale for the selected
assemblages are presented in Appendix D.
6-1
-------
Chapter 6
6.1 PRIMARY PRODUCERS:
TROPHIC STATE
ASSESSMENT
Phytoplankton are the base of most lake food
webs, and fish production is linked to phy-
toplankton primary production (Ryder et al.
1974). Excessive nutrient and organic inputs
from human activities in lakes and their water-
sheds lead to eutrophication, characterized by
increases in phytoplankton biomass, macro-
phyte biomass, nuisance algae blooms, loss of
water clarity from increased primary production,
and loss of oxygen in bottom waters. From a
human perspective, eutrophication problems
might include loss of aesthetic appeal, decreases
in desirable gamefish, loss of accessibility due to
increased macrophyte production, and in-
creased cost of treating drinking water.
Trophic state is assessed with 4 Trophic State
Indices (TSI)—chlorophyll a, Secchi depth, total
nitrogen, and total phosphorus (Carlson 1977,
Carlson and Simpson 1996)—and with Algal
Growth Potential (AGP) (i.e., nutrient availabil-
ity for algal growth). The chlorophyll TSI (Table
6-1) indicates whether algal biomass is low,
medium, or high; the Secchi TSI indicates if
algal growth may be limited by mineral turbid-
ity; and the nutrient TSIs can indicate excess
or limiting nutrient supply.
Level of Effort
Trophic state assessment is relatively inexpen-
sive. Sample collection requires approximately
10 minutes on station and can be done by a
single person. Filtration of chlorophyll samples
requires another 10 to 15 minutes in the field.
Chlorophyll, nutrients, and other water quality
chemical analyses are standard and costs are
well established in each region.
Table 6-1. Potential algal trophic state metrics.
6.2 SUBMERGED
MACROPHYTES
Macrophytes form an integral part of the littoral
zone of many lakes, providing cover for fish and
substrate for invertebrates. From a human
perspective, overabundant macrophytes (or
weedy conditions) can interfere with lake access
by fouling equipment, interfering with recre-
ational activities, and detracting from aesthetic
appeal. A conspicuous lack of native macro-
phytes in habitats where they are expected to
occur can result in reduced population of sport
and forage fish and waterfowl (Crowder and
Painter 1991). Potential macrophyte metrics are
listed in Table 6-2.
Level of Effort
Submerged macrophyte analysis, including an
estimate of total percent cover and identification
of dominant species, requires approximately 1 to
2 hours in the field for a 300- to 500-acre lake.
There is no laboratory analysis.
For the same size lake, macrophyte density or
biomass measurements would require 2 to 4
hours in the field to collect samples, and to sort
and weigh by species. Stem counts would likely
require a longer time. Again, there is no labora-
tory analysis.
6.3 SEDIMENTED DIATOMS
Phytoplankton cells continually grow and die,
and dead cells sink to the bottom. One group of
algae, the diatoms, have shells (called frustules)
made of silica (glass), which are preserved when
the dead cells fall to the lake bottom. The
preserved diatoms provide an integrated record
Metric
Response to Stress
Chlorophyll concentration.
Elevated under eutrophication.
Chlorophyll TSI.
TSI (CI) = 30.6 + 9.81 In (Chi)
Depressed under non-algal turbidity or toxicity (compared to
Secchi and nutrient TSIs).
Algal growth potential (AGP).
Increases with nutrient concentration.
6-2
-------
Biological Assemblages
Table 6-2. Potential macrophyte metrics.
Metric
Response to Stress
Total vegetated area (% of littoral).
Substantially more or less than reference.
% exotics or weedy species.
More than reference.
No. of exotic species.
High
Density or blomass in vegetated areas.
Substantially more or less than reference.
No. of taxa.
Low
% dominant species (by weight).
High
Maximum depth of plant growth.
Reduced under enrichment, deeper under acidification.
of the diatom assemblage in the lake. A sample
of the top 1 to 2cm of lake sediment contains a
representative sample of diatoms from the
most recent 1 to 3 years. Sedimented diatoms
can be sampled at any time and will always
yield a sample representative of the most
recent years. Potential sedimented diatom
metrics are listed in Table 6-3.
Level of Effort
Sedimented diatoms are sampled rapidly in the
field. A sample requires approximately 1 hour to
prepare and 2 to 4 hours to count and identify
300 to 500 cells.
6.4 BENTH1G
MACRO INVERTEBRATES
Benthic macroinvertebrates are long-term
indicators of environmental quality; they inte-
Table 6-3. Potential sediment diatom metrics.
grate water, sediment, and habitat qualities
(USEPA 1989b, USEPA 1990d). Macroinvcrte-
brate species have sensitive life stages that
respond to stress and integrate effects of both
short-term and long-term environmental
stressors. Classification of benthos according
to their relative sensitivity to pollution and
their functional feeding group level differenti-
ates effects on ecological health in response to
organic or toxic perturbations. Potential
metrics are listed in Table 6-4.
Macroinvertebrates are sampled from the
predominant substrate available in the sublit-
toral zone. The type of sampling gear will
depend on the substrate being sampled: each
substrate has its own optimal sampling gear
(Chapter 7).
Level of Effort
A benthic sample, consisting of several grabs,
requires 2 to 4 hours in the field. Sorting,
Metric
Response to Stress
No. of taxa.
Reduced
Diversity indices (Shannon-Weiner, Simpson's, etc.).
Reduced
% dominant taxon.
Increased
% centric diatoms.
Reduced
Pollution tolerance indices (e.g., Lange-Bertalot;
Bahls 1993).
Lower score under organic pollution.
% Nitszchia and Navicula (Bahls 1993).
Increased with sedimentation.
Indicator taxa (ecological categories).
Respond to specific stressors (acidity, salts,
metals, eutrophication).
Disturbance index (Dixit and Smol 1994).
increased
6-3
-------
Chapter 6
Table 6-4. Potential benthic macroinvertebrates metrics.
Metric
Response to Stress
No. of taxa.
Reduced
Mean number of individuals pertaxon.
Substantially lower or higher.
% contribution of dominant taxon.
Elevated
Shannon-Wiener diversity.
Reduced
% intolerant species.
Reduced
% oligochaetes.
Elevated under organic enrichment.
ETO taxa (ephemeroptera, trichoptera, odonates).
Reduced under enrichment, DO, stress.
% non-insects.
Reduced
Crustacean + mollusc taxa.
Reduced under acid stress.
% crustaceans and molluscs.
Reduced under acid stress.
Tolerance indices (e.g., HBI [Hilsenhoff 1987];
Hulbert's Lake Condition Index [LCI] [Frydenborg et
al. 1995]).
Reduced
% suspension feeders.
Reduced
% shredders.
Reduced under enrichment or in very large lakes.
counting, and identifying 100 organisms to
species requires approximately 4 to 6 hours in
the laboratory.
6.5 FISH
Fish assemblages include species that repre-
sent a variety of trophic levels (omnivores,
herbivores, insectivores, planktivores,
piscivores), and that exhibit a range of toler-
ance to water quality or habitat degradation.
Fish are long-lived and integrate short-term
temporal environmental changes, and also
integrate effects of lower trophic levels (e.g.,
primary producers and benthic
macroinvertebrates); thus, fish assemblage
structure is reflective of integrated environ-
mental health. Of all biological components of
lakes, fish probably receive the greatest public
attention because of sport and commercial
fishing and attendant concerns regarding fish
production success and safety for human
consumption.
Fish are the most difficult and time consuming
of all assemblages to sample; are wide-ranging
and might not reflect local conditions in large
lakes; and are actively and intensively managed
by stocking and angling. Each feasible gear type
suitable for their sampling in lakes is highly
selective (USEPA 1994a, USEPA 1994b). Unbi-
ased sampling methods such as explosives,
rotenone, and draining a lake are generally too
destructive. Because of lake fish assemblage
sampling method bias, the use of a combination
of more than one gear type is recommended
(Chapter 7). Potential fish metrics are provided
in Table 6-5. Among the most promising mea-
surements are indicators of fish health (external
gross pathology) and fish tissue contamination.
Level of Effort
Fish populations are generally nonrandomly
distributed and clumped in response to habitat
variables; therefore, the choice of sampling
methods and equipment, index period, and
sampling frequency depend upon waterbody
physical characteristics and specific study
objectives. An understanding of the attributes
and/or biases of sampling equipment and
methods used in fish assemblage surveys is
essential in order to draw valid conclusions from
the data. The relative labor intensity of fish
sampling techniques varies greatly, depending
on the specific method chosen and the abun-
dance and diversity of the catch. For example,
6-4
-------
Biological Assemblages
passive techniques (e.g., trap nets, gill nets,
etc.) generally require a deployment and
capture cycle of many hours (e.g., overnight
sets) to several days; and the processing of
catches from either passive or active sampling
techniques may require several hours depend-
ing on local abundances and method efficien-
cies.
6.6 PHYTOPLANKTON
ASSEMBLAGE
Phytoplankton assemblage data, consisting of
taxonomic identifications and abundances
(relative or absolute) can be analyzed in two
ways: by determining assemblage measure-
ments based on species structure or by perform-
ing multivariate assemblage analysis. Potential
phytoplankton metrics are listed in Table 6-6.
The recommended approach is to sample the
phytoplankton assemblage and to count and
identify cells to order or genus. Simplified field
and laboratory procedures are possible for
measurements based on higher taxonomic levels
such as division or order. Identification to
species is considered supplemental at this time
because it is not clear that the information
gained represents a substantial improvement
over higher levels of taxonomy.
The phytoplankton assemblage requires 4 to 10
samples during the growing season to obtain a
seasonal average of the phytoplankton assem-
blage. The exact number will need to be deter-
mined from preliminary or existing data sets.
Level of Effort
An integrated water column sample of phy-
toplankton requires approximately 10 minutes
to collect. Laboratoiy identification and counting
of 300 to 500 cells requires 1 to 4 hours in the
laboratory.
6.7 ZOOPLANKTON
In most lakes, zooplankton are the central
trophic link between primary producers and
fish. Zooplankton are ubiquitous in all lakes
and are quickly and easily sampled in the field.
Zooplankton species richness is reduced under
chemical stresses (Baker and Christensen
Table 6-5. Potential fish metrics.
Metric
Response to Stress
No. of taxa.
Reduced
No. of sunfish species.
Reduced
No. of sucker species
Reduced
No. of intolerant species.
Reduced
% tolerant individuals.
Increased
% piscivores.
Reduced
% ominivores.
Increased
% invertivores.
Reduced
% planktivores.
Increased
Reproductive
Reduced
Composition
Reduced
Total number of individuals.
Substantially different under stress.
Fish health (pathology)
- Lesions and deformations
- Histopathology
Reduced under severe organic pollution or contamination.
Tissue contaminants.
Elevated under contamination (e.g., mercury, organochlorines).
6-5
-------
Chapter 6
1991), and abundant large Daphnia are associ-
ated with clear lakes with healthy sport fish
populations (Mazumder 1994). Trophic struc-
ture measurements require knowledge of
feeding of zooplankton species—trophic links
and complexity measures require the most
detailed knowledge. Potential zooplankton
metrics are shown in Table 6-7.
Zooplankton are sampled with vertical or
oblique tows, using a plankton net equipped
with a 7:1 reducing cone (DeBemardi 1984). The
recommended approach is to sample 4 to 6
times during a growing season to obtain
seasonal averages.
Level of Effort
A zooplankton sample can be collected in
approximately 10 to 30 minutes in the field.
Identification and counting of 100 to 200
organisms requires approximately 1 to 2 hours.
Six samples in a growing season per lake thus
requires six trips and 6 to 12 laboratory hours.
6.8 PERIPHYTON
Periphyton, the algae growing on solid sub-
strates (rock, wood, sediment, macrophytes),
have a long histoxy of use in bioassessment of
streams (Patrick 1949). Diatoms are often the
group of choice among periphytic algae. Ecology
of periphyton is much like other algal assem-
blages: they respond to nutrient enrichment;
they are cropped by grazers; and their species
composition is affected by pH, metal concentra-
tions, trace elements, and contaminants. In
addition, periphyton are affected by the physical
and chemical characteristics of their substrate.
Like phytoplankton, periphyton are subject to
changing water chemistry and seasonal succes-
sion. Several sampling periods may be neces-
sary to characterize lake periphyton.
Whereas periphyton have been used success-
fully in streams (Bahls 1993, Patrick 1949),
their application as lake indicators is relatively
new. Measurements of periphytic diatoms (Table
6-8) have shown promise for bioassessment,
based on investigation of undisturbed reference
lakes in Montana (Gerritsen and Bowman 1994),
but actual responses to disturbance or pollution
are as yet unknown.
Level of Effort
Analysis of a periphyton sample requires 2 to 6
hours, similar to diatoms and phytoplankton.
Table 6-6. Potential phytoplankton metrics.
Metric
Response to Stress
% cyanobacteria.
% greens.
Elevated under eutrophication.
% diatoms.
% chrysophytes.
Depressed under eutrophication.
% Anabaena, Aphanizomenon, Microcystis.
% centric diatoms (of total diatoms).
% pennate diatoms (of total diatoms).
% colonial greens (Volvocales).
% euglenophyta.
% dinoflagellates.
Blue-green algae and colonial greens
elevated under eutrophicaton.
No. of taxa.
Low under stress.
Diversity
Low under stress.
% dominance.
High under stress.
Lange-Bertalot index (pollution tolerance index, Bahls 1993).
Lower value under organic pollution.
Indicator taxa (presence or percentage).
Respond to specific stressors.
6-6
-------
Biological Assemblages
Table 6-7. Potential zooplankton metrics.
Metric
Response to Stress
% targe Daphnia (> 1 mm).
Low under planktivorous fish predation.
No. of taxa.
Reduced under contamination or stress.
% dominance.
High under stress.
Trophic structure measurements
- No. of trophic links
- Complexity measures
- % large predators
- No. of predator species
Simplified trophic structure under stress.
Case Study: Florida Metric Selection and Index Development
;; Of 32 potential macroinvertebrate metrics exam-
C tried, 9 were selected as candidate metric® for an
invertebrate'index for FldridaAes.:; Responsive
metrics are shown in Figure 6-1. Most metrics have
different values among the three lake types, with
sandy-bottom faked having the greatest
macroinvertebrate numberoftaxa(and other related
metrics), the greatestproportlon of OET(Odonata,
Ephemeroptera, Trichoptera) drganismSf and the
greatest
^ -gatherers (Fig. 6-1). -
'other.
The Shannon Index was strongty cqrretated with lotaJ
Utaxa and with dominance. Graphic exarrifnation of
the relationships among the metrics showed that the
Shmnon-totaltaxa relatloriship was not entirely lin-
" / ear, and! the Shannon-dominance relationship had
large and asymmettric yariamelnthe middle o f the
J. range (Fig. 6-2). Because of the virtahce and non-
r. linemity of the relationships, aM^caml^ate metrics
were retained
^Referenceand' test lakes also differedIrr water cob
, umn measurements (&g; &&)< Test lakes as a group
had higher chlorophyll ihncentratidns arid reduced
Secchi transparency, and higher total phosphorus
titan the b6ri^orK^;m&iB^:'k^0^i^Hming
increased trophic state in the. test lakes; Total
KjeldMnitr^m^shig^rtnsahdy.andmM-rnuCk
test lakes, and algal growth potential was increased
" in the'mud-muck lakes.
•> Eight invertebrate metrics appear responsive to lake
stiessors, and bah be used in a lake invertebrate
: index (number of taxa, Shannon-Wiener diversity,
; Hulbert index, OET taxa, percent dominance, per-
cent filterers, percent OET, and percent gatherers).
Two trophic state indicators (Secchi depth and chlo-
rophyll a concentration) also had characteristic val-
ues under reference conditions, and am best used
In conjunction with an invertebrate index to deter-
mine status of a lake.
i
JS
i
EJ
ii
J
«
s-
•
•
*
•
•
t * •
: T
I s I
rn '
pL •
x m ¦
0 £ •
U 4- '
Faf Tut
M TMl
Pel Test
santfy
transit
mud-muck
Figure 6-1. Metrics related to algal production,
Florida lakes (from Florida DEP 1994). a; algal
density (cells/ml) b: chlorophyll a c:
chlorophyll TSI
6-7
-------
Chapter 6
Cme Study: Florida Metric Selection and Index Development (continued)
5.0
4.5
4.0
3.5
f 3"°
S 2.5
| 2.0
w 1.5
1.0
0.5
0.0
&
¦ °o f
¦ * * * "
Sandy
Transitional
Muddy
A
A
A i
A
A%
A
¦ *
20 40 60
Percent Dominance
80
100
Figure 6-2. Relationship of bra highly
correlated attributes, Shannon
diversity and percent dominance
(from Gerrifsen and Whits 1907),
s<- *¦ f • ~ 4 •' r ¦ • *
# -L I ¦
"lijri"-}'
~ k«. • rwr »•. ¦
Figure 6-3. Distribution of nutrients,
secchl depth, and chlorophyll a In
Florida reference and test labs.
3.4
c
2,8
a>
I
2,2
c
1.8
£
CO
1.0
©
0.4
•0.2
0
JS
01
a,
S
o
X
o
o
Ul
<0
o
£
o
8 -
0 jr 0:
4' o
14 ¦
10
06 p-<
02 ¦ gfa 111
ili
[J] i
$ J
^ s;
, L^-i t^i ,
3 •
) •
j jIEl
3 LJ
5. ill:
I a I
. T—.... L—J
m i
Rel Tail
sandy
Ref Teat
transit
Ret Test
muddy
6-8
-------
In This Chapter*
> Tier Structure
> Study Design Consideration
Chapter 7
Tiered Sampling
This chapter provides general guidance for
designing a sampling program for lake
bioassessment. Four sampling tiers are sug-
gested options, and will need to be modified to
meet the sampling objectives, project resources,
and local conditions of individual programs.
Options for lake biological sampling include two
sampling tiers, each with an "A" and "B" field
component (Figure 7-1). There is also a desktop
screening process, with no field sampling.
Although sampling effort increases from Tier 1A
to Tier 2B, quality of information is not neces-
sarily related to sampling effort. Selection of a
sampling tier must be based on the objectives of
the biocriteria program. Tier 1 includes chloro-
phyll a and submerged macrophytes, and is
consistent with Clean Lakes Program sampling.
Tier 1 may be a single sample during a summer
index period (Tier 1A) or monthly sampling
during the growing season (Tier IB). Tier 2A
consists of assemblages that can be sampled a
single time during the index period: submerged
macrophytes, benthic macroinvertebrates, fish,
or sedimented diatoms. Tier 2B consists of
assemblages that are sampled several times
during the growing season: phytoplankton,
zooplankton, and periphyton. Both Tier 2A and
Tier 2B require Tier 1 sampling. Although Tier
2A and 2B were developed as an "either or"
choice, it is possible to perform both surveys as
the assemblages sampled in each do not over-
lap. It should also be understood that al-
though Tier 2B requires more effort and yields
a greater quantity of data, due to multiple site
visits, it does not necessarily produce better
data than Tier 2A. A supplemental habitat
assessment that includes diagnostic elements
(as detailed in Chapter 5) can be added to any
of the tiers.
7. 1 DESKTOP SCREENING
The desktop screening assessment involves
documentation of existing
data without any observa-
tions in the field (Table 7-
1). No assessment can be
better than the data that
go into it; therefore,
desktop screening alone
might be unreliable. Its
use should be limited to
planning for more detailed
monitoring and assess-
ment. It incorporates cost and time efficien-
cies, allowing evaluation of a large number of
sites, and identifying potentially affected areas
for further investigation using higher tiers.
Information is obtained from land use data
Options for lake biological
sampling include two
sampling tiers each with
an "A" and "B" field
component.
7-1
-------
Chapter 7
Tier lr-Trophic State and Nlacrophyte Assessment
Macrophytes:
Trophic State:
Water Quality.
Watershed variables:
(desktop)
Percent cover, density, exotic or native species
Chlororphyil a
Secchi depth, total nitrogen, total phosphorus, ortho-
phosphorus, DO and temperature profilas, total
dissolved solids, algal growth potential
Lake morphology, drainage area, land-use, cultural,
discharges, population density
; -d
Tier 1A
Mt
AWu AP MngBisfl
a#-. . -f.
A single visit during Index period.
Tier IB
Multiple visits during Index period.
Tier 2—Biological Assemblage Assessment
Shoreline habitat variables:
Vegetation cover, vegetation type, bank
composition, bank features, human Influence
Littoral habitat variables: Bottom substrate, emergent plant zone
TierlA
Tier IB
Tier IB
#s
3
g
g
Tier 2 A
P
g
g
I
L-J
Hi
LhmJ
iln
glevt
¦It during Iik
flex pi
LJ
irl
LJ
od
L_
I.
tier 2A Incorporates Tier 1A OR Tier IB
(habitat and biology), all components of the
shoreline and littoral habitat variables
(above), plus two or mom of the following
biological assemblages.
Macrophytes:
Macroinvertebrates:
Relative abundance,
species composition
Sublittoral, relative
abundance, species .
composition, deformities
Fish: Species composition,
health Index
Sedlmented diatoms: Surface sample
mm
Tier 2B
Multiple visits during Index period.
Tier 2B Incorporates Tier IB (habitat and
biology), all components of the shoreline
and littoral habitat variables (above),
plus one or more of the following
biological assemblages.
Phytoplankton: Relative abundance,
species composition
Zoopiankton: Relative abundance,
species composition,
size
Perlphyton: Relative abundance,
species composition
Figure 7-1. Tiered Sampling Structure.
7-2
-------
Tiered Sampling
and from a questionnaire to identify known
problems in a lake (Table 7-1).
The questionnaire identifies existing known
problems in lakes, but does not address new
problems. An example questionnaire (Figure 7-2)
is modeled after one for stream bioassessment
(USEPA 1989b). Potential recipients of the
questionnaire include regional biologists from
natural resource agencies, the Cooperative
Extension Service (CES), and academic biolo-
gists. Land use, NPDES, and population density
data will identify lakes likely to have problems
requiring further attention (primarily from
eutrophication). but will not estimate biological
impairment in the lakes. Components of desktop
screening include the following:
Land. Use—Land use information indicates the
relative level of anthropogenic stresses in a lake
watershed, especially nonpoint sources of
pollutants. Many states estimate land use from
satellite images.
Discharges—USEPA maintains a data base of
NPDES discharges and their receiving waters.
Algae—Questions on the history of nuisance
algal blooms and perceived problems with high
Table 7-1. Desktop screening assessment.
Watershed Components
Component
Data Collection
Responds to or Indicator of
1. Watershed land use, NPDES.
Maps, existing database,
questionnaire.
GIS databases, e.g., EPA
Reach File; EPA BASINS;
Census Bureau TIGER;
USGS Land Use, Land
Cover.
Identification of potential point and
nonpoint source eutrophication,
toxicity problems.
Biological Components
1. Algal production
- Bloom history
Questionnaire
Identification of perceived problems
(eutrophication).
2. Plant assemblage
- Macrophyte cover
- Extent (% available habitat)
- Density (% cover)
- Known weed problems
Questionnaire
Identification of perceived problems
(weeds, exotic plants, loss of native
plants).
3. Fish assemblage
- Fishery problems
Questionnaire
Identification of perceived problems
(species imbalance, exotic species,
overfishing, overstocking, diseased).
turbidity due to algae are included in the
questionnaire (Figure 7-2).
Macrophyte Survey—Local professionals knowl-
edgeable of the macrophytes in the lake(s) are
canvassed for existing data and information
(Figure 7-2). The questionnaire can provide the
following information:
• Extent of coverage.
• Dominant species.
• Past and present characteristics of the
macrophyte assemblage.
• Factors believed to be limiting or expanding
the spread of macrophytes.
• Past or present management practices used
for control of macrophytes.
Fish Assemblage—Local professionals knowl-
edgeable about fish assemblages can provide the
following information:
• Expected condition of the fish assemblage.
• Likelihood of improvement and degradation.
7-3
-------
Chapter 7
LAKE BIOASSESSMENT QUESTIONNAIRE
This questionnaire is part of an effort to assess the biological health or integrity of lakes of this region.
Our principal focus Is on the blotic health of the designated waterbody as indicated by its biological
assemblages and watershed use. You were selected to participate in the study because of your expertise
In one or more of these areas and your knowledge of the waterbody identified in this questionnaire.
Please complete all statements. If you feel that you cannot complete the questionnaire but are aware of
someone who is familiar with the waterbody, please give this person's name, address, and telephone
number in the space provided below.
Waterbody name:
Waterbody location (see map):
State Countv Lona/Lat Ecoreaion
Lake size
acres at (circle one): <10 acres, 1000-10,000 acres,
10-100 acres, >10,000 acres
100-1000 acres,
SECTION A - OVERALL ASSESSMENT
(Instructions: Answer questions 1-4 using the following scale. Answer by circling only one score
for each question).
Score
5
Description
Species composition, age classes, and trophic structure comparable to non (or
minimally) Impacted sites of similar waterbody size In that ecoregion.
4
Species richness somewhat reduced by loss of some intolerant species; young of
the year of top carnivores rare; less than optimal abundances, age distributions,
and trophic structure for waterbody size and ecoregion.
3
Intolerant species absent, considerably fewer species and individuals than
expected for that waterbody size and ecoregion, older age classes of top
carnivores rare, trophic structure skewed toward omnivory.
2
Dominated by highly tolerant species, omnivores, and habitat generalists; top
carnivores rare or absent; older age classes of all but tolerant species rare;
diseased fish and anomalies relatively common for that waterbody size and
ecoregion.
1
Few individuals and species present, mostly tolerant species and small
Individuals, diseased fish and anomalies abundant compared to other similar-
sized waterbodies In ecoregion.
0
No fish.
Figure 7-2. Example of desktop screening questionnaire.
7-4
-------
Tiered Sampling
1.
Circle the score that best describes your impression of the current condition of the waterbody.
5 4 3 2 1
0
2.
Classify the condition of the lake 10 years ago.
5 4 3 2 1
0
3.
Given present trends, what score will be representative of lake conditions 10 years from now?
5 4 3 2 1
0
4.
If the major human-caused limiting factors were eliminated, how would the lake be rated 10 years from
now?
5 4 3 2 1
0
Subsection A.1 - WATER QUALITY
(Instructions: Complete subsections A.1 - A.4 by circling the single most appropriate limiting factor and
probable cause. If there is more than one limiting factor and cause, please rank them accordingly (by
assigning a "1" for the primary factor and cause, "2" for the secondary factor and cause, etc.).
Limiting Factor Bank
Probable Cause Rank
Temperature too high
Quality of tributaries
Temperature too low
In-lake processes
Turbidity
Salinity
Dissolved oxygen
Gas supersaturation
pH too acidic
pH too basic
Nutrient deficiencv
Nutrient surplus
Toxic substances
Excessive water level
fluctuation
Point source discharae
Industrial
Municipal
Combined sewer
Minina
Upstream dam release
Nonpoint source discharae
individual sewaoe
Urban runoff
Landfill ieachate
Construction
Other (specifv below)
Agriculture
Feedlot
Not ilmitlno
Grazing
Silviculture
Minina
Dam surface release
Shorezone disturbance
Natural
Unknown
Other (specify below!
Figure 7-2. Example of desktop screening questionnaire (continued).
7-5
I
-------
Chapter 7
Subsection A.2-HABITAT STRUCTURE
Limiting Factor Bank
Excessive siltation
Insufficient structure
Insufficient shallows
Insufficient macrophytes
Excessive macrophytes
Insufficient concealment
Insufficient reproductive
habitat
Other (specify below)
Not limiting
Probable Cause Rank
Agriculture
Silviculture
Mining
Grazing
Dam
Diversion
Channelization
Snagging
Natural
Aquatic weed
management
Unknown
Other (specify below)
Subsection A,3 - FISH COMMUNITY
Umitinq.Fac'or Baals
Overharvest
Undertiarvest
Fish stocking
Non-native species
Migration barrier
Tainting
Food limited
Habitat
Fish kills
Other (specify below)
Not limiting
Probable Cause Rank
Aquarists
Point source
Nonpoint source
Natural
Unknown
Management
State agency
Federal agency
Weed Control
Other (specify below)
Subsection A.4 - MAJOR LIMITING FACTOR
UmBlngEaflar Penh
Water quality
Water quantity
Habitat structure
Fish community
Other (specify below)
Figure 7-2. Example of desktop screening questionnaire (continued).
7-6
-------
Tiered Sampling
Subsection A.S-ALGAE
(Instructions: Please provide short answers to questions 1-7, as appropriate).
1. Is there a presence and history of nuisance algae blooms?
2. Have algae blooms resulted in fish kills or other adverse changes to the fish community?
3. Has algae caused odor problems or taste problems in drinking water?
4. Have algae blooms deterred swimmers or affected other forms of contact recreation?
5. Are there other problems caused by algae blooms; and If so, what are they?
6. What is the source of your information?
7. Are there other sources of information that the agency should be aware of such as fishery records and grey literature
studies?
SECTION B - AQUATIC MACROPHYTE COMMUNITY
(Instructions: Answer questions 1 - 3 using the following scale. Circle only one score for each question).
Score DasgipMoo
3 Extent and cover are comparable to non (or minimally) impacted sites of similar waterbody size
In that ecoregion.
2 Macrophyte beds appear weedy. The extent and/or cover are greater than non (or minimally)
impacted sites. The dominant species are those found in highly eutrophic waters.
1 Few macrophytes found compared to non (or minimally) impacted sites. Macrophytes that are
found are usually exotics and are tolerant of a wide range of water quality conditions and/or
fluctuations.
0 No macrophytes.
1. Circle the score that best describes your impression of the current macrophyte conditions of the lake.
3 2 10
2. Classify the macrophyte conditions of the lake 10 years ago.
3 2 10
3. Given the present trends, what score will be representative of lake conditions 10 years from now?
3 2 10
Figure 7-2. Example of desktop screening questionnaire (continued).
7-7
-------
Chapter 7
teflon B.1 - FACTORS EFFECTING MACROPHYTES
(Instructions: Complete subsection by circling the single most appropriate limiting factor and probable cause. If
thara Is more than one limiting factor and cause, please rank them accordingly (by assigning a "1" for the primary
factor and cause, "2" for the secondaiy factor and cause, etc.).
Limiting Factor Rank Probable Cause RanK
Grass carp introduction
Exotic species
Excessive siltatlon
Drawdowns
Weed control
Shoreline cleanup
Excessive epiphytes
Excessive turbidity
Insufficient shallows
Elevation or latitude
Macrophyte beds are expanding
Other (specify below)
Aquarists
Point source
Nonpoint source
Natural
Unknown
Management
State agency
Federal agency
Fisherman
Other (specify below)
Not limiting
SUb?WtlPfl P,2 - MACROPHYTE EXTENT AND SPECIES
(Instructions: Please provide short answers to questions 1 - 4, as appropriate).
1. What is the extent of macrophyte coverage in the photic zone?
2. What are the dominant species?
3. What is the source of your information on macrophytes?
4. Are there other sources of Information on the macrophyte community in this waterbody that the agency should be
aware of such as management reports or grey literature studies?
SECTION C - WATERSHED CHARACTERISTICS AND LAND USE
(Instructions: Please provide short answers to questions 1 -11, as appropriate).
1. Watershed size acres S. Urban %
2. Elevation difference ft 4. Agricultural %
[watershed divide to lake surface]
3. Forest or natural vegetation 6. Suburban/residential %
7. Human population density In lake watershed
8. Number of dischargers within the watershed (e.g., NPDES permits)
9. What is the source of your information on the watershed?
10. Are there other sources of information on the watershed and surrounding land use that the agency should be
aware of such as grey literature or land use planning documents?
Figure 7-2. Example of desktop screening questionnaire (continued).
7-8
-------
Tiered Sampling
• Major limiting factors.
• Water qualify
• Habitat availability
• Management, harvest, or mortality
Desktop Integration
Based on responses to the questionnaire,
perceived levels of impairment can be judged
from the three biological assemblages: algae,
macrophytes, and fish. The three evaluations
are kept separate. Perception of a problem, or a
substantial departure from expected conditions,
earns a rating of "impaired" for the respective
assemblage. The land use information is used to
identify potential stressors on a lake.
7.2 TIER 1: TROPHIC STATE
AND MACROPHYTES
Tier 1 requires sampling of primary producers to
assess trophic state and aquatic macrophytes. It
can be done with a single visit during an index
period when the objective is a synoptic survey
and screening of many lakes filer 1A). Tier 1A is
only appropriate for regional assessments—it
Table 7-2. Tier 1: Trophic state and macrophyte sampling.
Component
Data Collection
Responds to or
Indicator of
Tier 1A
Tier 11
Habitat Components
1. Watershed land use,
population, NPDES.
Maps, existing database,
questionnaire.
GIS databases, e.g., EPA
Reach File; EPA BASINS;
Census Bureau TIGER;
USGS Land Use, Land
Cover.
Desktop screening habitat.
Potential causes.
2. In-lake physical habitat
maximum depth area
inflow.
Maps or survey (single
visit).
Potential causes.
3. Water Quality
- DO, temperature profile
- pH, alkalinity, conductivity
- Secchi depth
- Total dissolved solids
- Nutrient concentration
- Algal growth potential
Single index period.
Surface or integrated.
Multiple visits.
Water column
sample.
DO problems,
eutrophication,
stratification,
acidification,
turbidity.
Biological Components
4. Algal chlorophyll a
concentration.
Single visit chlorophyll
sample from 0.5m.
Surface integrated water
sample.
Multiple visits.
Eutrophication
Sa. Submerged macrophytes
- % of available habitat with
macrophytes
- dominant species
Single visit, aerial photos if
possible; otherwise,
estimate from shorezone
survey.
Identify dominant species.
Multiple visits.
Eutrophication,
herbicides, exotics.
7-9
-------
Chapter 7
Design of a sampling
program Inevitably
requires compromises to
answer the Intended
questions in a reasonable
time and at a reasonable
cost.
cannot be used to assess single lakes. More
precise estimates for single lakes can be made
with Tier IB, comprising
several sampling visits to
determine growing season
averages. Her 1 consists
of the Desktop Screening
land use survey, lake
physical habitat, water
chemistry (dissolved
oxygen, nutrient concen-
trations, conductivity,
alkalinity, pH), Secchi
depth, chlorophyll a
concentration, and a
submerged macrophyte
survey (Table 7-2). The survey enables:
• Identification of trophic state based on
chlorophyll a concentration, nutrient
concentration, and Secchi depth.
• Detection of weed problems or loss of
aquatic macrophytes.
• Detection of midsummer oxygen stress.
7.2.1 Sampling Frequency for
TS1 Variables (Tier 1A
v®. Tier 1B)
Tier 1A consists of sampling during an index
period, typically mid to late summer for trophic
state variables (e.g., chlorophyll a, Secchi depth,
nutrients}. Tier 1A is adequate for characteriza-
tion of lakes In a region, when many lakes must
be sampled to develop the characterization and
assessment. Tier 1A will yield a good character-
ization of a region or a population of lakes, but
precise characterization of individual lakes, for
site-specific management, will require Tier IB,
with more frequent sampling. Tier IB takes into
account the changes in chlorophyll and nutri-
ents that can occur in a short time and is used
to estimate seasonal averages of the variables by
sampling several times during the growing
season. Trophic State Indices (TSI) are calculated
from the seasonal average estimates of chloro-
phyll, Secchi depth, and nutrients. The number
of sampling visits required depends on the
temporal variation in the lake and the desired
precision of the estimated seasonal average.
Monthly sampling appears to be adequate for
most purposes (Knowlton and Jones 1989).
7.2.2 Sample Locations
Trophic State
Measu rements
for
Design of a sampling program inevitably re-
quires compromises to answer the intended
questions in a reasonable time and at a reason-
able cost. In lake biosurveys, the unit of interest
(sampling unit) may be the whole lake, a lake
basin, a tributary arm, or an embayment. In
some situations, the unit of interest may be an
area of the lake receiving discharges or runoff.
The object of sampling is to characterize the
sampling units with sufficient precision and
accuracy to meet the needs of the program.
Sample sites are selected to be representative of
the lake. Single sites have traditionally been
located in the middle of the lake, usually over
the deepest area. For unbiased characterization,
multiple sites should be selected randomly.
Sampling may be stratified by zones, e.g.,
littoral, pelagic, and inflows; or riverine, transi-
tional, and lacustrine (Figure 7-3). Estimation of
mean values for the whole lake should be
weighted by the relative area or volume of each
zone. Figure 7-4 shows an example of sampling
locations for all tiers in a relatively simple lake
(natural or impoundment).
Lakes may be characterized by single or mul-
tiple sample sites in each lake, depending on the
objectives of the survey.
Single sample site
If the objective is to characterize a large popula-
tion of lakes, as in a statewide survey, then a
single sample per lake is most cost-effective. A
single site is typically chosen as the midpoint of
the central basin of the lake, and is usually
sufficient to prioritize lakes within a region.
H
Large riverine reservoirs have
known gradients of nutrients and
productivity from the river inflow to
the dam (Kennedy and Walker
1990), and a single site is not
appropriate. Large reservoirs would
require a minimum of three sites,
corresponding to the riverine,
transitional, and lacustrine zones,
respectively (Figure 7-3}.
7-10
-------
Tiered Sampling
inflow east arm
inflow west aim
S transition g
3 east arm 5
transition
west arm
forebay Sfj&y
Figure 7-3. Sampling zones in large or complex
lakes (large reservoirs, multi-basin lakes).
Multiple sites
If turbidity, nutrients, and algae are known to
be variable across the surface of a lake, then
multiple sample sites are required (Figure 7-4).
Hlf gradients are known to occur, as in
many large reservoirs, then sampling
should be stratified by zones. For
example, in a reservoir one could
define the three reservoir zones
(riverine, transitional, lacustrine) as
sampling strata, and take two or
more samples from each zone.
The exact number of sampling sites in a lake or
lake zone is determined by the spatial variability
of nutrients, turbidity, and chlorophyll; and the
desired precision. In general, within a basin or
reservoir zone, variation in time is larger than
variation in space (Knowlton and Jones 1989).
Thus, chlorophyll sampled 2 weeks apart may
differ by several fold, but samples on the same
day 500m apart are likely to differ much less.
If precise characterization of individual lakes is
an objective of the biological survey then it is
more cost-effective to sample repeatedly during
the growing season (Tier IB) than to sample
multiple sites at a single time (Tier 1A).
Composite samples
Composite samples are taken from several sites
in a lake or lake zone, and combined into a
single sample for laboratory analysis. For example,
water samples may be taken from four sites in a
lake, and poured into a single clean bucket. The
composite sample is subsampled for chlorophyll
a and nutrients. Secehi depth temperature, and
DO are measured at each of the four sites. Care
must be taken that the methods and volume
sampled are the same at each site. Composite
samples characterize the lake better than a
single sample and they save laboratory analysis
costs. The principal disadvantage of composite
samples is that they do not allow estimation of
spatial variability within a lake.
7.2.3 Trophic State
The Tier 1 Trophic State Indices (TSI) are
estimated from Seechi depth, chlorophyll a, and
nutrient concentrations. Field methods for
Secchi transparency and chlorophyll a are
outlined below and summarized in Table 7-4.
Seechi Depth (SD)
Secchi depth is a measure of transparency.
Turbidity caused by suspended sediments and
algae decreases Secchi depth.
| W«Mr<»3uwssa?npk» 1 andTt*f2B)
BsntWc macfotRveftsbratd and fish
aampUng*
Cn»r2A)
Macrophyte transect (Tw 2A}
Store fractal transact (Tkic 2)
X SocSmwrtsd cfiatom ctompto
Figure 7-4. Integrated sampling, Tiers 1 and 2.
7-fl
-------
Chapter 7
Sampling Location—Secchi disk transparency
can be measured at one or more representative
locations.
Frequency—1Tiers 1A and 2A: single determina-
tion, midsummer. Tiers IB and 2B: 6 to 10
samples during the growing season (e.g., March
through October).
Sampling Procedure—Readings are obtained with
a 20cm plastic or metal Secchi disk that is
divided into black and white quadrants on a
nonstretchable line, calibrated in centimeters.
The disk is lowered into the water until it
disappears from view, then is raised slowly to
the point where it reappears. Secchi depth is the
average of the two depths.
Observations are made from the sunny side of
the boat or dock, during midday, without
sunglasses, and as close as possible to the
water in order to reduce glare.
Data Analysis—Secchi depth can be used in deter-
mining trophic state along with chlorophyll a.
Chlorophyll
Chlorophyll a sampling and analysis follow
standard protocols (USEPA 1994a, USEPA
1994b).
Presampling—Samples must be collected in a
clean container, without using acid washes or
phosphorus detergents. Before sample collec-
tion, bottles and collectors should always be
double or triple rinsed with the lake water to be
sampled.
Sample Location—One location or several
representative locations for composite sample.
Frequency—The same as Secchi depth.
Depth—Chlorophyll a concentration may be
estimated from surface samples taken at 0.5m,
from integrated epilimnion samples, or from
integrated water column samples. Half-meter
surface samples require the least equipment
and can be taken by hand; epilimnion and
integrated water column samples are taken with
a flexible hose.
Sampling Procedure—Surface sample, 0.5m. A
rinsed, 1-liter sample bottle is inverted and held
at depth (arm's length) by hand, turned up to
fill, and brought to the surface.
Hose sample—A flexible hose is an easy method
to obtain an integrated sample over the whole
water column or over a defined portion, such as
the epilimnion. The weighted end of a plastic
hose is lowered to a given depth. The upper end
is stoppered or clamped at the surface, and the
weighted end is hauled to the surface with an
attached line. The hose is emptied into a clean
sample bucket, and chlorophyll and chemical
subsamples can be drawn from the integrated
sample. The hose may be lowered to lm above
the bottom for a water column sample, to the
metalimnion, to twice Secchi depth as an
estimate of the photic zone, or to a fixed depth
(e.g., 5m). Each standard depth method has its
own advantages and disadvantages (Carlson
and Simpson 1996). Consistency of sampling
method is more important than selecting the
"best" standard depth.
Water samples are filtered for chlorophyll a
extraction. A "rule of thumb" for the quantity to
filter is 100ml for every foot of Secchi depth
(330ml for every meter; D. Canfield, personal
communication). Samples are vacuum-filtered
on glass-fiber paper, and the filter papers are
stored frozen in the dark. Detailed instructions
Table 7-3. Sampling summary for chlorophyll, water quality, and phytoplankton.
Habitat
Open water, 1 to 5 sites per lake or lake stratum.
Sampling Gear
Hand-held bottle or flexible hose.
Index Period
Single mid-season sample (Tier 1 A) or monthly samples during growing season
(Tier 1B, Tier 2B).
Sampling
Bottle: invert bottle at arm's length depth (0.5 m); turn.
Uphose: lower open weighted hose through water column to predetermined depth,
stopper, and haul up.
Analysis
Chlorophyll and water quality: standard methods.
Phytoplankton: filter or settle and identify 300 to 500 cells to genus.
7-12
-------
Tiered Sampling
for filtering and analysis are in APHA (1992)
and USEPA (1994a, 1994b).
Water Chemistry
Samples of water for chemical analysis are
collected in the same manner as chlorophyll
samples. Sampler bottles should be cleaned in a
phosphate-free detergent prior to use and rinsed
two to three times in lake water in the field.
Samples may need to be preserved or filtered in
the field depending upon which chemicals are to
be analyzed.
Dissolved Oxygen and Temperature Profiles—A
dissolved oxygen/temperature electrode (EPA
Method 360.1) is used to measure both dis-
solved oxygen and temperature. Using the
electrode, dissolved oxygen and temperature
may be measured at 0.5m intervals to produce
dissolved oxygen and temperature profiles.
Dissolved oxygen electrodes should be cali-
brated against standard chemical titration
methods before and after field use.
pH, Alkalinity and Acid Neutralizing Capacity—A
calibrated pH meter may be used to determine
pH. Acid neutralizing capacity is important to
the ability of a waterbody to resist changes in
pH due to addition of acid and is based upon
the alkalinity of the water and dissociated
organic compounds present. Carbonates,
bicarbonates and hydroxides are the major
contributors to alkalinity which is determined
using calorimetric titration methods (APHA
1992). For more precise determination of acid
neutralizing capacity, the Gran plot method is
used (USEPA 1987a).
Total Dissolved Solids (TDS)—Total dissolved
solids consist of inorganic solutes such as
nitrates, sulfates, and carbonates, and organic
substances dissolved in water (APHA 1992). TDS
is measured by first filtering a measured volume
of sample water through a filter, and weighing
the dried residue. See APHA (1992) for specific
methods.
Algal Growth Potential Test (AGPT)—Because
nutrients are not always present in a form
available to algae, direct chemical measure-
ments may not be predictive of the actual
potential for algal growth. The Algal Growth
Potential Test (also know as a biostimulation
study, APHA 1992) was developed to directly
measure in a standardized way the potential of
waters to support algal growth.
Total Nitrogen—Nitrogen is an important plant
nutrient and may serve as a limiting factor in
some waters, especially subtropical lakes. Total
nitrogen is a combination
of nitrate/nitrite nitrogen
and total Kjeldahl nitro-
gen (organic and reduced
nitrogen). Total Kjeldahl
nitrogen is measured
Trophic state determinations
provide a method for
determining whether
increased nutrients or
sediments (loading) are
causing changes in a lake.
using a digestion tech-
nique that converts
organic nitrogen to
ammonia and includes
any other ammonia
present in the sample.
Nitrate plus nitrite is measured with standard
colorimeter methods (APHA 1992).
Total Phosphorus—Phosphorus is a limiting
nutrient in many fresh waters. Total phosphorus
can be analyzed using the automated procedure
outlined in USEPA Method 365.1.
Estimation of Trophic State
Trophic state determinations provide a method
for determining whether increased nutrients or
sediments (loading) are causing changes in a
lake. Carlson's TSI uses Secchi depth, chloro-
phyll a, and total phosphorus, each producing
an independent measure of trophic state
(Carlson 1977). Index values range from ap-
proximately 0 (ultraoligotrophic) to 100
(hypereutrophic). The index is scaled so that TSI
= 0 represents a Secchi transparency of 64 m.
Each halving of transparency represents an
increase of 10 TSI units. For example. TSI of 50
represents a transparency of 2m, the approxi-
mate division between oligotrophic and
eutrophic lakes (USEPA 1990b). A TSI is
calculated from each of Secchi depth (SD),
chlorophyll concentration (Chi), and total
phosphorus concentration (TP) (Carlson 1977,
Carlson and Simpson 1996).
TSI(Chl) = 30.6 + 9.81 In(Chl)
TSI(TP) = 4.15 + 14.421 In (TP)
TSI(SD)= 60-14.41 In(SD)
Trophic state indices are used to infer trophic
state of a lake and whether algal growth is
nutrient limited or light limited. If the three
indices are approximately equal, then phospho-
rus limits algal growth. If the three are not
equal, then other interpretations exist (Carlson
7-13
-------
Chapter 7
and Simpson 1996). A trophic state index has
also been developed for total nitrogen (TN)
(Kratzer and Brezonik 1981, Carlson 1992):
TSI(TN) = 54.45 + 14.43 In(TN)
For a more complete discussion of trophic state
indices and their interpretation, see Carlson
(1992) and Carlson and Simpson (1996).
7.2.4- Aquatic Macrophytes
The Tier 1 macrophyte survey is a visual esti-
mate of percent cover of submerged and floating
macrophytes in shallow water, and identification
of the most dominant species and weedy or
exotic species. The survey can be done with
aerial photographs (if available); a visual whole-
lake survey in small lakes, or examination of
transects in large lakes. Three to ten transects
should be sufficient for most lakes or
embayments too large to survey in their entirety.
Large lakes with known differences within the
lake should be sampled by lake zone; for ex-
ample, the shallow riverine zones of a reservoir
may have greater macrophyte cover than the
lacustrine zone.
To avoid bias, transects should be selected
randomly within each lake zone. A method of
selecting transects is to divide the shore into
equal segments (corresponding to the number of
transects). A point is selected randomly in each
segment as the starting point for transects.
Transects are perpendicular to shore to deeper
water.
Total vegetative cover is estimated visually. The
presence of algae mats and epiphytes should be
noted. Cover might be difficult to estimate in
turbid waters. Vegetation samples may be
collected with a rake and total abundance
estimated from the material raked in (ordinal
scale: sparse, moderate, abundant). The most
dominant species, and any weedy or exotic
species, are identified.
7.3 TIER 2AS BIOLOGICAL
ASSEMBLAGE
ASSESSMENT
Tier 2A sampling requires two or more lake
biotic assemblages: macrophytes, sedimented
diatoms, fish, or macrobenthos (Table 7-4). Tier
1 variables, including DO, chlorophyll a, and
Secchi depth, are also critic components of the
Tier 2A survey. Tier 2A may be built on either
Tier 1A or IB. Macrophytes are the easiest of
these assemblages to identify and count in the
field (using wet weight instead of relative abun-
dance). Sedimented diatoms are also relatively
easy to sample, although identification and
enumeration must be done in the laboratory.
The choice of which plant assemblage to sample
clearly depends on the importance of the
assemblage in lakes of the region—diatoms
would be the choice in regions where macrophytes
are minor components of the lake system.
The habitat components of the Tier 2A survey
build on the desktop screening and Tier 1 habitat
assessment and also include a semi-quantitative
shore zone habitat evaluation (Table 5-3). Tier
2A requires estimates of shorezone land use,
riparian vegetation, emergent macrophyte extent
and cover, and floating macrophyte extent and
cover at several transects from the shore.
The Tier 2A faunal component consists of the
benthic macroinvertebrates. Macroinvertebrates
are sampled from the sublittoral zone, below the
floating macrophyte zone, yet above the ther-
mocline to avoid sampling predominantly anoxic
areas. Tier 2A sampling typically consists of a
single visit during an index period. Benthic
macroinvertebrates may optionally be sampled
more frequently to obtain growing season
averages. Macrophytes are best sampled mid- to
late in the growing season when plant biomass
is near its annual maximum. Sedimented
diatoms, which represent sedimentation of at
least a year or more, may be sampled at any time.
Tier 2A allows more precise detection and
identification of problems and potential causes
than Her 1, as well as detection of biological
effects on the biotic assemblages selected for
assessment.
7.3.1 Tier 2A: Transect
Sampling
Establish Transects
Tier 2A sampling of macrophytes and benthic
macroinvertebrates, and the shorezone habitat
are surveyed along 3 to 10 transects perpen-
dicular to the shoreline (Figure 7-4). Transects
7-14
-------
Tiered Sampling
Table 7-4. Tier 2A: Routine biological sampling.
Component
Data Collection
Responds to or
Indicator of
Habitat Components
1. Watershed land use,
population, NPDES.
Desktop screening habitat.
. v;' •'
2. Lake physical.
Tier 1 habitat.
¦ ¦ .V- -
-
•
3. Shorezone habitat
assessment.
3-10 transects:
- land use
- bank stability
- riparian vegetation
- emergent vegetation
'• •:
r r •_ •.
' • * ,
4. Water quality DO seasonal
or annual mean, %
depth-time Mean pH,
alkalinity Secchi depth.
Tier 1 water quality (1A or 1B).
Trophic state,
turbidity.
Biological Components
5. Algal chlorophyll a.
Tier 1 chlorophyll (1A or 1B).
Trophic state.
6,7. Assemblages (minimum
2):
a. Macrophyte species.
2-3 samples from transects; identify plants to
species and weigh cumulative sample of each
species, or count stems.
Trophic state,
exotics,
herbicides.
b. Macrobenthos
Sublittorai surface sediment grab at end of
each transect; identity to lowest practical level,
100-200 organisms.
DO, siltation,
toxicity,
productivity.
c. Fish assemblages.
Littoral electrofishing sample at the end of each
transect; sublittorai netting; identify to species,
enumerate, weigh, and record incidence of
external anomalies.
DO, toxicity,
productivity.
d. Sediment diatoms.
Surface sediment grab In deepest part of lake;
identify to species and variety.
Nutrient
enrichment,
toxicity.
are the same as the Tier 1 macrophyte
transects: the lake (or lake zone) shoreline is
divided into equal length segments correspond-
ing to the number of transects, and a transect
start point is randomly selected in each segment.
7.3.2 Shorezone
Measu r ements
Each transect is extended visually on the lake
shorezone, and the condition of the shorezone is
determined. Shorezone measurements include
riparian vegetation cover estimates, lake bank
substrate and erosion, and human modifica-
tions. Figure 7-5 is an example scoring sheet for
habitat measurements showing how the vari-
ables are scored.
7.3.3 Aquatic Macrophytes
Tier 2A macrophyte sampling is more systematic
and detailed than Tier 1. The objective is to
obtain relative abundances of macrophyte taxa
to develop assemblage measurements. Relative
7-15
-------
Chapter 7
abundance can be estimated by stem counts
(number of stems of each species) or biomass.
Biomass Is preferred because a stem does not
correspond to an individual plant, and biomass
is a good indicator of species dominance in the
habitat. An alternative to relative abundance is
scoring presence and absence of species in
quadrat.
One to four macrophyte sampling locations are
established on each transect within depth zones
between shorezone emergent and the
unvegetated. sublittoral bottom. For example,
location may be identified in Q-lm depth, l-2m,
2-3m, and 3-4m depth (Weber et al. 1995).
Stem counts—May be done with the transect
method, by counting stems touching a line held
on the transect. Stems may also be counted in
quadrants, where all stems within a 1/4 m2
quadrat are counted and identified. Stem counts
may require diving in water deeper than 1 m.
One or more sampling stations (for quadrat
sampling) are selected on each transect between
the emergent macrophyte zone and the deepest
extent of submerged macrophytes.
Biomass Sampling—The easiest method to
estimate macrophyte biomass is with an aquatic
weed rake (Table 7-5). At each station on the
transect, an aquatic weed rake or thatching
rake is dragged a set distance (e.g., 1m) to
sample vegetation (Trebitz et al. 1993). Plants
from all stations on the lake are identified and
sorted by species, and the total of each species
collected is weighed (wet weight) to obtain
estimates of biomass and proportion of biomass
of each species. Algae mats
Tier 2A macrophyte and epiphytic growth on
leaves and stems are
sampling Is more systsm- described. Voucher speci-
mens of each species
should be kept for com-
ber 1. The objective is Plete identification and for
permanent record. Depth is
to obtain relative abun- sounded at the lakeward
edge of submerged vegeta-
dances of macrophyte uon
atlc and detailed than
taxa to develop assem-
blage measurements.
Aquatic weed rakes are
biased against macro-
phytes that can slip
through the tines of the
rake. Therefore, a more accurate estimate of
biomass would be to clip all plants in the
quadrat for wet weight determination. Clip
plots would require diving or snorkeling in
water more than 1m deep, Biomass can be
estimated more accurately by drying the sorted
plant material for dry weight determination, at
the cost of additional processing.
The weed rake and wet weight determination is
likely to be the most cost-effective methodology
for most purposes. Although it undersamples
certain species, it is likely to be consistent
enough to use for biological surveys, as long as
the same sampling methodology is used in all
lakes.
Presence-Absence—Instead of estimating
biomass, species can be scored for presence or
absence within quadrants (Weber et al. 1995).
Each sampling location along a transect is
divided into four quadrants. Each quadrat is
sampled with the rake, and each species re-
ceives one point for every quadrat in which it
occurs.
7.3.4. Macroinvertebrates
The macroinvertebrate assemblage beyond the
macrophyte zone is sampled with gear appropri-
ate to the bottom type and depth (e.g., Ponar,
Ekman grab sampler, dome sampler); and the
assemblage is identified and characterized
(Table 7-7).
Sampling Period—Two sampling periods have
been identified either the most stressful period
(usually late summer) or a period after recruit-
ment (usually early spring) but before major
emergence of adult insects.
Sampling Location—Along transects, the sublit-
toral habitat is recommended as the most
appropriate habitat for sampling due to its
relatively stable nature.
Sampling Gear—The type of gear will depend on
the substrate being sampled (Table 7-7). A
standard mesh size of 595 (Am (No. 30 mesh) is
required.
Sample Replication—To characterize the
macroinvertebrate assemblage, multiple grabs
are taken from several sites. Each transect ends
in a macroinvertebrate sample site, and two to
three grabs are taken at each site. Grabs may
be composited into a single composite sample.
7-16
-------
Tiered Sampffng
Table 7-5. Sampling summary for submerged macrophytes.
Habitat
Littoral zone.
Sampling Gear.
Tierl: none.
Tier 2A double-headed rake on chain (Trebitz et al. 1993), or 1m2 quadrats and
diving gear.
Index Period.
Late summer. (Macrophytes are sampled once regardless of tiers.)
Sampling
Tier 1: Estimate of area covered by macrophytes.
Tier 2A; 2-3 semiquantitative rake samples to determine relative biomass of
species; on randomly placed transects perpendicular to shore.
Alternative: 1-3 randomly tossed quadrate on each transect, then stem count and
identification of each species in quadrat.
Analysis
Tier 1: Dominant species Identifed, % estimated.
Tier 2A: All species identified, relative abundance of each estimated from wet
weight or stem count (Trebitz et al. 1993).
Sample Processing—To process the sample,
organisms are removed from sticks, rocks, and
similar size objects. The remainder of the
sample is placed in a tub and mixed into a fine,
uniform slurry. After mixing, the slurry is sieved
using a U.S. No. 30 sieve (595 um) to remove
organic and mineral material. The benthic
organisms are retained by the sieve, which can
be emptied into a light-color, gridded sorting
tray. Grid cells are selected at random and
sorted until at least a 100-organism subsample
has been counted and identified to the appropri-
ate taxonomic level. The last grid cell is sorted
completely until all organisms from the grid are
identified to the lowest practical level. Further
description of sorting is presented in EPA/440/
4-89-001 (USEPA 1989b).
7.3.5 Fish Assemblage
Fish assemblages can be sampled by
electrofishing in and/or beyond the macrophyte
zone. Sampling effort for fish should be kept
constant between transects. Electrofishing is
generally the single most cost-effective sampling
Habitat
Preferred: sublittoral.
Alternative: profundal (if hypoxia is rare).
Sampling Gear.
Regionally most appropriate for substrate (Table 7-8); 595mm mesh (No. 30 sieve).
Index Period.
Regionally most appropriate.
Preferred: Late summer (most stressful; most regions).
Alternative: Early spring; winter (subtropical lakes).
Sampling
Lakewide composite samples of 2-3 grabs at each of 3-5 sublittoral sites (7 to 15
grabs total) or keep sites as replicates if an in-lake variance estimate will be used in
assessment.
Analysis
Preferred: lowest practical taxonomic level, 100-organism subsample.
Alternatives: more than 100 organisms.
7-17
Table 7-6. Sampling summary for benthic macroinvertebrates.
-------
Chapter 7
Lakeshore Habitat Measurements and Metrics
Lake Name: Date of Visit: Visit #:
Lake ID: Team ID:
Riparian Vegetation Measurements Transect
ID:
AreaJ Coverage Categories: 0 = Absent, 1 = Sparse (< 10%), 2 = Moderated 0-40%)
3 = Heavy (40 to 75%), 4 = Very Heavy (>75%)
Canopy (%) cover
Understory (%) cover
Ground Cover (%) cover
Barren (%) cover
Bank Measurement
Rocky (%)
Soil (%)
Vegetated (%)
Other (%)
Bank Erosion Score (0-4): 0 = None, 4 = Severe
Human influence Measurements 0 = Absent, 1 = Present within Transect
0.5 = Observed Adjacent to or Behind Transect
Buildings
In-Lake Structure
Roads, Railroads
Agriculture
Lawn
Dump or Landfill
Figure 7-5. Example scoring sheet for shorezone habitat.
7-18
-------
Tiered Sampling
Case Study: Florida sampling methods
; In 1995, FDEP adopted anew sampling protocol to
obtain more representative samples of each lake,
In part based on results from the earlier samples.
Lakes greater than 1000 acres are divided into two
or more basins, usually by separating at constric-
tion p^nts w between bathymetricsdly identif}s±>le
: basins (Fig.7-6).'Th§2-4rri sublittoral zone of each
lake basin is divided into 12 equal segfriehts, and a
grab is taken in eachsegment with-a Petite Ponar
or Ekmansampler (0.02nf) (Fig. 7^6}. Positions of
segments and sampling sites are estimated by eye
in the field. The 12 grabs are combined into a single
composite sample, which is randomly
subsampled to a count of 100 organisms,
identified to the lowest practical taxo-
nomic level. Basins (in lakes greater'than :
1000 acres) are retained as separate
sample unite: Lakes smaller than 1000
f acres are represented by a single 100-
brganism sample. A second grs0 sample
is takenateach oftiie 12 stations for sedi-
ments, which am likewise combined into
a single representative sample.
In fixed organism subsampling, a targeted
number of organisms (typically 100to500) ;
is identified, if fixed organism subsam-
> piing for benthos is conducted in an unbi-
ased manner using a random Selection
mepiod.the resultinginfonrtationonrich-
: ness end relative abundance is compa-
rable among samples. For benthic
samples, the targeted number is readied
by randomly choosing several fractions or "grids" from
a~jj^^af..or§?misms enclosed within the grids; are
sorted'to: avoid bias toward large and easily seen
individuals. Ideally, several (four or more) grids arej
sorted to ensure proper representation^ ;
Surface and bottom water chemistry samples, and
phytopianktpnsamples, are taken hear the center
ofeachlkife. Ci^rva^sii^ikled field me^ur^^
ments and laboratory analyses, and identification
ofphytoplariktontogenus.
For lakes with a surface area
of 1,000 acres or less.
Benthic dredge at 2m to
4m depth.
Surface grab for water
chemistry parameters.
Figure7-6.Florida lakes sampling scheme. (The lake
is divided into 12 approximately equal segments. A
Ponar grab lit taken from each segment, at a random
location In the 2 to 4 meterdepth zone. Water
chemistry, chlorophy ll,andSecchidepth are
measured from the center of the lake.)
Substrate
Gear Types
Submerged aquatic vegetation.
Dip net.
Rocks, gravel.
Diver operated dome sampler.
Sand
Peterson, Van Veen grabs.
Mud
Ponar, Ekman grabs.
Clay
Peterson, Van Veen grabs;
Table 7-7, Benthic macroinvertebrate sampling gear appropriate for major substrate types.
7-19
-------
Chapter 7
Case Study: TVA BenthSc Macrolnvertebratea Collection Methods
Benthic macroinvertebrate assemblage samples
wereHbRocted In the spring (March and April) at 69
locations on 30 TVA reservoirs. Sample locations
were selected In each of the forebav, mid-reservoir,
and inflow areas, corresponding to lacustrine, tran-
sitional, and marine conditions, respectively (Fig-
ure 7-3). At each sample location, a line-of-sight
*: transect was established across the width of the
reservoir, and one Ponar grab sample collected at
10 equally spaced locations along this transect.
When rocky substrates were encountered, a
Petemon^mige was used. Cam was taken to col-
orSyTran the permanently wetted bot-
tom portion of the reservoir (i.e., below the eleva-
tfon of the minimum winter pool level}. Samples were
washed In the field, transferred to a labeled collec-
tion Jar, and fixed with 10 percent buffered formalin
solution. Samples were sorted and identified in the
field, to the lowest practical taxon, typically genus,
and reported as number per square meter.
To assess the reproducibility of benthic macroin-
vertebrate sampling results, replicate samples were
collected at 13ofthe 69 sampling locations in 1994,
' with all types of reservoir locations (i.e., forebay,
transition zone, embayment and inflow) Included.
At each of the replicate sampling locations, the sam-
pling protocol involved collection of a first set of 10
samples, leaving the sampling location, and then
" returning as near as possible to the original transect
site (on the same day) and repeating the collection
of a second (replicate) set of 10 samples. Results
from sets of replicate samples were evaluated for
reproducibility.
method for fish (Scott et al. 1992) but it is not
effective in deep water. If deep water fish are an
endpoint of concern, then gill nets, fish traps, or
trawls can be used. A combination of nets and
electrofishing often provide a more representa-
tive sample of the fish assemblage; however,
multiple methods translate to a substantial cost
for field effort. A variety of nets may be used to
sample littoral and sublittoral areas. Fish
sampling methodologies are further outlined in
EPA 600/R-92/ 111 (USEPA 1992b) and Table 7-8.
Sampling Procedures
Etecti-qflshing—Multiple habitats are selected in
littoral areas for electrofishing. Habitat distinc-
tions are based on substrate (e.g., rocks, sand,
clay) and on available cover (e.g., vegetation,
woody debris).
Nets—A variety of nets are used to sample
littoral and sublittoral areas. It is recommended
that trapping nets (gill nets, trammel nets, fyke
nets, trapnets) be set for 2 to 5 days with
collection once or twice a day.
Table 7-8. Sampling summary for fish assemblage.
Habitat
Littoral and sublittoral zones.
Sampling Gear
Boat electrofisher (for available microhabitats within shallow littoral areas).
Experimental gill nets (extended for littoral to sublittoral zones).
Index Period
Regionally most appropriate.
Preferred: Late summer - early fall.
Alternative: Early spring; winter (subtropical lakes).
Sampling
Littoral electrofishing sample reach of shoreline at the end of each transect.
All microhabitats sampled within each measured littoral reach.
Experimental gill nets (five panel nets) set perpendicular to shore at the end of each
transect, extending from littoral to sublittoral zones.
Analysis
Preferred: All specimens Identified to species, enumerated, measured, weighed, and
examined for incidence of external anomalies.
Alternative: Abundant species (e.g., greater than 50 individuals per sample) may be
subsampled, measured, weighed, and data extrapolated for the species total.
7-20
-------
Tiered Sampling
• Gill nets or trammel nets are set in littoral
areas, perpendicular to shore, and usually
extend into sublittoral areas. To reduce size
selectivity, an experimental gill net consist-
ing of panels of five different mesh sizes is
commonly used. Smaller mesh size (0.5in) is
used in shallow areas and up to 2-2.5in
mesh farther out.
• Fyke nets, trap nets, and fish traps can be
used in shallow areas.
• Trawl and sonar can be used to sample
pelagic areas.
Sample Processing—Fish samples are processed
as recommended to the RBP manual EPA 440/4-
89-001 (USEPA 1989b). Sampling duration and
area or distance sampled are recorded in order to
determine level of effort. Specimens are identified
to species, then total numbers and weights, and
the incidence of external anomalies is recorded
for each. Voucher specimens of each species from
each site are preserved in a 10 percent formalin
solution, in a labeled jar. The voucher collections
are placed with the state ichthyologlcal museum
to confirm identifications and to constitute a
biological record. This is especially important for
uncommon species, for species requiring verifica-
tion of identification, and for documentation of
new distribution records. If kept in a live well,
most fish can be identified and counted in the
field by trained personnel and returned to the
lake alive. Additional information on field meth-
ods is presented in Karr et al. (1986) and EPA
600/R-92/111 (1992b).
7.3.6 Sedlmented Diatoms
Diatom frustules are preserved in lake sedi-
ments that are not disturbed or resuspended.
Field sampling for sediment diatoms can be
relatively fast. Field methods outlined below and
Case Study TVA Fish Colfection Methods
Shoreline ebctrofishing samples were collected
s during daylight hotim from inflow, transition, and
¦ fiveforebayztmeisofmiost reservoirs frorriSep-
tembertomid-November (Figure 7-fy:Onlypne
^or^Mnes^ms^rn^^dn^mfymrs Where
zones
werei$ampie& in Mbutaiyrmeiyohs.
A toted of-:15 ele^rofishing transects, each cov-
ering00m of shoreline, was collected from each!
of the sampled zones:Allhabitats were sampled i
- in proportion to their occurrencejnthezone.
Where: conditions permitted experimental gill ¦
reservoirzone. v
Ex&ssive puirent prevented use of gill n^tsjn
mainstream Mow fhforebay and transi- ]
¦ tion zones, nets were setin all habitat ty0s,al-
' temating mesh sizes toward the shoreline be-
tween sets.
Total length (mm) and weight (g) were obtained
for all sport Species and channel catfish.. Reii-
prior to release. During electrofishing, fish ob-
servedbut not, captured were included if posh
- tive identification could be made and counts
were:estim0e4 when high densities of idpntifi
able fish were entxtuMemd; Ypung-pf-^ .
calculations jKarr 1981), were excluded from ,
proportional and aburidance metrics (due to
fishaxammed ctoseiy to obtain length and
weight measurements were inspected exter--.-
nallyfor signs of disease, parasites, and
l'aricmalies. " -
Table 7-9. Sampling summary for sedimented diatoms.
Habitat
Mid-lake, deep depositional area.
Sampling Gear
Grab (surface only).
Corer (paleolimnology).
Index Period
None. Samples may be taken annually, biennially, triennially, etc.
Sampling
Single sample in mid-lake.
Analysis
Samples are divested of organic matter and 300-500 diatom frustules are identified to
lowest pratical level.
7-21
-------
Chapter 7
in Table 7-10 are similar to those used in EMAP
(USEPA 1994b),
Sample Location—Sediment samples are ob-
tained from or near the deepest area of the lake.
A single core sample is sufficient (Charles et al.
1994).
Sampling and Analysis—Sediment diatoms can
be sampled with a corer that is able to reliably
sample and retain the top 1 cm of sediment. The
top 1 cm of sediment is carefully removed from
the sampler and kept at 4°C in a plastic bag.
Diatom samples are prepared, enumerated, and
identified following the procedure from the
EMAP manual (USEPA 1994b).
7.41 TIER 2B: SHORT-TERM
INDICATORS (REPEATED
SAMPLING)
Tier 2B consists of phytoplankton and zoop-
lankton sampling in addition to Tier IB sam-
pling, and is conducted at the same sites and
times as Tier IB (Table 7-10). Sampling fre-
quency may range from three samples during
the growing season to monthly samples,
depending on the objectives of the program.
The number of sampling sites is the same as
Tier IB, and samples may be composited
among the sample sites to economize labora-
tory effort, if within-lake spatial variability is
not an Issue.
7.4.1 Phytoplankton
Phytoplankton are subsampled from the same
water sample collected for chlorophyll and
nutrients in the Tier 1 sampling protocol. The
water sample may be a surface sample or an
epilimnion or photic zone hose sample. The
large sample is mixed thoroughly before
subsamples are taken from it.
A sample of 150 to 500ml is sufficient for
phytoplankton. The phytoplankton sample is
preserved In the field with Lugol's solution
(APHA 1992). Cells are identified and counted
using the Utermohl method on an inverted
microscope, or by filtration onto a membrane
filter (APHA 1992). The Utermohl method
requires settling chambers and an inverted
microscope, and the filter method requires a
compound microscope and filtering apparatus.
7.4.2 Zoo plankton
Sampling Procedure—Zooplankton are sampled
with a vertical tow at the same sites as phy-
toplankton, trophic state, and water quality
(Table 7-11). Nets of 118|im mesh and 30 em
diameter will sample most crustacean zooplank-
ton. The net should be equipped with a cone to
prevent spill and escape of active organisms.
Zooplankton are anesthetized with carbonated
water, and preserved in 4 percent formaldehyde.
After fixing, long-term storage should be in 70
percent ethanol.
Analysis—1The sample is split until 100 to 200
organisms remain in the subsample. Zooplank-
ton are identified to genus; equipment includes
dissecting microscope and keys. Lengths of
Daphnia are recorded.
7.4.3 Periphyton
Periphyton should be sampled several times
during the growing season: certain species
might be dominant depending on the time of
year. Field methods are outlined below and
summarized in Table 7-12 (after Bahls 1993).
Sampling Location—A minimum of two random
sampling points along each transect is sug-
gested; a determination of greater sampling
effort should be based on lake size and profes-
sional judgment.
Sample Collection—Collection can be from
natural or artificial substrates depending on
the preference of the investigation team or
agencies. Natural substrates include rocks,
logs, macrophytes. and mud. A composite
sample of three to five substrates (e.g., fist-
sized rocks) is obtained from each sample site.
The area scraped from each substrate should
be approximately equal. Use a pocket-knife or
similar tool for scraping solid substrates. A
spoon or large-bore eyedropper can be used for
lifting microalgae from mud or silt substrates.
Maeroalgae can be picked by hand. Epiphytic
algae can be dislodged from maeroalgae, moss,
and aquatic macrophytes by placing a portion
of the higher plant in the sample container and
7-22
-------
Tiered Sampling
Table 7-10. Tier 2B: Water column biological sampling.
Component
Data Collection
Responds to or Indicator of
Habitat Components
1. Watershed land use, population,
NPDES.
Tier 0 habitat
¦¦ «
2. Lake physical.
Tier 1 habitat.
3. Shorezone habitat assessment.
3-10 transects:
- land use
- bank stability
- riparian vagetation
- emergent vegetation
. .
4. Water quality
DO seasonal or annual mean, %
depth-time
Mean pH, alkalinity
Secchi depth
Tier 1B water quality
(seasonal average).
Trophic state, turbidity.
Biological Components
5. Algal chlorophyll a.
Tier 1B chlorophyll (seasonal
average).
Trophic state.
6,7. Assemblages
(minimum 2):
a. Phytoplankton
Surface samples (0.5 m) or
integrated samples (hose)
Identify to genus; count
100-500 cells.
Trophic state acidity, metals,
water column toxicity.
b. Zooplankton
Vertical tows; identify to
genus; count 100-200
organisms, measure
cladocerans.
Trophic state, contamination,
trophic imbalance.
c. Periphyton
O-;,'
shaking vigorously. The moss or macrophyte is
then removed and discarded (Bahls 1993).
Sample Preservation—Preserve samples in
watertight, unbreakable jars. Water is added
from the sample site to cover the sample; then
enough Lugol's solution is added to impart a
reddish-brown tint. Artificial substrates can be
preserved intact in a suitable container or
scraped in the field.
Sample Preparation—Extracellular organic
matter is decomposed by oxidation, leaving only
the diatom shells (frustules) as described in
APHA (1992). Using the cleaned diatoms (frus-
tules), a permanent mount is prepared and a
proportional count is made of 300 to 500 cells
(APHA 1992). Counts for each species are
divided by the total count and multiplied by 100
to obtain percent relative abundance (PRA).
7.5 DIAGNOSTIC HABITAT
SURVEY
More detailed habitat procedures allow monitor-
ing agencies to focus on specific water and
Habitat
Open water, 1 to 5 sites in lake.
Sampling Gear
Plankton net, 300mm (12in) mouth; 118mm mesh.
Index Period
Mid-summer index.
Sampling
Single vertical tow through water column from 0.5m above bottom to surface.
Analysis
Tier 1: Identify to species, measure Daphnia. 100 to 200 organisms.
Table 7-11. Sampling summary for crustacean zooplankton.
7-23
-------
Chapter 7
Table 7-12. Sampling summary for periphytic diatoms.
Habitat
Rock, wood, silt, macrophyte substrates, 0.5 to 1 m depth (wading depth).
Sampling Gear
Spatula, toothbrush for scraping cells from substrates, eyedropper or spoon for lifting
(Bahls 1993). Samples preserved in Lugol's solution.
Index Period
Preferred: mid-summer.
Alternative: growing season, average of 7 to 10 samples.
Sampling
3-5 substrates (rock, wood, sand, mud, macrophytes) are sampled in the proportion
of their occurrence at 3-5 sites around the lake. Single composite sample from all
substrates and sites.
Analysis
300-500 diatom frustules are identified to species and enumerated.
sediment quality problems in a lake, and
specific land use practices In the watershed, for
identification of probable cause of impairment
fTable 7-13). Supplemental habitat components
may include: a detailed watershed assessment
(soils and geology, detailed land use, agricul-
tural practices); a stream assessment for
migratory fish habitat; additional water quality
analysis (nutrients, contaminants); and sedi-
ment quality (sediment grain size, sediment
organic carbon, contaminants, toxicity).
Tiers 2A and 2B will allow detection of effects of
toxic substances on the respective biological
assemblages, but will not provide positive
identification of toxicity as a probable cause of
Impairment. Positive identification of contamina-
tion and toxicity as a probable cause will require
the supplemental survey, particularly habitat
contaminant analysis and toxicity assays. The
detailed land use measurements In the habitat
assessment allow identification of more specific
nonpolnt source probable causes of impairment.
The tiers allow detection of biological effects on
at least two assemblages, and hence detection of
effects at multiple levels (Including cascades of
effects).
7.6.1 Watershed and
Shorezone Components
The diagnostic habitat survey is similar to the
Her 1 and Tier 2 habitat survey but evaluates
more detailed components. In searching for
probable causes of impairment, land use is
broken down into more detailed land use
categories, Including high- and low-density
residential, industrial and commercial transpor-
tation, cropland, pasture, orchard, mines, etc. If
agriculture is thought to contribute to impair-
ment, then the dominant agricultural practices
should be documented, as well as their distribu-
tion in the watershed. If the fish assemblage
shows impairment (particularly migratory fish),
then fish spawning habitat in inflowing streams
can be evaluated.
7.5.2 Sediment Analyses
TTie Sediment Classification Methods Compen-
dium (USEPA 19920 discusses various aspects
of sediment analyses including sample collection
and handling, quality assurance/quality control
issues, and toxicity testing. In addition, this
guide furnishes references for specific methods.
Sampling
There are three main types of devices used to
collect sediment samples. The choice of sampler
to be used for a particular study depends upon
the nature of the sample needed. Grab samplers
and core samplers can be used in toxicity
testing and in evaluating chemical and physical
properties of the sediment. Additionally, cores
can be used In evaluating historical sediment
records.
Equipment should be thoroughly cleaned
between samples to prevent cross contamina-
tion. In some cases, preservation methods such
as pH control or addition of chemical preserva-
tives will need to be done. Standard methods for
sample handling can be found in ASTM (1990).
Sediment Particle Size
Sediment particle size is measured using stacks
of different sized sieves. The sediment to be
analyzed is first heated to dryness. Samples
may need to be stored cold, frozen, or preserved.
7-24
-------
Tiered Sampling
Table 7-13. Supplemental components.
Component
Data Collection
Responds to or Indicator of
1. Watershed
- Soil and bedrock characteristics
- Hydrology
- Agricultural practices
- Detailed land use categories
(roads, mines, impervious
surface, cropland, pasture, etc.)
Maps; survey of state and county
agencies.
Physical classification.
Probable cause.
2. Shore
- Migrating fish spawning habitat
Tributary stream habitat survey.
Disturbance, habitat
destruction.
3. Sediment quality
- Toxicity, contaminants, total
organic carbon, particle size
Annual grab in depositional
environment (deepest point).
Exposure to toxics,
contaminants.
Then a known weight of dried sediment is
poured into a stack of sieves of different sizes to
separate the particles. Each size fraction is then
weighed and expressed as a percentage of the
total dry sample weight.
Sediment Contamination
Chemical analyses that can be measured
include metals, polyaromatic hydrocarbons
(PAHs). polychlorinated biphenyls (PCBs),
pesticides, and volatile and semivolatile organic
pollutants. Metals are typically measured using
atomic absorption spectrophotometry. Other
constituents should be analyzed using USEPA
approved methods (USEPA 199If, ASTM 1990).
Although it is not a contaminant, total organic
content (TOC) should also be analyzed since it is
an important indicator of the bioavailability of
nonionic hydrophobic organic pollutants.
Likewise, acid volatile sulfides (AVS) are impor-
tant in determining the bioavailability of metals.
Sediment Toxicity Evaluation
Several approaches are recognized by USEPA for
evaluating sediment toxicity. These approaches
may be used separately or in combination to
provide evidence of toxicity and to generate
sediment quality criteria. (USEPA 1994j).
Whole (bulk) sediment toxicity testing is a
method of evaluating the level of toxicity of a
sediment sample. Typically, test organisms are
exposed to sediment for 10 to 14 days. End-
points used are growth and survival. The most
often used organisms in freshwater sediment
toxicity tests are the amphipod Hyctlella azteca
and larvae of the midge Chironomus teutons.
Other organisms that have been tested
include other benthic infauna such as the
mayfly Hexagenta spp; and the worms Tubifex
tubifex and Lumbricuhis variegatus; and two
cladocerans, Daphnta magna and
Cetiodaphnia dubia. Results of exposure to
contaminated sediments is compared with
control (uncontaminated) sediments (USEPA
1994J, ASTM 1998, PSEP 1995, Environment
Canada 1994).
7-25
-------
In This Chapter...
> Characterization
> Metric Selection
> Index Development
Chapter 8
Index Development
a. 1 OVERVIEW
The approach taken here for development of an
index for assessment is called the multimetric
approach. Biological attributes, or metrics, are
calculated from the measurements, A score is
assigned to each metric corresponding to its
deviation from the expected value in reference
sites. The multimetric index is the sum of all the
metric scores. A separate index is developed for
each assemblage sample (e.g., macrophytes,
benthic invertebrates, fishj.
The multimetric approach has been successfully
applied to assessment of stream fish assem-
blages (Karr 1981, Karr 1991, Karr et al. 1986)
and stream invertebrate assemblages (Ohio EPA
1987, USEPA 1989b, Barbour et al. 1995, Yoder
and Rankin 1995). The approach appears to be
statistically robust (Fore et al. 1994) and is
straightforward to apply. Alternative methods of
analysis and assessment are discussed in
Appendix E.
Development of a multimetric index is the final
step toward operational bioassessment. Three
steps are necessary for development of an index:
characterization of reference conditions, evalua-
tion and final selection of metrics, and
multimetric index building.
Development of a
tional bioassessment.
The basis of the multimetric approach is
comparison of a metric to an expected (refer-
ence) distribution of values and a judgement of
whether the value is within the expected
range. Each metric is given an ordinal score of
5, 3, or 1, depending on whether it is similar
to reference values (within the expected range),
is somewhat different, or
is very different, respec-
tively (Figure 8-1). The
expected range is usually multimetric Index Is the
expressed as a percentile
of the reference distribu- final step toward opera-
tion. Two methods of
scoring are commonly
used. The first is based on
a lower percentile of a representative sample of
reference sites (Figure 8-la). The second
method is used if predetermined reference
conditions are not definable or if there are too
few reference sites, and it is preferred for
defining reference conditions for reservoirs. In
the first method (Figure 8-la), the 25th
percentile of the reference site distribution is
often used as the dividing line between optimal
(similar to reference) and less than optimal. In
the second method (Figure 8-lb), the 95th
percentile of the entire population distribution
is often used as the reference mark for trisect-
ing metric values (e.g., Karr et al. 1986).
8-1
-------
Chapters
The index consists of the sum of all metric
scores, and the total index value of a site is
compared to the distribution of index values in
reference conditions. Development of an index
thus requires characterization of reference
conditions to obtain the distributions of metric
values, final selection of metrics based on metric
response to stressors, and, finally, characteriza-
tion of the index distribution in reference
conditions.
Selection of metrics and development of a
multimetric index requires a test data set
composed of reference sites and nonreference
(test) sites. The best sites may be impaired or
may simply not meet the criteria for reference
sites. Ideally, the test sites should include at
least some lakes that are severely impaired by
a. max
METRIC
VALUE
MIN
T
X
25th
%i!e 3
b. max
REFERENCE SCORE
DISTRIBUTION
95th
METRIC
VALUE
MIN
I
%ilB
DISTRIBUTION SCORE
OF ALL SITES
Figure 8-1. Basis of bioassessment scores •
unimpaired reference sites; population
distribution.
different stressors. If, for example, all test sites
are eutrophic lakes, then the response of
metrics to other stressors cannot be determined.
Reference condition characterization uses only
the reference site data—metric evaluation and
index development use both reference and test
site data.
8.2 CHARACTERIZATION OF
REFERENCE CONDITION
The objective of reference characterization is to
finalize the classification of the reference sites
and to describe (characterize) each of the lake
categories in terms of metrics and other descrip-
tive variables.
Several statistical tools can assist in the classifi-
cation of sites, but there is no one set proce-
dure. If the preliminary classification is rela-
tively certain (based on well-developed prior
knowledge and professional judgment, and
graphical analysis of metrics) followed by
necessary modifications and tests of the result-
ant classification, is usually sufficient to finalize
the classification. If the preliminaiy classifica-
tion is less certain, it might be necessary to
develop a classification from the data, using one
of several classification methods. These methods
include cluster analysis and several ordination
methods (e.g., principal components analysis,
correspondence analysis, multidimensional
scaling; Appendix E). Ordination is also useful
for visualizing alternative a priori classification
schemes.
8.2.1 Graphical Analysis
A key analysis method for biological metrics is
graphical displays using box-and-whisker plots
(e.g., Figure 8-1). In the form used here, the
central point is the median value of the variable;
the box shows the 25th and 75th percentiles
(interquartile range); and the whiskers show the
minimum to the maximum values (range). A
common alternative is whisker extending to
values within the "inner fence" (see Tukey 1977
for explanation); this method also plots outliers.
Box-and-whisker plots are simple, straightfor-
ward, and powerful, and the interquartile ranges
are used to evaluate whether there is a real
difference between two areas and whether a
metric is a good candidate for use in assess-
8-2
-------
Index Development
ment. Graphing the data should always be a
first step in data analysis.
Statistical methods used by biologists are
frequently tests of whether two or more popula-
tions have different means using t-tests,
analysis of variance, or various nonparametric
methods. However, the fundamental problem of
biological assessment is not to determine
whether two populations (or samples) have a
different mean, but to determine whether an
individual site (lake) is a member of the least-
impaired reference population. If it is not, then
a second question is how far it has deviated
from that reference. Therefore, biological
assessment requires the entire distribution of a
metric, which is effectively displayed with a
box- and - whisker plot.
In operational bioassessment, metric values
below the lower quartile of reference conditions
are typically judged impaired (e.g., Ohio EPA
1990). The actual percentile chosen (25, 10, or
5) is arbitrary and reflects the amount of
uncertainty a monitoring program can tolerate.
8.2.2 Characterization
The preliminary classification is refined through
inspection of plotted data (graphical analysis),
professional judgment, and
statistical tests of final classifica-
tion hypotheses. First, the values
and distribution of metrics are
compared among ecoregion or
lake type. Regions that appear to
be similar to each other can be
lumped together for final classifi-
cation. For two regions to be
lumped, most of the metric
distributions must be similar. In
addition to box plots of metrics,
it is also useful to examine
scatter plots of selected metrics
and habitat variables such as
lake size, salinity, or alkalinity.
The number of taxa in a
waterbody is often dependent on
its size, for example, large lakes
have more zooplankton species
than small lakes (Dodson 1992).
Salinity also influences the
number of species found in
aquatic systems, as do pH and
alkalinity.
Refining the Classification
In sampling fish from reservoirs of the Tennes
see Valley Authority, the number of fish
species was found to
vary by reservoir class
and ecoregion (Hickman
and McDonough 1996).
Figure 8-2 (after
Hickman and
McDonough 1996) shows
the number of fish
species in different parts
of four groups of TVA
reservoirs. First, the
number offish taxa is
relatively homogenous
between forebay, transi-
tion, and inflow zones
(Figure 8-2). The
reservoir types differ in
number of fish species,
with the mainstream reservoirs having the
most species, and the Blue Ridge reservoirs
being relatively depauperate. Based on num-
ber of species, the Interior Plateau reservoirs
are not significantly different from Ridge and
Valley reservoirs, and TVA reservoirs could be
considered to be in three groups (dotted lines)
However, on the basis of other considerations,
TVA has kept Interior Plateau reservoirs
The fundamental problem
of biological assessment is
not to determine whether
two populations have a
different mean, but to
determine whether an
individual site is a member
of the least-Impaired
reference population.
50-
45**
-------
Chapters
separate from Ridge and Valley reservoirs.
Refining the Classification-Covariates
Certain physical or chemical attributes can have
a strong influence on biological metrics, espe-
cially number of taxa metrics. The most impor-
tant of the physical-chemical attributes to test
are lake size, salinity (in arid regions), and
alkalinity or pH. The example (Figure 8-3) shows
number of taxa of benthic macroinvertebrates as
a function of salinity in the littoral zone of
Montana lakes and wetlands (Stribling et al.
1995). Finding a relationship as in Figure 8-3
requires adjusting reference expectations as a
function of the covariate salinity in this case.
8.3 INDEX DEVELOPMENT
Following classification and characterization of
reference conditions, metrics are evaluated for
suitability in a multimetric index. Suitable
metrics are those that respond in a predictable
way to stressors on the system and that have
low noise or variability.
S.3.1 Metric Variability
Metrics that are too variable within the reference
sites are unlikely to be effective for assessment.
A measure of metric variability Is the ratio of the
interquartile range to the distance between the
lower quartile and the minimum possible value
of the metric.
In operational bioassessment, metric values
below the lower quartile of reference conditions
are typically judged as not meeting reference
expectations (e.g., Ohio EPA 1990). The range
from 0 to the lower quartile can be termed a
"scope for detection." For those metrics with low
values under reference conditions and high
values under impaired conditions, the scope for
detection is the range from the 75th percentile
to the maximum possible value (e.g.. 100
percent) (Figure 8-4). The larger the scope for
detection, compared to the interquartile range,
the easier it will be to detect deviation from the
reference condition. The "interquartile coeffi-
cient" is thus defined here as the ratio of the
interquartile range to the scope for detection.
The interquartile coefficient is analogous to the
coefficient of variation and is used the same
Salinity vs. Total Taxa
40
35
30
25
20
& 15
.o
10
5
0
-5
-0 j
o
o
o o
CD
o o
CD
0.5 1.5 2.5 3.5
Salinity (%)
4.5
5.5
Figure 8-3. Benthic macroinvertebrate taxa richness in littoral zone of Montana lakes and wetlands.
8-4
-------
Index Development
a. Max
interquartile
range
scope for
detecting
impairment
Min
b. Max
scope for
detecting
impairment
Min
~
interquartile
range
Figure 8-4. Assessing candidate metrics, a. Metrics that have high values under unimpaired conditions,
b. Metrics that have low values under unimpaired conditions.
8-5
-------
Chapters
way, but it is bidirectional and uses percentiles
in the same way that assessment uses percen-
tiles. In general, an interquartile coefficient
greater than 1 indicates excessive variability of a
metric.
8.3.2 Metric Response
Response of metrics to stresses is evaluated by
comparison of reference sites to test sites. The
simplest comparison is using box-and-whisker
plots of the metric distribution in reference and
test sites (Figure 8-5}. Alternatively, it may be
possible to develop an empirical model of metric
response to stressors. Several approaches are
available including multiple regression, canoni-
cal correlation, canonical correspondence
analysis, and log-linear models (Ludwig and
Reynolds 1988, Jongman et al. 1987). For
multivariate model building, refer to the above
references or any statistical software package—it
will not be outlined further in this document.
Metrics are judged responsive if there are
significant differences in central tendency or in
variance between reference and test sites (Figure
8-5). If the test sites are known to be impaired,
then the mean or median values should be
significantly different (Figure 8-5). If the test
sites are simply lakes that do not meet reference
criteria (i.e., they might be a mix of impaired
and unimpaired lakes; shown as "unknown test
I
I
REFERENCE
SITES
IMPAIRED
TEST
SITES
UNKNOWN
TEST
SITES
Figure 8-5. Responsiveness of metrics. A large
difference between reference and impaired test
sites Indicates a responsive metric. Unknown sites
are a mixture of Impaired and unimpaired sites.
Variability and Uncertainty
Variability in values of measurements and metrics
results in uncertainty of the assessment. Uncer-
tainty can be reduced by increasing the sampling
effort (repeated measurement) to obtain a better
estimate of the mean value. This is especially im-
portant for the measurements that are the most
variable: chlorophyll, nutrient concentration phy-
toplankton and zooptankton. Algal abundance and
biomass may vary tenfold within the growing sea-
son (i.e., Wetzel 1975, Hecky and Kling 1981). A
tenfold change in chlorophyll corresponds to 22.6
points in the TSI range, a substantial change.
Because of this variability, Tier 1A Is unreliable
for assessment of an individual lake and Tier2A
is recommended. Tier 1A is appropriate for as-
sessing a class of lakes or a region, to answer
questions such as: what Is the status of lakes in
the region, or how many lakes are oligotrophia?
As long as many lakes are sampled, the effect of
errors in individual lakes is reduced in the evalu-
ation of all lakes.
sites" in Figure 8-5), then the variance in the
test sites should be larger than that in the
reference sites.
Metrics that are responsive to known or un-
known stresses are retained for index develop-
ment. Finally, responsive metrics are evaluated
for redundancy. A metric that is highly corre-
lated with another metric might not contribute
new information to the assessment. Pairs of
metrics with correlation coefficients greater than
0.9 should be examined carefully to determine
whether both metrics are necessary. Often,
strongly correlated metrics are calculated from
the same raw data, or their method of calcula-
tion ensures correlation. For example, Shannon-
Wiener diversity and percent abundance of the
dominant taxon are strongly correlated in any
data set.
A correlation alone (say, r >0.6) is not sufficient
to eliminate one of a pair of correlated metrics.
Some metrics might be sensitive only at severe
or moderate stress; others might be sensitive
across the entire range of stresses (Karr 1991).
These would all contribute information, in spite
of strong correlation, A scatterplot of correlated
metrics is examined; if there is an apparent
8-6
-------
Index development
nonlinear or curved relationship, then both
should be retained. If the points all fall close to
a straight line, then one of the metrics can be
safely eliminated.
8.3.3 Scoring and Index
(development
Combining unlike measurements is possible
only when the values have been standardized by
a transformation through which measurements
become unitless (Schuster and Zuuring 1986).
Standardization of these measurements into a
logical progression of scores is the typical means
for comparing and interpreting unlike metric
values.
Two methods are commonly used for scoring
metrics, which are based on the metric distribu-
tion in defined reference sites or in the popula-
tion of sites, respectively. Each metric is given a
score of 1, 3, or 5, corresponding to impaired,
intermediate, or unimpaired biota, respectively
(Figure 8-1).
Bisection scoring—(Figure 8-la) Based on a lower
percentile of the reference distribution; for
example, the 25th percentile (Barbour et al.
1996b). In this method, values above the 25th
percentile are considered unimpaired (similar to
reference conditions) and values below the 25th
percentile are considered impaired to some
degree. The range from 0 to the 25th percentile
is bisected, with values in the top half receiving
a score of 3 and those in the bottom half receiv-
ing a score of 1 (Figure 8-la).
Trisection scoring—(Figure 8-lb) Based on the
95th percentile of the population distribution
(Karr et al. 1986). Metric values from 0 (or the
lowest possible value) to the 95th percentile are
trisected; values in the top one-third receive a 5,
values in the middle third receive a 3, and
values in the bottom third receive a 1 (most
impaired).
The scoring method should reflect how well the
reference sites represent unimpaired conditions.
If reference sites are unimpaired and considered
to be representative, bisection is recommended
(Figure 8-la). This method assumes that the
reference sites are representative of relatively
unimpaired conditions and that the metric
distribution reflects natural variation of the
metric. A value above the cutoff is then assumed
to be similar to reference conditions. The lower
quartile (25th percentile) is most frequently
taken as the cutoff (e.g., Barbour et al. 1996b).
Two methods are com-
monly used for scoring
metrics which are based
on the metric distribution
in defined reference sites
or in the population of
sites respectively.
The trisection method
(Figure 8- lb) is best for
scoring in regions where
impacts might be so
pervasive that nearly all
reference sites are thought
to be impacted or for
assessment of reservoirs
where reference sites
cannot be defined. In
trisection, it is assumed
that at least some refer-
ence lakes attain an
excellent value for the
metric, but that many reference lakes are
impaired and hence the lower limit of the
reference distribution is not known. The 95th
percentile is taken as the "best" value, and the
range is trisected below it (Figure 8-lb).
Choice of scoring method should be based on
confidence in the reference sites, rather than on
the method that will produce the most conserva-
tive or most liberal scoring. If confidence is high
that reference sites are representative of rela-
tively unimpaired conditions, then the lower
percentile cutoff and bisection are preferred. If
confidence is low, then trisection below the 95th
percentile is preferred.
If covariates such as lake size determine metric
values, then the scoring should be adjusted for
the covariates. Reference data are plotted as in
Figure 8-6, and a locally weighted estimate is
made of the appropriate percentile (95th or
25th) and the range below it is trisected or
bisected accordingly. Figure 8-6 shows total
zooplankton taxa in North American lakes
ranging in size from 4m2 to nearly 10um2 (Lake
Superior) (Dodson 1992). Few state assessment
programs are likely to include lakes smaller
than 104m2 (lha; 2.47 acres), nor larger than
109m2 (1000km2; 247,000 acres). In this ex-
ample, considering only the middle range from
104m2 to 109m2, the slope is not apparent and
adjusting for lake area would not be necessaiy.
The index is the sum of the scores of the se-
lected metrics. The number of metrics in an
index affect the variability of the index—those
with more metrics tend to be less variable (Karr
1991). Index values are evaluated by compari-
8-7
-------
Chapters
Log Lain Area (mA2)
Figure 8-6. Total crustacean and zooplankton taxa in North American lakes (redrawn after Dodson 1992). if
metrics show a relationship such as this with area, elevation, or some other physical covariate, then reference
expectations must be adjusted to the covariate. The three lines show one possible method for scoring. In
practice, most state assessment programs are not likely to span 10 orders of magnitude in lake area.
son to index values of the reference sites. Even
the best reference sites do not receive perfect
scores of the index. The final index scores are
compared to the
Choice of scoring method
should be based on confi-
dence In the reference sites
rather than on the method
that will produce the most
conservative or most liberal
scoring.
distribution of scores in
the reference sites.
Criteria for assessment
are based on the
distribution of index
scores in reference
sites. Those that
correspond to the range
of index values in
reference sites support
life use; those that are
clearly below index
values in reference sites do not support life use.
Following appropriate review and revision, they
can be established as biocriteria.
8.4 LAKE TIER INDICES
An index is calculated for each assemblage
sampled. Each tier has three to six indices,
which should all be reported. The indices can be
summed into an overall lake index, which can be
used to report overall condition but would not
reveal the condition of the component assem-
blages. Indices within each tier might or might
not be multimetric; Her 1 indices are primarily
single metrics, whereas indices of Tiers 2A and
2B might be composed of 3 to 12 metrics.
8.4.1 Tier 1
Tier 1 assessment consists of trophic state algal
growth potential and macrophyte indices. Three
TSI (chlorophyll, Secchi depth, and total phos-
phorus) are recommended; the fourth (total
nitrogen) is also recommended in regions where
nitrogen is suspected to be a limiting nutrient
for algal growth. The TSI and AGPT are scored
as metrics for their similarity to reference
conditions, and the scores are summed for a
"Trophic Reference Index."
The trophic metrics are unique in that they may
be scored lower if their values are substantially
higher as well as lower than reference values. For
example, an unproductive (oligotrophic) lake in
a region where lakes are expected to be produc-
tive (mesotrophic) would be given a lower score.
Tier 1 has two or more submerged macrophyte
metrics, percent cover of macrophytes, and
8-8
-------
Index Development
dominance of exotic species. More metrics can
be developed if macrophyte species are identified
and relative abundances are estimated. Percent
cover is scored by comparison to reference
expectations, but dominance of exotic species is
rated 5 if none are present, 3 if exotics are
subdominant, and 1 if exotics are dominant.
The two macrophyte metrics are summed for the
Tier 1 macrophyte index.
Lakes are assessed from the scores of the two
Tier 1 indices. Tier 1A and Tier IB use the same
metrics and indices; Tier IB trophic metrics are
estimated from seasonal mean measurements.
Biocriteria can be established for further investi-
gation or remedial action, based on the scores.
8.4.2 Tier 2A
Tier 2A assessment may consist of three to live
indices:
• Trophic reference index of either Tier 1A or
IB.
* Macrophyte index of Tier 1 or a mere
detailed Tier 2A macrophyte index.
• Benthic macroinvertebrate index.
• Fish assemblage index.
• Sedimented diatom index.
The macroinvertebrate, fish, and diatom indices
are developed from metrics as described in
Chapter 6.
8.4.3 Tier 2B
Tier 2B consists of three to five indices:
• Trophic reference index of Tier IB (seasonal
averages).
• Macrophyte index of Tier 1.
• Phytoplankton index.
• Zooplankton index.
• Periphyton index.
The phytoplankton, zooplankton. and periphy-
ton indices are developed from metrics as
described in Chapter 6.
Case Study: TVA ScorlngCriterla and Index Development
Chlorophyll The classification scheme used b" levels not become too great because of the associ-
: develop expectations for chlorophyll in Tennessee ated undesirable conditions—dense algal blooms,
: Valley reservoirs was based oh the "natural" nutri- ^poot water clarity, low DO, and noxlous^blue-greeh
ent level in a watershed. Pmfessibnafjudgmentwas ' algae. Conversely, in oases where sufficient nutii-
used to select concenti^iOnsconsi0ere0n^cative entsare available but" chlorophyll concentrations
of good, fair, and poor conditions. ReservOits were remain low, them 'is $mlysdm^inghinderihg0s:{
placed into one ©x* natural process.
pectations; those expected to be oligotrophia bo- minimum level for the "gOptf range of expectations
cause they are in watersheds tilth naturally low for mesotrophio reservoirs,
nutrient concentrations, and those expected to be ; ¦
mesotrophio because they are in watersheds which ' Sediment Quality The sediment quality scoring
naturally have greater nutrient availability. The res- criteria -
ervoirs expected to be oligotrophia are those in the monia' heavymetals> pesticides, and PCBs. : ;
Blue Ridge Ecoregion. T^e remaining reservoirs, Benthic Nlaetolhyehebrates Seven assemblage
.both mainstream reseivpiis andiri^tJtary reservoirs, characteristics (or
ff® expected to be mesotmphic. ^ ate the benthic macroinvertebrate assemblage. Six
J The ^^cOh^n^m^iBctod to represent of the m0tric^ are an avera^e of thB "> samples -
good, fair, and poor conditiorisis much lower for taken at0achsite.
Ridge) ' j, Number of taxa.'
~ jhah for the other reservoirs. For reservoirs expected ,' " -
S " to be rrieisotrophlc, Vte concern is that chlorophyll 2- EPTtaxa.
8-9
-------
Chapters
- CSttsa Study: TVA Scoring Criteria and Index Development (Continued)
3. Percent of samples with long-lived species.
4. Proportion as Tubiftcidae.
5. Proportion as two dominant taxa.
6 Total abundance excluding Chironomidae and
Tubifiddae.
7. Percentage of samples with no organisms
present.
Scoring criteria for each of the seven metrics were
developed using the 5 years of Vital Signs monitor-
ing data (1994-1996). Scoring ranges were devel-
oped as follows:
• Individual criteria were developed for each type
of sampling location (forebay, transition zonef
mid-reservoir, embayment and inflow) for each
of the fourctasses of reservoirs.
• Results from the 10 samples along a transect
for each sample year were combined (averaged
for most metrics) and outliers deleted.
• The range of average values was then trisected;
w^t^up^erorw-third oi'therange represent-
ing desirabfeTconditions assigned a value of 5
(good), the middle one-third assigned a 3 (fair),
and the lower one-third representing undesir-
able conditions assigned a 1 (poor).
Professional judgment and supplementary statisti-
cal analyses were used to adjust the cutoffs for each
range as appropriate. Sample results at each site
were compared with these criteria for each metric
andassigned the rating described above—5=good;
3s fair;! = poor if they fell within the top, middle, or
bottom group, respectively. Numerical ratings for the
seven metrics wore then summed. This resulted in
a minimum score of 7 if all metrics at a site were poor,
and a maximum score of 35 if all metrics were good.
Reservoir Fish Assemblage Index The current
RFAI uses 12 fish assemblage metrics from five gen-
eral categories, including:
Species Richness and Composition
1. Total number of species.
2. Number of piscivore species.
3. Number of sunfish species.
4. Number of sucker species—suckers are also
insectivorous but Inhabit the pelagic and more
riverine sections of reservoirs.
5. Number of intolerant species.
6. Percentage of tolerant individuals (excluding
young-of-year).
7. Percentage of dominance by one species.
Trophic Composition
8. Percentage of individuals as omnivorous,
9. Percentage of individuals as insectivorous.
Reproductive Composition
10. Number of lithophllic spawning species.
Abundance
11. Total catch per unit effort (number of individu-
als).
Fish Health
12. Percentage of individuals with anomalies (dis-
eases, lesions, tumors, external parasites, de-
formities, blindness, and natural hybridization).
Establishing scoring criteria (reference conditions)
by trisecting observed conditions requires a sub-
stantial data base for each class of reservoir and
assumes the data base contains reservoirs with con-
ditions ranging from poor to good for each metric.
The smaller the number of reservoirs within a class,
the less likely these assumptions can be met and
the greater the need for sound professional judg-
ment based on extensive knowledge of the reser-
voir assemblages being studied.
Because some reservoir classes contained relatively
few reservoirs, the approach used to develop scor-
ing criteria for RFAI was to Include all sampling re-
sults from Vital Signs monitoring (1990-1994). A
slightly different approach was used for species rich-
ness metrics th^rifpr ^bup^ance and proportional
metrics. For species richness metrics, a list was
made of all species collected from comparable lo-
cations within a^esarypiLctess from 1990 to 19.94, r
This species list was adjusted using inferences of
experienced biologists knowledgeable of the reser-
voir system, resident fish species, susceptibility of
8-10
-------
Index Development
Case Study; TVA Scoring Criteria and index Development (Continued)
each species to collection methods being used, and
effects of human-induced impacts on these species.
This effort resulted in a list of the maximum nurhtier.
of species expected to occur at a sampling location
and be captured by collection devices in use. Given
that samples are collected once each year, this
maximum number of species would not be expected
to be represented in that one collection. Therefore,
the range from 0 to 95 percent of the maximum was
trisected to provide the three scoring ranges (good,
fair, and poor). Although 95 percent of the maximum
number of species at a site would not be expected
to be collected in one sampling event, thishigh"
expectation was adopted to keep these metrics con-
servative in light of potential uncertainties introduced
by relying heavily on professional judgement.
Scoring criteria for proportional metrics and the
abundance metric were determined by trisecting
Observed ranges after omitting outliers. Next, cut-
off points between tlie three ranges were adjusted
based on examination of frequency distributions, of
observed data for each metric along with profes-
sional judgment. In some cases, the narrow range
of observed conditions required further adjustment
based on knowledge of metric responses to human-
induced impacts observed in other reservoir classes..
Scoring criteria for the fish health metric are those
described by Karr et al. (1906). '
To develop metric scores for number of taxa, re-
productive composition, and fish health metrics,
electrofishlng and experimental gill net sampling re-
sults were pooled prior to scoring. For abundance
and proportional metrics, electrofishlng and gill net-
ting results were scored separately, then the two
scores averaged to arrive at a final metric value.
These: scoring criteria separated sites into three
categories assumed to represent relative degrees
of degradation. Sample results are compared to
these reference conditions and assigned a corre-
sponding value: good = 5, fair = 3, arid poor = 1.
Overall Assessment
To arrive at an overall health evaluation for a reser
voir, the sum of the ratings from all sites are totaled,
divided by the maximum potential ratings for thai
: reservoir, and expressed as a percentage: For ex-
ample, for a small reservoir with only one sample
side, the health evaluation would be 20% (ail five
indicators rated poor—1 for a total score of 5 di-
vided by tbe maximum possible total of 25) and the
maximum would be 100% (all five indicators rated
good—5). This same range of 20 to 100 percent
applies to all reservoirs regardless of the number
of sample sites, and the same calculation process
is used. v
The next step is to divide the 20 to 100 petcent
scoring range into categories representing good,
fair, and poor
has been achieved as follows: , ;
f. Results are plotted and examined for appar- '
ent groupings. ' '
2. Groupings are compared to known, a priori con-
'¦ ditiohs (focusing on reservoirs with known poor ,
conditions), and good- fair and fair-poor bound-
aries are istabttshed subjectively^
3. The groupings amcomparedtoatriseGtionof
the overall scoring range. A scoring range is
~ - adjusted up or down a few percentage points
to ensure imown conditions
' i falls within the appropriate category. This Is
; - done only in circumstances where a nominal
adjustment is necessary. i
These methodshavebeen in use for 6 years. Each
year slight modifications are made in the original
evaluation process and the numerical scoring cri-
teria for each, of the five, ecological health indica-
tors (Table 8-i) based on experience gained from
working with this process, reS/iew of the evaluation
scheme by other professionals, and results of an-
other year of monitoring. As a result, scoring Ganges
have changed slightly over the years. Low DO and
poor benthos quality contributed most to poof
scores among tributary reservoirs in 1994 (Figure
8-7). Reservoir health raiings^also differed among
ecoregions (Figure &-8), with ruh-df-river reservoirs
typically scoring highest
8-11
-------
Chapters
Case Study: TVA Scoring Criteria and Index Development (Continued)
Table 8-1. Example of TVA's computational method for evaluation of reservoirs: Wilson Reservoir
1994 (run-of-the-rlver reservoir).
Aquatic Health Indicators
Observations*
Ratings*
Forebay
Inflow
Forebay
Inflow
Dissolved Oxygen
- Less than 2 mg/L (summer avg.)
- % of X-sectionai area
- % of X-secttonal bottom length
- Less than 5 mg/L at 1.5 m (Yes/No)
0.4[5]
10.7 [2]"
No
Tailrace DOs
No
3.5 (fair)
5 (good)
Chlorophyll a mg/L
- Summertime average
- Maximum concentration
13.5
30.0
No samples
3 (fair)
No rating
Sediment Quality
- Toxicity
- Ceriodaphnia survival
- Rotifer survival
- Chemistry
- Metals/NH/pesticides
Rating = 1
Yes-0% sur.
Yes-30% sur.
Rating = 5
None
No samples
3 (fair)
No rating
Benthic Assemblage
- Dominance
- Tublficidae
- Chironomldae
- EFT
- Long-lived
- No. of taxa
- Zero in sample
- Non-tolerant density
TotaJ
1
5
1
1
1
1
5
1
20
5
3
5
3
5
5
5
3
34
2 (poor)
5 (excellent)
Rsh Assemblage
- RFAI
45
40
4 (good)
3 (fair)
Sample Location Sum.
15.5 of 25
13 of 15
Reservoir Sum. .
28.5 of 40 (71%)
Overall Reservoir Evaluation. ......
"fair" (yellow)
* No samples taken from transition zone
" DO was 0 mg/L on the bottom in forebay
;;
8-12
-------
-------
00
1
~Jk
A
• .< I ' ¦ L
, Ca«a Sfutfy; TV4 Scaring Criteria and Index Development (Continued)
100
90
80
70
60
Run-of-River
Reservoirs
Nickajack (90) H
Chickamauga (87) B
Pickwick (84)
Guntersville (83) ¦
Watts Bar (79) „
Wheeler (75) u
Kentucky (71}-pfc ¦
Teiiico giyy
Wilson (71)
Fort Loudoun (61) a
Blue Ridge
Ecoregion
Reservoirs
Blue Ridge (86) a
Chatuge (77)
Hiwassee (68) _
Fontana (67) m
Parksville (60)
Nfteiy(56)'
Ridge & Valley
Ecoregion
Reservote
Norris (69)
S. Hols ton (66)
Watauga (65)*
Douglas (64)
Fort Pat (60)
Boone (59)
Cherokee (53)
Interior Plateau
Ecoregion
Reservoirs
I
l|:|.
Cedar (80)
¦
¦
Normandy (68) H
Little Bear (64) a
Tims Ford (58)** B
Beech (56^ ¦
oeecn \pojTr
Bear (56J
Benthos & Fiih corop«red to Blue Ridge Ecorcgion rarevoits although WtUuga phy»ically within Ridge & Valley Ecoregion.
Ben&os & Fish equipped to Ridge & Vtltey Ecotegioa naervoin iltboughTinu Fori phytictlly within Interior Plaam Ecoregion.
O.
: O:
a
:¦ V"H
'B
Ph
u
O
~o
Oh
Figure 8-8.1994 TVA ecological condition summary.
-------
in This Chapter.
> Program Design
> Sampling Design
> Precision
> Quality Assurance
> Operational Quality Control
Chapter 9
Quality Assurance: Design,
Precision and Management
Qualify assurance (QA) is an integrated program
for ensuring the reliability of monitoring and
measurement data and includes quality control.
Quality control (QC) refers to operational proce-
dures for obtaining prescribed standards of
performance in the monitoring and measure-
ment process. Specific QC elements can be
developed for most, if not all, project activities.
All project activities, from sampling (data
collection) and laboratory analysis to statistical
analysis and reporting, are potential error
sources (Peters 1988). Because error is cumula-
tive and can significantly affect the results of a
project, all possible efforts must be made to
control it. Therefore, quality assurance is a
continuous process that should be implemented
throughout the entire development and opera-
tion of a program.
The purpose of an overall quality assurance
project plan (QAPP), containing specific QC
elements and activities, is to minimize—and
when possible eliminate—the potential for error.
Additionally, there are objective mechanisms for
evaluating activities relative to pre-established
measurement quality objectives and other project
goals. The appropriateness of the investigator's
methods and procedures and the quality of the
data to be obtained must be ensured before the
results can be accepted and used in decision
making. QA is accomplished through:
• Program design.
• Investigator training.
• Standardized data gathering and process-
ing procedures.
• Verification of data reproducibility.
• Instrument calibration and maintenance.
As outlined below, QA
requirements apply to all
activities in an ecological
study. More detailed
guidance and examples
for QA activities should be
obtained from USEPA
(1994d, 1995, 1996c);
Quality assurance is a
continuous process that
should be implemented
throughout the entire
more general guidance is development and operation
outlined by USEPA
(1993b). of a program.
9.1 PROGRAM DESIGN
A central component of QA is overall study
design, which includes formulation of ques-
tions and hypotheses, experimental design,
and development of analysis approaches. The
classical approach by which scientists plan
research consists of the following steps:
-------
Chapter 9
Statement of the problem to be resolved.
A central component of
0-4 is ovarall study
design, which Includes
formulation of questions
and hypotheses,
experimental design,
and development of
analysis approaches.
• Formulation of alterna-
tive hypotheses that will
explain the phenomena
or, in the case of prob-
lems that do not Involve
elaboration of processes,
formulation of specific
research questions.
• Establishment of bound-
aries within which to
resolve the problem.
• Formulation of an
experimental or study design
that will falsify one or more hypotheses or
answer the specific research questions.
• Establishment of uncertainty limits including
setting acceptable probabilities of Type I and
Type II errors for statistical hypothesis testing.
• Optimization of the study design including
power analysis of the statistical design.
Experimental advances In basic sciences have not
included the last two steps because uncertainty
limits were inappropriate or unknown. Examina-
tion of experimental advances also reveals that a
high degree of creativity and insight is required to
formulate hypotheses and study designs; no
formal planning process or "cookbook" can
guarantee creativity and insight. Nevertheless,
documentation of the planning process and a
complete explanation of the conceptual framework
help others evaluate the validity of scientific and
technical achievements.
9.1.1 Specifying the
Questions
The first task in developing
a sampling and assessment
program is to determine,
and be able to state in
simple fashion, the principal
questions that the sampling
program will answer:
The first task in devel-
oping a sampling and
assessment program is
to determine, and be
able to state in simple
fashion, the principal
questions that the
sampling program will
answer. Questions may
or may not be framed as
hypotheses to test,
depending on program
objectives. For example, suppose that a sam-
pling program objective is to establish reference
conditions for biological criteria for lakes in
state Y. Typically, the initial objectives of a
survey designed to develop criteria are to
identify and characterize classes of reference
lakes. Initial questions may then Include;
• Should state Y*s minimally disturbed lakes
be divided into two or more classes that
differ in biological characteristics and
dynamics?
• What are the physical, chemical, and
relevant biotic characteristics of each of the
lake classes?
After state Y*s monitoring and assessment
program has developed biological criteria, new
questions need to be developed that encompass
assessments of individual lakes, groups of lakes,
or lakes of an entire region or state. Specific
questions may include:
• Is lake Z similar to reference lakes of its
class (unimpaired), or Is it different from
reference lakes (altered or impaired)?
• Overall, what is the status of lakes in state
Y? How many (or what percentage) lakes
are similar to reference conditions? How
many lakes are impaired?
• Has lake Z changed over a certain period?
Has it improved or deteriorated?
• Overall, have lakes in state Y Improved or
deteriorated over a certain period? Have
Individual lakes improved? Are more lakes
similar to reference conditions now than
some time ago?
Finally, resource managers often wish to deter-
mine the relationships among variables, that Is,
to develop predictive, empirical (statistical)
models that can be used to design management
responses to perceived problems. Examples of
specific questions include;
• Can trophic state of a lake be predicted by
areal phosphorus loading rate (e.g.,
Vollenweider 1968)?
• Can the biota of a lake be predicted by
watershed land use (e.g., Dillon et al. 1994)?
9-2
-------
Quality Assurance: Design, Precision and Management
These same models (e.g., analysis of variance,
regression) are also used to help develop hy-
potheses on causal relationships between
stressors and responses of systems. Establish-
ing cause requires manipulative experiments,
and since surveys and monitoring programs
preclude experimental investigations, inference
of causal relations will not be considered here.
Often, there is enough experimental evidence
available from other studies so that additional
causal experiments are not necessaiy and would
be superfluous (e.g., current knowledge of
nutrients and trophic state generally makes it
unnecessary to "prove" experimentally which
nutrients are limiting).
9.1.2 Specifying the
Population and Sample
Unit
Sampling is statistically expressed as a sample
from a population of objects. In some cases, the
population is finite, countable, and easy to
specify, e.g., all lakes in state Y, where each lake
is a single member of the population. In other
cases, the population is more difficult to specify
and may be infinite, e.g., lake waters of state Y,
where any location in any lake defines a poten-
tial member of the population (Thompson 1992).
Sampling units may be natural units (entire
lakes, cobbles in a littoral zone), or they may be
arbitrary (plot, quadrat, sampling gear area or
volume) (Pielou 1977). Finite populations may
be sampled with corresponding natural sample
units, but often the sample unit (like a lake) is
too large to measure in its entirety, and it must
be characterized with one or more second stage
samples of the sampling gear (bottles, benthic
grabs, quadrats, etc.)
In most sampling designs, each sample unit is
assumed to be independent of other sample
units. The objective of sampling is to best
characterize individual sample units in order to
estimate some attributes (e.g., number of taxa,
DO) and the statistical parameters (e.g., mean,
median, variance, percen-
tiles) of a population of
sample units. The objective The objective of sampling
of the analysis is to be able
to say something (estimate)
about the population. It is individual sample units in
critical to distinguish
between making an infer- order to estimate some
ence about a population of
many lakes (e.g., "Reser-
voirs in the Blue Ridge are
deep and oligotrophies
/s to best characterize
attributes and the statisti-
cal parameters of a
versus an inference about a population of sample units.
single lake (e.g., "Lake Z has
fewer fish species than
unimpaired reference lakes"). These two kinds
of inferences require different sampling designs:
the first requires independent observations of
many lakes and does not require repeated
observations within sample units (pseu-
doreplication) (Hurlbert 1984); while the second
often does require repeated observations within
a lake. Table 9-1 depicts some examples of
sample units and populations.
9.1.3 Specifying the Reporting
Unit
Finally, it is necessary to specify the units for
which results will be reported. Usually, these
units are the population (e.g., all lakes), but
often subpopulations (e.g., lakes within a given
lake district) and even individual locations (e.g.,
lakes of special interest) can be used. Subpopu-
lations, or strata, are more homogeneous than
Table 9-1. Examples of sample units and populations.
Sample Unit
Sample Population
Infinite or Finite Population
A point in a specific lake.
All points in the lake
Infinite
A point in any lake of
a state or region.
Total surface area or
volume in a state or region
Infinite
A lake or a definable subbasin
of a lake as a single unit.
(NOTE: Because lakes are most often
discrete environments, this is likely to
be the most common sample unit)
All lakes in a state or region
Finite
9-3
-------
Chapters
the entire population, and are separated to
facilitate comparison among them (see Section
9.2.1). In order to help develop the sampling
plan, it Is useful to create hypothetical state-
ments of results in the way that they will be
reported, for example:
• Status of a place: Lake Z is degraded.
• Status of a region: 20% of the lake area in
state Y has an elevated trophic state, above
reference expectations; or 20% of lakes in
state Y have an elevated trophic state.
• Trends at a place: Benthic species richness
in lake Z has decreased by 20% since 1980.
• Trends of a region: Average lake trophic
state in state Y has increased by 20% since
1980; or Average benthic index values in
20% of lakes of state Y have increased by
15% or more since 1980.
• Relationships among variables: 50% in-
crease of P loading above natural back-
ground is associated with decline in number
of taxa of benthic macroinvertebrates, below
reference expectations; or Lakes receiving
runoff from large impervious parking lots
have 50% greater probability of elevated
trophic state above reference than lakes not
receiving such runoff.
Specification of reporting units helps to focus
the study design on relevant questions. Alterna-
tive designs can be examined for their ability to
address the questions within the specified
reporting units. Elements of the design that are
not relevant to questions and reporting units are
identified as superfluous.
trolled measurement error and, (2)characterize
and partition the natural variability. For
example, we may stratify lakes by soil phospho-
rus content of the surrounding watersheds (e.g.,
Rohm et al. 1995) so that lakes within a soil P
class may be likely to have similar water column
total P concentrations. Typically, we stratify so
that observations (sample units) from the same
stratum will be more similar to each other than
to sample units in other strata.
When sampling lakes we often measure some-
thing (say, chlorophyll concentrations) at single
points in space and time (center of the lake, 2m
depth, 10 AM on 2 July). If we make the same
measurement at a different place (littoral zone,
1 m) or time (30 January), the measured value
will be different. These two natural components
of variability (space and time in this example)
are called sample variability or sampling error
(Fore et al. 1994). A third component of variabil-
ity, called measurement error, refers to our
ability to accurately measure the quantity we
are interested in. Measurement error can be
affected by sampling gear, instrumentation,
errors in proper adherence to field and labora-
tory protocols, and the choice of methods used
in making determinations. The three basic rules
of efficient sampling and measurement are:
1. Sample so as to minimize measurement
error.
2. Characterize the components of variability
that have influence on the central questions
and reporting units.
3. Control other sources of variability that are
not of interest and thus minimize their
effects in the observations.
9.2 SAMPLING DESIGN
9.2.1 Sources of Variability
Variability of data justifies the existence of
statistics. Variability has many possible
sources. The intent of sampling designs is to
collect a representative
Variability of data Justifies of the P°P"latlon-
For bioassessment,
the existence of statistics. we also wish to (l)minimize
variability due to uncon-
In our example of chlorophyll concentrations, we
may want to sample each of several lakes in the
deepest part, with a vertically integrated pump
sample taken in early spring before stratification
appears. Many lakes are sampled in order to
examine and characterize the variability due to
different lakes (the sampling unit). Each lake is
sampled in the same way, in the same place,
and in the same time frame in an attempt to
minimize variability due to location, depth, and
season, which are not of interest in this particu-
lar study.
In the above example, chlorophyll concentra-
tions vary with location within a lake, among
9-4
-------
Quality Assurance: Design, Precision ansf Management
lakes, and time of sampling (day, season, year).
If the spatial and temporal components of
variability within lakes are large, then it is best
to use either an index period sample or to
estimate a composite from several determina-
tions. For example, measurements of chloro-
phyll concentrations typically vary more be-
tween spring and fall samples within a lake than
they do between lakes. Therefore, lake chloro-
phyll concentrations are often estimated as a
growing season average, taken from several
determinations (for instance, monthly) during
the growing season.
In analyses, especially hypothesis testing,
multiple determinations within lakes may be a
form of pseudoreplication (Hurlbert 1984), and
should be used with caution. If the hypothesis
refers to a single lake (e.g., chlorophyll concen-
tration of lake Z is higher than a biocriterion),
multiple determinations are often necessary for
the test. If the hypothesis refers to many lakes
(e.g., lakes in state Y have elevated chlorophyll
compared to state Q), multiple determinations
within lakes are pseudoreplication if they are
used as independent observations in the test,
rendering the test invalid (Hurlbert 1984). If
multiple determinations for each lake are used
to calculate a single seasonal mean or median,
which is then used as an independent observa-
tion for the hypothesis test, there is no
pseudoreplication. Repeated measurement
designs—analysis of variance (ANOVA-ANalysis
Of VAriance) or regression—can be used (e.g.,
Underwood 1994) as a single analysis that takes
into account multiple determinations. These
methods estimate means of repeated measures
to maintain independence.
A less costly alternative to multiple measures in
space is to use spatially composite determina-
tions. In nutrient or chlorophyll determinations,
a water column pumped sample, where the
pump hose is lowered through the water col-
umn, is an example of a spatially composite
determination, Benthic macroinvertebrates are
often sampled with spatial composite determina-
tions. For example, benthic macroinvertebrates
in Atlantic Coastal Plain streams are typically
sampled by 20 sweeps of a dip net in multiple
habitats, and composited into a single sample
(e.g. USEPA 1997b, Barbour et al. 1996a,
Barbour et al. 1996b, Roth et al. 1997), Benthic
sampling of Florida lakes is a composite of 12
Petite Ponar grabs made throughout the sublit-
toral zone of a lake or a sample unit (Gerritsen
and White 1997) (see Florida case study in this
chapter).
Multiple observations within a sample unit (e.g.,
within a lake) should not be considered inde-
pendent observations unless they are taken to
examine an explanatory variable of interest,
such as effects of depth, lake zone, season, or
year. The principal use of multiple measure-
ments is to estimate measurement error, that is,
the variability we should expect when a single
determination is made in a lake.
Analysis of variance is used to estimate mea-
surement error. All multiple observations of a
variable are used (from all lakes with multiple
observations), and lakes are the primary effect
variable. The root mean square error (RMSE) of
the ANOVA is the estimated standard deviation
of repeated observations within lakes. A hy-
pothesis test (F-test) is not of interest in this
application because it tests the trivial hypoth-
esis that lakes are different from one another.
Measurement error is the result of methodologi-
cal biases and errors: gear bias; improper use of
gear or improper training; variability in use of
gear; laboratory errors
(chemical analysis
errors); and natural
variability that is not of minimized with methodologi-
interest and is not
being sampled. Mea- cat standardization; selection
surement error is ,
minimized with meth- of cost-effective, low-
odological standardiza-
tion: selection of cost-
effective, low variability
sampling methods;
proper training of
personnel; and quality
assurance procedures
designed to minimize
methodological errors.
Measurement error is
variability sampling methods;
proper training of personnel;
and quality assurance
procedures designed to
minimize methodological
Natural variability that is not of interest for the
questions being asked, but may affect ability to
address these questions, should be estimated
with the RMSE method above. If the variance
estimated from RMSE is unacceptably large (i.e.,
as larges or larger than variance expected
among sample units), then it is often necessary
to alter the sampling protocol, usually by
increasing sampling effort in some way, to
further reduce the measurement error. Mea-
9-5
-------
Chapter 9
surement error can be reduced by multiple
observations at each sample unit, e.g.: multiple
Ponar casts at each sampling event, multiple
observations in time during a growing season or
index period, depth-integrated samples, or
spatially integrated samples.
Sampling design Is the
selection of a part of a
population in order to observe
the attributes of interest, so
that the values of those
attributes can be estimated for
the whole population.
Spatial integration of
sample material and
compositing the
material into a single
sample is almost
always more cost-
effective than retaining
separate, multiple
observations. This is
especially true for
relatively costly labora-
tory analyses such as
organic contaminants
and benthic
macroinvertebrates. The Florida invertebrate and
TVA fish methodologies include the compositing
of multiple sampler casts into a single sample,
which is then counted and identified.
For quality assurance, some effort will always be
required for repeated samples so that measure-
ment error can always be estimated from a
subset of sites. Repeated measurement at 10%
or more of sites is common among many moni-
toring programs, and is recommended.
9.2.2 Alternative Sampling
Designs
Sampling design is the selection of a part of a
population in order to observe the attributes of
interest, so that the values of those attributes
can be estimated for the whole population.
Classical sampling design makes assumptions
about the variables of interest; in particular, it
assumes that the values are fixed (but un-
known) for each mem-
The most basic probability- ber of the population,
until that member is
based design is simple observed (Thompson
1992). This assumption
is perfectly reasonable
possible sample units in the for some variables, say,
length, weight, and sex
population have the same of members of an
animal population, but
probability of being selected. it seems less reasonable
random sampling, where all
for more dynamic variables such as nutrient
concentrations, loadings, or chlorophyll concen-
trations of lakes. Designs that assume that the
observed variables are themselves random
variables are model-based designs, where prior
knowledge or assumptions (a model) are used to
select sample units.
9.2.3 Probability-based
Designs (Random
Sampling)
The most basic probability-based design is
simple random sampling, where all possible
sample units in the population have the same
probability of being selected, that is, all possible
combinations of n sample units have the equal
probability of selection from among the N units
in the population. If the population N is finite
and not excessively large, a list can be made of
the N units, and a sample of n units is randomly
selected from the list. This is termed list frame
sampling. If the population is very large or
infinite (such as locations in a lake), one can
select a set of n random (x,y) coordinates for the
sample.
All sample combinations are equally likely in
simple random sampling. There is no assurance
that the sample actually selected will be repre-
sentative of the population. Other unbiased
sampling designs that attempt to acquire a more
representative sample include stratified, system-
atic, multistage, and adaptive designs. In
stratified sampling, the population is subdivided
or partitioned into strata, and each stratum is
sampled separately. Typically, partitioning is
done so as to make each stratum more homoge-
neous than the overall population. For example,
lakes could be stratified by ecoregion. System-
atic sampling is the methodical selection of
every fcth unit of the population from one or
more randomly selected starting units, and
ensures that samples are not clumped in one
region of the sample space. Multistage sampling
requires selection of a sample of primary units,
such as fields or hydrologic units, and then
selection of secondary sample units such as
plots or lakes within each primary unit in the
first stage sample.
Estimation of statistical parameters requires
weighting of the data with inclusion probabilities
(the probability that a given unit of the popula-
9-6
-------
Quality Assurance: Design, Precision and Management
tion will be in the sample) specified by the
sampling design. In simple random sampling,
inclusion probabilities are by definition equal,
and no corrections are necessary. Stratified
sampling requires weighting by the inclusion
probabilities of each stratum. Unbiased estima-
tors have been developed for specific sampling
designs, and can be found in sampling text-
books, such as Thompson (1992).
9.2.4 Model-based Designs
Use of probability-based sampling designs may
miss relationships among variables (models),
especially if there is a regression-type relation-
ship between an explanatory and a response
variable. As an example, elucidation of lake
response to phosphorus loading with the
Vollenweider model (e.g., Dillon and Rigler 1974)
required a range of trophic states from ultraoli-
gotrophic to hypereutrophic. A random sample
of lakes is not likely to capture the entire range
(i.e., there would be a large cluster of me-
sotrophic lakes with few at high or low ends of
the trophic scale), and the random sample may
be biased with respect to the regression model.
In model-based designs, sites are selected based
on prior knowledge of auxiliary variables, such
as estimated phosphorus loading, lake depth,
elevation, etc. Model-based designs may pre-
clude an unbiased estimate of the population
(e.g., regional trophic state), unless the model
can be demonstrated to be robust and predic-
tive. The population value is then predicted from
the model and from prior knowledge of the
auxiliary (predictive) variables.
Identifying and sampling selected least stressed
reference sites to develop an index is an ex-
ample of samples for a model. The model is the
index (e.g., IBI) and the responses of its compo-
nent metrics. Reference sites alone cannot later
be used for unbiased estimation of the biological
status of lakes. Ideally, it may be possible to
specify a design that allows both unbiased
estimation of a population and index or model
development. Statisticians should be consulted
in developing the sample design for a biocriteria
and biological monitoring program. However,
managers should be aware that there is strong
disagreement among statistical schools of
thought on the subject of sampling design.
biological monitoring program;
disagreement among statistical
9.3 EVALUATION OF
STATISTICAL POWER
A principal aspect of probability sampling is
determining how many samples will be required
to achieve the monitor-
ing goals and what is Statisticians should be con-
the probability of
making an incorrect suited in developing the sample
decision based on the , .
monitoring results. desi9n for a b'ocntena and
The primary tool for
conducting these
analyses is statistical however, managers should be
power analysis.
Evaluating statistical aware that there is strong
power is key to devel-
oping data quality
criteria and perfor- schools of thought on the
mance specifications
for decision making subject of sampling design.
(USEPA 1996c) as well
as evaluating the performance of existing
monitoring programs (USEPA 1992d). Power
analysis provides an evaluation of the ability to
detect statistically significant differences in a
measured monitoring variable. The importance
of this analysis can be seen by examining the
possible outcomes of a statistical test. The null
hypothesis (HJ is the root of hypothesis testing.
Traditionally, null hypotheses are statements of
no change, no effect, or no difference. For
example, the mean abundance at a test site is
equal to the mean abundance of the reference
sites. The alternative hypothesis (Ha) is counter
to Ho, traditionally being statements of change,
effect, or difference. Upon rejecting Ho, Ha would
be accepted.
The two types of deci-
sion errors that could be
made in hypothesis
testing are depicted in
Table 9-2. A Type I
error (i.e., false positive)
occurs when H is
o
rejected although Ho is
really true. A Type II
error (i.e., false negative)
occurs when H„ is
accepted although Ho is
really false. The magni-
tude of a Type I error is represented by a and
the magnitude of a Type II error is represented
by $. Decision errors are the result of measure-
Evaluating statistical power is
key to developing data quality
criteria and performance
specifications for decision
making as well as evaluating
the performance of existing
monitoring programs.
9-7
-------
Chapter 9
Table 9-2. Errors in hypothesis testing.
Decision
Stale of affairs In the population
H„ Is True
H0 is False
Accept H0
1-a
(Confidence level)
P
(Type II error)
Reject H0
a
(Significance level)
(Type I error)
1-p
(Power)
merit and sampling design errors that were
described in Section 9.2.1. A proper balance
between sampling and measurement errors
should be maintained because accuracy limits
effective sample size and vice versa (Blalock,
1979).
9.3.1 Comparison of
Significance Level and
Power
Regardless of the statistical test chosen for
analyzing the data, the analyst must select the
significance level of the test. That is, the analyst
must determine what error level is acceptable.
The probability of making a Type I error is equal
to the significance level (a) of the test and is
selected by the data analyst. In many cases,
managers or analysts define 1-a to be in the range
of 0.90 to 0.99 (e.g., a confidence level of 90 to
99 percent), although there have been environ-
mental applications where 1-a has been set to
0.80. Selecting a 95 percent
confidence level implies
that the analyst will reject
positive) occurs when H0 ^ H0 when Hois really
true (i.e., a false positive) 5
is rejected although H0 is percent of the time.
A Type I error (i.e., false
really true. A Type II error
(I.e., false negative)
occurs when H0 Is
accepted although Hg is
really false.
Type II error depends on
the significance level,
sample size, number of
replicates, variability, and
which alternative hypoth-
esis is true. The power of a
test (1-J3) is defined as the
probability of correctly
———— rejecyng when Ho is
false. In general, for a fixed sample size, a and 3
vary inversely. Power can be Increased 03 can be
reduced) by increasing the sample size or
number of replicates. Figure 9-1 illustrates this
relationship. Suppose the interest is in testing
whether there is a significant difference between
the means from two independent random
samples. As the difference in the two sample
means increases (as indicated on the x-axis), the
probability of rejecting H0, the power, increases.
If the real difference between the two sample
means is zero, the probability of rejecting Ho is
equal to the significance level, a. Figure 1A
shows the general relationship between a and £>
if a is changed. Figure IB shows the relation-
ship between a and fi if the sample size is
increased. The tradition of 95% confidence (a =
0.05) is entirely arbitrary; there is no scientific
requirement that confidence be set at 95%.
Indeed, for environmental protection, power is at
least as important—and possibly more impor-
tant—than confidence (Peterman 1990,
Fairweather 1991).
sic Assumptions
Usually, several assumptions regarding data
distribution and variability must be made to
determine the sample size. Applying any of the
equations described in this chapter is difficult
when no historical data set exists to quantify
initial estimates of proportions, standard
deviations, means, or coefficients of variation.
To estimate these parameters, Cochran (1963)
recommends four sources:
* Existing Information on the same population
or a similar population.
• A two-step sample. Use the first-step sam-
pling results to estimate the needed factors,
for best design, of the second step. Use data
from both steps to estimate the final preci-
sion of the characteristic(s) sampled.
9-8
-------
Quality Assurance: Design, Precision and Management
• A "pilot study" on a "convenient" or "mean-
ingful" subsample. Use the results to
estimate the needed factors. Here the
results of the pilot study generally cannot be
used in the calculation of the final precision
because often the pilot sample is not repre-
sentative of the entire population to be
sampled.
• Informed judgment, or an educated guess.
For evaluating existing programs, proportions,
standard deviations, means, etc. would be
estimated from actual data.
Some assumptions might result in sample size
estimates that are too high or too low. Depend-
ing on the sampling cost and cost for not
sampling enough data, it must be decided
whether to make conservative or "best-value"
assumptions. Because of the fixed mobilization
costs, it is probably cheaper to collect a few
extra samples the first time than to realize later
that additional data are needed. In most cases,
the analyst should probably consider evaluating
a range of assumptions regarding the impact of
sample size and overall program cost. USEPA
recommends that if the analyst lacks a back-
ground in statistics, he/she should consult with
a trained statistician to be certain that the
approach, design, and assumptions are appro-
priate to the task at hand.
9.3.3 Simple Comparison of
Proportions and Means
from Ttafo Samples
The proportion (e.g., percent dominant taxon)
or mean (e.g., mean number of EPT taxa) of two
data sets data sets can be compared with a
number of statistical tests including the para-
metric two-sample t-test, the nonparametric
Mann-Whitney test, and two-sample test for
proportions (USEPA 1996c). In this case, two
independent random samples are taken and a
hypothesis test is used to determine whether
there has been a significant change. To compute
sample sizes for comparing two proportions, p,
and p2, it is necessary to provide a best estimate
for Pj and p2, as well as specifying the signifi-
cance level and power (1-jJ). Recall that power is
equal to the probability of rejecting HB when Ho
is false. Given this information, the analyst
substitutes these values into the following
equation (Snedecor and Cochran, 1980)
- ,2(P1+P2q2)
(za +z2b) ~ ~2~
(P2 "Pi)
Equation 1.
where Z and Z,^ correspond to the normal
deviate. Common values of (Z + Z^f are sum-
marized in Table 9-3. To account for p, and p2
being estimated, t could
be substituted for Z. In
lieu of an iterative
EPA recommends that if the
calculation, Snedecor
and Cochran (1980)
propose the following
approach: (1) compute
n0 using Equation 1; (2)
round n0 up to the next
highest integer, /; and (3)
multiply no by (f+3)/(f+l)
to derive the final
estimate of n.
To compare the mean
from two random
analyst lacks a background
In statistics, he/she should
consult with a trained
statistician to be certain that
the approach, design, and
assumptions are appropriate
to the task at hand.
Power (1-0)
0.0 increasing Difference Between
the Mean of Two Random Samples
A) Increasing Significance Level from a, to a.
i>2> tii
0.0
Increasing Difference Between
the Mean of Two Random Samples
B) Increasing Sample Size from n, to n
Figure 9-1. Illustration of significance (a) and
power (1-p).
9-9
-------
Chapters
Table 9-3. Common values of (Zo + Z2fl)2 for estimating sample size for use with Equations 1 and 2
(Snedecorand Cochran 1980).
Power,
1-/7
crfor One-sided Test
a for Two-sided Test
0.01
0.05
0.10
0.01
0.05
0.10
0.80
10.04
6.18
4.51
11.68
7.85
6.18
0.85
11.31
7.19
5.37
13.05
8.98
7.19
0.90
13.02
8.56
6.57
14.88
10.51
8.56
0.95
15.77
10.82
8.56
17.81
12.99
10.82
0.99
21.65
15.77
13.02
24.03
18.37
15.77
samples to detect a change of (i.e., x^-x^), the
following equation is used:
n,
Equation 2.
: (^o ^2p)
(Sj2 + sa2)
For large sample sizes or
samples that are normally
distributed, symmetric
confidence intervals for
the mean are appropriate.
n0 (^cc ^2b)
(Si )
52
Equation 3,
In Equation 3, Z is most often one-tailed,
because the concern is only whether the value is
below the threshold.
9.3.4 Sample Size
Calculations for
and Proportions
Means
where s, and s2 are standard deviation of
samples 1 and 2.
Common values of [Z + Z^)2 are summarized in
Table 9-3. To account for s, and s2 being
estimated, Z should be replaced with £ In lieu
of an iterative calculation, Snedecor and
Cochran (1980) propose the following approach:
(1) compute n0 using Equation 2; (2) round n0 up
to the next highest
integer,/; and (3) multiply
nB by (f+3}/(f+l) to derive
the final estimate of n.
For large sample sizes or samples that are
normally distributed, symmetric confidence
intervals for the mean are appropriate. This is
because the distribution of the sample mean will
approach a normal distribution even if the data
from which the mean is estimated are not
normally distributed. The Student's t statistic
(t/2 n j) is used to compute symmetric confidence
intervals for the population mean, ji:
x-ta/2,n.1Vsi7n< H
-------
Quality Assurance: Design, Precision and Management
Example Sample Size Calculations for Comparing Proportions and Population Means
Example 1—Sample size calculation for compar-
ing proportions
To detect a difference in proportions of 0.20 with a
two-sided test, a equal to 0.05, X- B equal to 0.90,
and an estimate of p, and p2 equal to 0.4 and 0.6, n0
is computed from Equation 1 as
Jm -4X0.6) ^
Da "-"10 *5 'A : "1 I Zu • I
(0.6-b.4)2 ;.x-
Rounding 126.1 to the next highest integer, ! is equal
to 127, and n is computed as 126.1 x 130/128 or
128.1, Therefore 129 samples in each random
sample, or258 total samples, are needed to dStect a
difference in proportions of 0,2. Since these are pro-
portions, the result means that the total count in the
sample must be at least 129, For example, to detect
the above difference in the proportion of dominant
taxon (e.g., benthicmacroirivertebratesorfish) of two
lakes, at least '129 individuals must be counted and
identified in each lake.
The example illustrated that a statistically significant
difference can be easily detected in proportions if
sufficient Individuals are'sampled: However, it is
doubtful that a difference between 40% and 60% in
dominant taxori is biologically meaningful.
Example 2
ingpopulationmeans
To detect a difference of 20 in mean abundance with
a twfhsidad teslL The standard deviation; s, was es-
timated as 30 for both samples based on previous
studies; was selected as 0.05; and 1-0 was selected
as 0.90. Substituting these values into Equation 2
yields
(30
-302)
20
= 47,3
Rounding 47.3 to the next highest integer, f is equal
to 48, and n is computed as 47,3 x 51/49 or 49,2.
Therefore 50 samples in each random sample, or100
total samples, are needed to detect a difference of 20,
rely on these assumptions. Conover (1980) also
provides a procedure for smaller sample sizes.
To calculate the confidence interval correspond-
ing to the median, lower quartile, or upper
quartile, the following procedure is used,
1. Order the data from smallest to largest
observation such that
X, < ••• < Xr :
-------
Chapter 9
Case Study; Optimization of Benthic Sampling in Florida Lakes
To optimize its lake sampling protocols, Florida DEP
performed a pilot study on 9 lakes. Each lake was
sampled with twelve petite Ponar grabs (0.02 rrf) dis-
tributed approximately equidistant in the sublittoral
zone of the lake (2-4 m depth; Fig. 7-6). Each grab
was kept separate In laboratory Identification and enu-
meration. The lakes spanned a wide range in benthic
r macrglnvertebrate diversity and abundance (Table 9-
4). from 7 to 63 taxa, and228 to 3540 organisms in
0 24 m1 sampled. In seven of the nine lakes, the num-
ber of taxa continued to increase with sampling ef-
fort, and did not reach an asymptote with twelve Ponar
samples.
To illustrate the effects of compositing sample casts,
each sample of 12 grabs was composited into 2 repli-
cate samples of 6 casts, so that each sample consisted
of alternate casts (Fig. 7-6). This yielded 2 alternative
sampling protocols: 12 Ponar replicates for each lake,
andtworepScates of 6 Ponars each. 4 candidate metrics
- were calculated: number of taxa (cumulative for
composited samples), percent dominance, sensitive
taxa (ephemeroptera, trichoptera, odonata), and log
abundance. Standard deviation ofeach metric, as mea-
surement error In determining the "true' value for each
-lake, was estimated with the foot mean square error
(RMSE) from an analysis of variance (Table 9-5).
All metrics had a tower coefficient of variation (CV) in
the composited protocol than in the uncomposited,
showing the advantages of compositing multiple de-
ployments of small sample gear such as Ponars.
Composited samples reduce costs because fewer jars
and records am required, and sampling time is re-
duced some. Laboratory analysis can be reduced
by subsampling a fixed number of organisms (e.g.,
100,200, or300) from the composite sample for iden-
tification. It has been shown with the same Florida
data (Barbour and
volume) (Hurlbert 1971). Subsamples that are larger
than the target number can be reduced
computationally by rarefaction (Hurlbert 1971, Vinson
and Hawkins 1996, Barbour and Gerritsen 1996).
Based on these results, Florida DEP adopted the fol-
lowing sampling protocol for lake benthic inverte-
brates:
• 12 Ponars randomly deployed in 12 segments of
the2-4m depth zone of lakes less than 1000acres.
• Ponar casts are composited into a single sample
and sieved through a 500 m mesh screen.
* A subsample of 100 benthic macroinvertebrates
is sorted and identified to the lowest practical
taxonomic level.
* The sampling protocol is duplicated at approxi-
mately 10% of sites to estimate measurement
error.
Table 8*4. Number of taxa and Individuals In 12
Laka
Cumulative taxa
Cumulative
Individuals
Overstreet
63
768
Post
54
454
Camel
54
3540
Logan
42
1648
Mic
34
2828
Ocho
31
1848
Dei
18
228
Pickett
9
370
Adams
7
485
Gerritsen 1996)
that subsampling
5 :a fixed number of
organisms (100
or more) yields
adequate esti-
mates of number
of taxa, which am
actually more pre-
cise than taxa
density (total taxa
In a fixed area or
Table 9-5. Comparison of two sample processing protocols, Florida lakes.
mean of 12 Ponars
mean of 2 samples of 6 composited Ponars
Population
mean
(Slakes)
Range
(9 lakes)
B.d.
{Individual
laka)
CV
(average
lake)
Population
mean
(Stakes)
Range
(9 lakes)
S.&
(Individual
lake)
CV
(average
lake)
No. of taxa
8.85
2-19
3.62
40.9%
25.7
5.6-44.5
4.30
16,9%
%
dominance
58.8%
40%-96%
14.8%
25.2%
50.4%
16%-96%
8.9%
17.7%
Sensitive
taxa (ETO)
0.39
0-1.7
0.628
161%
1.6
0-55
1.27
79.4%
Total iodlv
On)
4.13
2.78-5,60
0.717
17.4%
8,12
4.68-7.48
0.145
2.4%
9-12
-------
Quality Assurance: Design, Precision and Management
Case Study: Estimation of Power for TVA Fish Samples
TV A samples reservoir fish, benthic macro-,
invertebrates, water column chlorophyll, dissolved oxy-
gen, and sediment contamination to rate the overall
health of its reservoirs. 5 indices am calculated, one for
each indicator group, Measurements are duplicated at
selected reservoirs to obtain estimates of Variability *
in 1996, fish sampling was. repeated at seven reser-
voirs. The TVA Reservoir Fish Assembly Index (RFAI)
is composed of 12 metrics (see Chapter s). Ranges
of metric values in 1996 (for all reservoirs) and metric
standard deviations (from multiple determinations at
single reservoirs are given in Table 9S.
From the standard deviation of the RFAI score, we
can estirnate the number of samples required to de-
tect differences among lakes.
1. Difference between two lakes (or between two sam-
pling times within a lake) ,
To detect st difference of 10 in mean RFAI score with
a two-sided test The standard deviation, s; was es-
timated as 4.027forboth samples (Table9-6); awas
selected as 0.05; and 1 *8 was selected as 0.80; Sub-
stitutingthese values into Equation 2 yields
2. Test whether a lake is below a threshold
(biocriteria): - c' .-
To detect a difference of 10 in mean RFAl score^be-
low a threshold. The same standard deviation esti-
mate is used as above (4.027; Table 9-6); a and 1-8
were selected as. 0.05 and 0.80, respectively, but
is now one-sideB. Substituting thesevalues into-
Equation2yields:, '*
V'-"" .4 0272 :
6.18—i— = 1.002 : :
n„
10"
n„ = 7.as
(4.027 +4.027
' 102
- = 2.54
: Rounding 2.54 to the next highest integer, f is equal
to 3; and n is computed as 2.54 x 6/4 or3 82. There-
fore, 4 samples in each reservoir, or 8 total samples,
arp needed to detect a difference of 10 in RFAI score
between two reservoirs, witit a probability of 0.80 of
finding a true difference.
Rounding 1.002to the next highest integer, f is equal
to 2, arid n is computed as 1.002 x 5/3 or 1.67.
Therefore, 2 samples are needed id* detect a differ-
ence of 10 ih RFAI scorn below.a threshold, with a
probability of 0.80 of finding a true difference. If the
effect size, or distance below the threshold, were
increased to 15, then the required sample size would
beCf. Thus if we find §n RFAI value from a single
unreplicated sample to be f 5points below a thresh-,:
old, then we would expect that replication would not
Cchange a conclusion thai the^reservoir RFAt is be -
lowihe threshold, 95% of the time.' This example
shows the potential value of adaptive sampling strat-
egies; where a decision to increase sampling effort
is, based on the vafue of the first replicate, if the
index value is very far below a threshold, there is no?
need to replicate.;'-:Asi the index value approaches
the threshold, sampling,effort needs p increase in:
order lo make a decision at the prescribed power
andsignificance: At some point, the sampling effort
becomes so costly that judgement is reserved; i.e.,
no decision is made.
Table 9-6. Minimum and maximum values, and standard deviations of repeated
measures, of reservoirfish metrics anel the RFAI.
Metric
Minimum
{all reservoirs,
1990-96)
Maximum {all)
s of repeated
measures (n=7)
Total taxa
12 ¦-...'V:' -
47
1.389
Piscivore species
1.309 '
Sunfish spades"
.V • '
0.756 • .
Sucker species
- - - ' y -1V " ^
0.463
Intolerant soecies
0.463
Percent tolerant
0.118
Percent dominant ¦
¦ -
i' - " . '. * " \ I
0.122
Percent omnivores
J-:-'
0.1.18 :
Percent inse
-------
Chapter 9
9.4 MANAGEMENT
9«4.1 Personnel
Trained and experienced biologists should be
available to provide thorough evaluations,
provide support for various activities, and serve
as QC checks. They should have training and
experience commensurate
with the needs of the
Protocols should ha
developed for designing a
data base and for screen-
ing, archiving, and
documenting data.
program. At least one staff
member should be familiar
with establishing a QA
framework. QA programs
should document person-
nel responsibilities and
duties and clearly delineate
project organization and
lines of communication
(USEPA 1995). A time line illustrating comple-
tion dates for major project milestones or other
tasks can be a tremendously useful tool to track
project organization and progress.
9.4.2 Resources
Laboratory facilities, adequate field equipment,
supplies, and services should be in place and
operationally consistent
with the designed pur-
poses of the program so
aspect of an ecological that high-quality environ-
mental data can be
study the major QC generated and processed
. , , in an efficient and cost-
eements are instrument „ ,T j,.™.
effective manner (USEPA
calibration and mainte- 1992b). Adequate taxo-
nomic references and
nance, crew training and scientific literature should
be available to support
evaluation, field equipment, laboratory work, data
processing, and interpre-
tation.
For the field operations
sample handling, and
additional effort checks.
9.5 OPERATIONAL QUALITY
CONTROL
Protocols should be developed for designing a
data base and for screening, archiving, and
documenting data. Data screening identifies
Six qualitative and quantitative data characteris-
tics usually employed to describe data quality:
1 Precision—The level of agreement among
repeated measurements of the same char-
acteristic.
2. Accuracy—The level of agreement between
the true and the measured value, where the
divergence between the two Is referred to as
bias.
3. Representativeness—The degree to which
the collected data accurately reflect the true
system or population.
4. Completeness—The amount of data collected
compared to the amount expected under ideal
conditions.
5. Comparability—The degree to which data
from one source cm be compared to other,
similar sources.
6. Measurability—The degree to which mea-
sured data exceed the detection limits of the
analytical methodologies employed; often a
function of the sensitivity of instrumentation.
questionable data based on expected values and
obvious outliers. Screening is especially impor-
tant if data are gathered from a variety of
sources and the original investigators and data
sheets are no longer available. The following
text box defines the qualitative and quantitative
data characteristics that are most often used to
describe data quality.
These measurement quality indicators require a
priori consideration and definition before the
data collection begins. Taken collectively, they
provide a summary characterization of the data
quality needed for a particular environmental
decision. Duplication of approximately 10
percent of the total sampling effort is a common
level for operational QC. Replication of samples
at a randomly selected subset of field sites
(usually, 10 % of the total number is considered
appropriate) is used to estimate precision, and
representativeness of the samples and the
methods; splitting samples into subsamples can
be used to check precision of the methodology,
and reprocessing of finished samples is used to
check accuracy of laboratory operations.
9-14
-------
Quality Assurance: Design, Precision and Management
9.5.1 Field Operations
For the field operations aspect of an ecological
study, the major QC elements 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
deficiencies to equipment problems. Field notes
are integral to the documentation of activities
and can be used to help locate potential record-
ing errors. Training is one of the most impor-
tant QC elements for field operations. Estab-
lishment and maintenance of a voucher speci-
men collection should be considered for biologi-
cal data. Transcription errors during data entry
can be reduced with double data entry. Table 9-
7 gives examples of QC elements for field and
laboratory activities.
9.5.2 Laboratory Operations
The QC elements in laboratoiy operations
include sorting and verification, taxonomy,
duplicate processing, archival procedures,
training, and data handling. Potential error
sources associated with sample processing are
best controlled by staff training. 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 can be
controlled by training and by duplicate process-
ing, sorting, and verification procedures. See
Table 9-7 for examples for QC elements for
laboratory activities,
9.5.3 Data Analysis
Errors can occur if inappropriate statistics are
used to analyze the data. Undetected errors in
the data base or programming can be disastrous
to interpretation. Problems in managing the
data base can occur if steps are not taken to
oversee the data handling, analysis, and sum-
marization. The use of standardized computer
software for data base management and data
analysis can minimize errors associated with
tabulation and statistical analysis. A final
consideration is the possible misinterpretation
of the findings. These potential errors are best
controlled by qualified staff and adequate
training.
9.5.4 Reporting
QC in reporting includes 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 sound scientific document.
9-15
-------
Chapters
Table 9-7. Example QC elements for field and laboratory activities
Project
Activity
QC Element
Evaluation Mechanism
Field Sampling
Replicated samples at 10 percent of
sites by same field crew.
Calculate relative percent difference
(RPD) of Index value or individual metric
score
Replicated samples at one to two of total
sites by different field crew using same
methods.
Calculate RPDs as above; use to
evaluate consistency and bias.
Physical
Habitat
Assessment
(Qualitative)
Ensure appropriate training and
experience of operators; multiple
observers.
Resume or other documentation of
experience; discuss and resolve
differences in interpretation.
Physical
Habitat
Assessment
(Quantitative)
Replicated measurements at 10 percent
of sites.
Calculate RPDs between replicate
measurements; compare to
preestablished precision objectives.
Laboratory:
Sample
Sorting
Sample residue checked for missed
specimens to estimate sorting efficiency;
check completed by separate lab staff.
Calculate percent recovery; compare to
preestablished goals.
Laboratory:
Sample
Tracking
Logbook with record of all sample
Information.
Not applicable.
Laboratory:
Taxonomic
Identification
Independent identification and/or
verification by specialist; ensure
appropriate and current taxonomic
literature available; adequate training
and experience in invertebrate
identifications; reference collection;
exchange selected samples/specimens
between taxonomists.
Calculate percent error; compare to
preestablished goals.
Data
Management
Proofreading; accuracy of transcription.
All transcribed data entries compared
by hand to previous form—handwritten
raw data, previously computer-
generated tables, or data reports.
Data Analysis
Hand-check of reduced data.
For computer-assisted data reduction,
approximately 10 percent of reduced
data recalculated by hand from raw data
to ensure integrity of computer
algorithm.
Appropriate statistics; training.
Review by statistician or personnel with
statistical training.
9-16
-------
in This Chapter...
> Effective Biocriteria
> Cost and Design Trade-Offs
> Cooperation for Cost-Effective Programs
Chapter 10
Biocriteria lniDlementation
Hi ^WH^ Hi Hi VI HI B HHIi^H B Hi ¦ H Hi ^H(^ ™ Hi H ^¦¦li^ H M HH Hi ^HH^ Bl M
1 0.1 CHARACTERISTICS OF
EFFECTIVE BIOCRITERIA
Development of narrative or numeric biocriteria
depends on the premise that biota provide a
sensitive screening tool for measuring the
condition of a water resource. Properly defined
biocriteria can be used to protect the biological
integrity of waterbodies and establish aquatic
life use classifications.
Following the development of biocriteria, sites
are evaluated to determine how well they meet
the biocriteria or whether they have been
significantly degraded. This determination is
made by comparing the aquatic biota at poten-
tially disturbed sites to the biocriteria, which are
in turn based on minimally impaired reference
conditions. The greater the anthropogenic
impact in a watershed, the greater the impair-
ment of the water resource. A corollary is that
drainage basins not subject to anthropogenic
impacts 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 that
result from human disturbances.
The establishment of formal biocriteria warrants
careful consideration of planning, management,
and regulatoiy goals. Effective biocriteria
function to:
Provide for scientifically Properly defined
sound evaluations.
biocriteria can be used
Protect the most sensitive
biological value. to Protect the biological
Support and strive for integrity of waterbodies
protection of chemical and establish tic
physical, and biological
integrity. //fe use classifications.
Generally, optimal biocriteria
share several common characteristics;
• They include specific assemblage charac-
teristics required for attainment of desig-
nated use.
• They are clearly written and easily under-
stood.
• They adhere to the philosophy and policy
of antidegradation of water resource
quality.
• They are defensible in a court of law.
In addition, biocriteria should be written to
consider the best attainable condition at a site.
10-1
-------
Chapter 10
The best balance is
represent the natural
biota, protect against
degraded sites.
Overly stringent criteria that are unlikely to be
achieved serve little purpose. Similarly,
bloeriteria that support a degraded biological
condition defeat the intent of the Clean Water
Act. Well-designed biocriteria are set at levels
sensitive to anthropogenic impacts; they are
not set so high that sites
that have reached their
full potential are consid-
achiaved by developing ered in nonattainment
or so low that unaccept-
blocriteria that closely ably impaired sites are
scored as meeting the
criteria. It will be diffi-
cult to determine the full
potential of a given lake.
further degradation, and Balanced biocriteria will
allow multiple uses to be
stimulate restoration of considered 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.
Several kinds of biocriteria are possible, and
both narrative and numeric biocriteria have
been effectively implemented. Narrative
biocriteria consist of statements such as
"aquatic life as it naturally occurs" or "changes
in species composition may occur, but struc-
ture and function of the aquatic community
must be maintained." Numeric values, such as
measurements of community structure and
function, can also serve as biocriteria as such
or as quantitative refinements of narrative
biocriteria. To account for a measure's natural
variability in a healthy environment, the
numeric criterion should be a defined range
rather than a single number. Numeric criteria
may also combine several such values in an
index. Regardless of which kind is chosen,
biocriteria should be both quantitatively based
and supported by effective implementation
guidelines and adequate capabilities including
people, resources, methods, historical data,
and management support. Additional general
guidance regarding the writing of biocriteria is
provided in EPA 440/5-90-004 (1990a) and
EPA 822-B-92-002 (1992e).
10.2 STEPS TO
IMPLEMENTATION
r
The first phase in a biocriteria program is the
development of narrative biological criteria
(USEPA 1992e). These criteria are essentially
statements incorporated into water laws and
regulations to formally consider the fate and
status of aquatic biological communities. These
statements of intent should include the following
objectives;
1. Support the goals of the Clean Water Act to
provide for the protection 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 commu-
nity possible by emphasizing the protection
of its most sensitive components.
3. Refer to specific community characteristics
that must be present for the waterbody to
meet a particular designated use; for
example, natural diverse systems with their
respective communities or taxa indicated.
4. Include measures of community characteris-
tics, based on sound scientific principles,
that are quantifiable and written to protect
or enhance the designated use.
5. In no case should impacts degrading
existing uses or the biological integrity of the
waters be authorized.
The use of multiple measures, or metrics, to
develop biocriteria is a systematic process
involving discrete steps. The process includes
site classification (Chapter 4), a biological
survey, evaluation of metrics with aggregation
Into indices (where indicated), formulation of
biocriteria, and monitoring and assessment. The
conceptual model for processing biological data
into a biocriteria framework is adapted from
EPA 822-B-96-001(USEPA 1996a) and summa-
rized in Table 10-1. The process is as follows:
Step 1: Preliminary Classification of the Re-
source—The first decision that a resource
agency must make is to determine the resource
classes to which biocriteria will apply. Success-
ful classification will result in less variation
within a class, leading to more refined charac-
terization of the reference condition and, there-
10-2
-------
Biocriteria Implementation
fore, to criteria with better resolution in detect-
ing impairment. The preliminary classification
should be based on lake characteristics that are
not subject to pollution or disturbance, such as
size, depth, morphology, or characteristics of the
lake watershed.
Multijurisdictional collaboration is encour-
aged so that common methods and metrics
can be established among states or other
monitoring entities, and common reference
conditions for multijurisdictional ecoregions
can be characterized.
A set of reference sites are selected for each
resource class; the reference sites are those
least impacted by human influence, and they
are characteristic of the resource class.
Step 2: Biological Survey—To determine the
discriminatory power of the metrics within a
lake class, the best-quality sites available, as
well as those known to be impaired, are sur-
veyed for biota and physical habitat. The use of
standardized field collection methods allows a
better interpretation of the raw data than does
the use of a conglomeration of techniques.
Step 1
Preliminary Classification to Determine Reference Conditions and Regional
Ecological Expectations
- Resource classification
- Determination of best representative sites (reference sites representative of
class categories)
Step 2
Characterization of Reference Condition
- Historical data
- Survey of reference sites and selected test sites.
- Applicable models
- Expert consensus
StepS
Final Classification
- Test preliminary classification
- Revise if necessary
Step 4
Metric Evaluation and Index Development
- Data analysis (data summaries)
- Testing and validation of metrics by resource class
- Evaluation of metrics for effectiveness in detecting impairment
- Aggregation of metrics into index
- Selection of biological endpoints
StepS
Biocriteria Development
- Adjustment by physical and chemical covariates
- Adjustment by designated aquatic life use
Step 6
Implementation of Monitoring and Assessment Program
- Determination of temporal variability of reference sites
- Identification of problems
Step 7
Protective or Remedial Management Action Initiate
- Programs to preserve exceptional waters
- Implement management practices to restore the biota of degraded waters and
to identify and address the causes of this degradation
StepS
Continual Monitoring and Periodic Review of References and Criteria
- Biological surveys continue to assess efficiency of management efforts
- Evaluate potential changes in reference condition and adjust biocriteria as
management is accomplished
10-3
Table 10-1. Sequential progression of the biocriteria process.
-------
Chapter 10
Step 3: Final Classification—The preliminary
classification Is tested with biological data to
determine whether it Is reflected in the biota. If
necessary, the classification is revised.
Any characterization of a reference condition
should allow for the variability in biological data
by using measures of central tendency and
variability. Statewide or broader characterization
of reference condition can be expected to exhibit
high variance. The goal of classification is to
minimize variability within classes by allowing
the variability to be
attributed to differences
among classes.
Step 4: Metric Evaluation
and. Index Development—
Potential metrics that
have ecological relevance
are identified in this step.
Metrics are then evalu-
ated for the ability to
differentiate between
impaired and
nonimpaired sites. Values
from various scales of
measurement are trans-
formed to scores, which are normally incorpo-
rated into an index, such as an Index of Biologi-
cal Integrity OBI) or an invertebrate index, which
in turn becomes part of the final assessment.
Metrics may also be used individually as indica-
tors of biological condition in the overall assess-
ment.
Step 5: Biocriteria Development—Biocriteria may
be based on an aggregated index, or established
for several biological metrics and adjusted by
aquatic life uses. The component information
and data should always be retained so future
indexes or improvements In initial indexes can
be calibrated with the data, and continuity of
information preserved over time.
For example, a biocriterion for "Class A" lakes
might be "a biotic index greater than the 25th
percentile of least-impacted reference condi-
tions." A "Class A" lake would be rated impaired
if its biotic index fell below the 25th percentile of
reference condition.
Step 6: Implementation of the Monitoring and
Assessment Program—Use of biocriteria requires
an operational monitoring and assessment
program for two primary reasons: assessment of
Biocriteria may be based
on an aggregated index, or
established for several
biological metrics and
adjusted by aquatic life
uses. The component
Information and data
should always be retained.
Outline of Evaluation Criteria for Bloassessment
Programs.
1. Development of quality assurance and qualify
control bloassessment program plans.
2. Careful preparation of data quality objectives
(DQOs) and design of field and laboratory stud-
ies to ensure the collection of representative
data that will enable the biologists to achieve
the objective of their program.
3. Preparation of standard operating procedures
(SOPs) for field and laboratory methods.
4. Staff with adequate training and experience;
division of labor within the program that per-
mits specialization.
5. Use of approved methodology, use of techni-
cally defensible methodology If approved meth-
odology is not available.
A.
Sample collection
B.
Sample processing
C.
Organism identification
D.
Counting
E
Biomass measurements
F.
Data analysis and interpretation
6. Adequate space and physical facilities.
7. Adequate state-of-the-art field equipment, labo-
ratory'instrumentation, and supplies.
8. Adequate safety procedures.
9. Use of replication in sample collection and
analysis to determine the precision.
10. Frequent calibration of field and laboratory in-
struments; log book documentation.
11. Chain-of-custody procedures for proper sample
identification, handling, and logging to prevent
..... misidgrftifjcation anq intermixing of samples.
12. Development and use of a taxonomlc reference
library for identifying specimens to the lowest
possible taxonomlc level.
13. Development and use of a reference specimen
collection and use of outside experts to solve
difficult problems in specimen identification.
14. Careful editing of data before they are placed
in a computer file or used in reports.
15. Use of appropriate statistical analyses and
other methods of data evaluation and interpre-
tation.
10-4
-------
Biocriteria Implementation
potentially impaired test sites arid continued
monitoring of selected reference sites to deter-
mine seasonal and annual variability and
trends, A biocriteria program is the basis for a
representative sampling program to determine
statewide status and trends of the resource.
The resources required to initiate a monitoring
and assessment program are presented in the
text box entitled "Outline of Evaluation Criteria
for Bioassessment Programs."
Step 7: Protective and Remedial Management
Action—The purpose of the entire process is to
improve the water resource quality. Where
problems have been identified through this
effort, land use changes, discharges, abate-
ments, and in-lake use adjustments are part of
the management response. This may be done to
improve degraded lakes or reservoirs or to
protect exceptionally good ones from future
damage. It should be recognized that imple-
menting management action is potentially a
multi-year process.
Step 8: Continual Monitoring and Periodic Re-
views—The biocriteria-biomonitoring effort is
designed to be a continuing process. Progress is
expected but failures must be documented so
monitoring and management efforts can be
improved. The process progressively improves
water resources by cycling back through the
sequence.
10.3 TECHNICAL
CONSIDERATIONS
The technical design of a biocriteria program
affects the program's total cost. The sampling
and analysis effort and data storage are two
major cost elements of a biocriteria program. An
optimal design balances the information needs
of the monitoring agency with the cost of obtain-
ing the information.
10.3.1 Taxonomlc Level
Assemblages in Tiers 2A and 2B are identified to
the lowest practical taxonomic level. Species
level identification can be time consuming,
especially for phytoplankton and benthic
macroinvertebrates. and identification to family
or genus might be more cost-effective.
10.3.2 Subsampllng
Consistency of sampling methods and effort is
critical in bioassessment. A sample is usually
subsampled, in a random manner, to obtain a
reasonable number of organisms for identifica-
tion and enumeration (typically 100 to 500).
Using fewer than 100 organisms might yield
unreliable results, whereas using more than 500
is not cost-effective.
Taxonomic richness metrics, such as total taxa,
diversity indices, and number of orders are
sensitive to sample size. These values increase
asymptotically with subsample size up to 500
organisms. Percent composition metrics (e.g.,
feeding groups, higher tax
metrics) are less sensitive
to subsample size; that is,
the precision of an esti- balances the information
mate for percent composi-
tion does not improve with needs of the monitoring
subsamples greater than
100 organisms.
An optimal design
agency with the cost of
obtaining the information.
In order to control for the
effects of sample size, it is
critical that the methods are consistent in the
number of organisms identified. For example, if
the target subsample is 100 organisms, then
subsamples smaller than 80 organisms should
be rejected and subsamples larger than 120
organisms should be reduced mathematically by
rarefaction (Hurlbert 1971) to make them
comparable.
10.3.3 Spatial Variability and
Replication
Replicating field samples by repeated measure-
ments at a site is integral to biological surveys.
These analyses have typically tested for signifi-
cant differences between upstream and down-
stream pairs of sites. Significant differences
were inferred to be due to discharges. However,
Hurlbert (1984) pointed out that treatment of
multiple measurements as replicates to infer
cause is incorrect use of statistical inference. He
pointed out that the site is the sampling unit,
and repeated measurement of a sampling unit is
not replication. True replication is achieved by
replicating independent sampling units.
Repeated measurements, however, do have
benefits, which must be weighed against the
10-5
-------
Chapter 10
cost. Repeated measurements are used to
estimate measurement error, which is variability
among measurements at the same site. Mea-
surement error is due to spatial and temporal
variability within a lake as well as actual errors
made in sampling and analysis. It may be
necessary to determine whether the measure-
ment methodology adequately characterizes the
site, and to determine the precision of metrics
and indices (Fore et al. 1994). If measurement
error is too large, it may be reduced by repeated
measurements at a site or by a change of
methodology (sample more microhabitats for a
larger composite sample; increase subsample
size). If the measurement error is acceptable, it
is necessary only to take repeated measure-
ments (replicates) for quality assurance at
randomly selected sites (typically 10 percent of
all sites). The QA replicates are then used to
estimate measurement error.
Repeated measurements are used to sample
more microhabitats at each site because the
spatial distribution of organisms at a site can be
patchy and a single measurement might not
represent the composition of the assemblage.
This usually results in a better estimate of the
assemblage at the site. Since the site is the
sampling unit, the mea-
Index period sampling, In surement methodology can
be altered to reduce
which measurements are measurement error. This is
usually done by sampling
made during the same muitipie locations or
habitat types, with several
deployments of the speci-
mldsummer), Is Intended fled sampling gear, and
combining the hauls into a
to control short-term single composite sample.
With a composite sample, a
V""b""y- single measurement is
taken, but the measure-
ment is thought to be more representative of the
site than a single, non-composited sample.
Composite sampling that is representative of the
sites is usually the most cost-effective sampling
methodology. It avoids the costs of multiple
measurements, allowing more sites to be
sampled and increasing statistical sample size.
10.3.4 Temporal Variability
All aquatic assemblages go through annual
cycles of composition and abundance changes.
period each year (e.g.,
In addition, short-lived species also exhibit
short-term temporal variability. Index period
sampling, in which measurements are made
during the same period each year (e.g., mid-
summer), is intended to control short-term
variability. Index period sampling is effective if
assemblage composition and abundance are
relatively stable and predictable among years. If
the assemblage is not stable within the index
period, it might be necessary to make repeated
measurements during a season or year to obtain
growing season or annual average estimates of
the metrics. Repeated measurements over a
season or year are more expensive and reliable
than index-period sampling. For cost-effective-
ness, assemblages that can be adequately
characterized using index period sampling are
therefore preferable to those which require
repeated sampling, unless the information from
the repeated sampling is more valuable.
In view of major seasonal changes in lakes, it Is
possible to have more than one index period.
Warm temperate and subtropical lakes, in
particular, might require two or more index
periods because biological activity remains high
year-round. Multiple index periods must be
analyzed separately. Therefore, there will also be
separate reference expectations and biocriteria
for each season represented by an index period.
Two index periods require double the sampling
effort of a single index period but provide greater
information on biological variability throughout
the year.
1 0.3.5 Classification
Each lake class requires a separate reference
characterization (hence, separate reference sites)
and separate biocriteria. For better statistical
validity, each class should have a minimum of 5
or 10 reference sites (preferably up to 30 sites).
Excessive proliferation of lake classes results in
an unwieldy and expensive biocriteria program.
10.3.6 Status and Trends
Estimating status and trends of lakes as a
resource requires a different sampling design
from that proposed here. Unbiased estimation of
status requires random selection of sampling
units (lakes) within sampling strata (lake
classes). One approach is to assign all lakes to
the classes and then randomly select a sample
10-6
-------
Biocriteria Implementation
of lakes from each class (list-frame sampling).
An alternative approach is to use a grid and
sample lakes nearest the grid points, as is being
done in EMAP (USEPA 1991e).
Trends can be assessed in single lakes or in a
region. Several years of sampling are required
for trend assessment. EPA 841-JR-93-003
(USEPA 1993d) outlines trend analysis methods
for lakes.
specialists from this above pool of talent can
also provide the "expert consensus" referred to
earlier in defining reference conditions and
developing biocriteria.
A cost-effective way to develop a bioassessment
and biocriteria program is to coordinate efforts
and share date with adjacent states or tribes,
especially when lake or reservoir systems cross
political boundaries.
1 O.a PROGRAM RESOURCES
A successful bioassessntent and biocriteria program
depends on (1) a clear definition of goals, (2) the;
active use of biomonitoring data in decision mafa
ing, and (3) the allocation of adequate resources to
ensure a high-quality program.
The implementation of a bioassessment and
biocriteria program requires proper management
and the appropriate combination of resources
and expertise. Agencies already having well-
developed programs usually have experienced
and well-trained biologists, appropriately
equipped facilities, and properly maintained
sampling gear. Areas just beginning a
bioassessment and/or biocriteria program need
to evaluate their existing biological expertise,
facilities, and equipment and expand accord-
ingly. A cost-effective way to accomplish this is
to coordinate efforts and share data with adja-
cent states or tribes, especially when lake or
reservoir systems cross political boundaries.
10.4.1 Program Elements
Monitoring agencies can and should enhance
their programs through cooperation with other
agencies. For example, they should seek coordi-
nation with staff from state fishery, land man-
agement, geology, agriculture, and natural
resource 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 could be initiated. Scientists at
universities should also be included in the
planning and monitoring phases of the pro-
gram—their students make excellent field
assistants and future ecologists and natural
resource managers. The selected team of
10.4.2 Personnel and
Resources
Several trained and experienced biologists and
natural resource specialists should be available
to provide thorough evaluations, support
various activities, and manage quality. They
A ¦biocriteria and biqmohitoring program has sev-
eral required elements, as well as optional ele-
ments, that determine the posts and resources of ;
the program. Program elements include:
• Quality assurance arid quality, control (e.g.,
standard operating procedures, training).
. • Delineated reference conditions with annual
• monitoring of selected sites.' '
• Multiple assemblage biosurvey.
• Habitat assessment.
• Status and trends monitoring of a representa-
_ tive sample of iaikes (optional).
• Computer hardware arid software (database
management, d0a analysis) aridshff tr0nirig^\
;• bomjmentatipn of program and study plansy pe-
riodic urates of analyses, and periodic review
, of reference conditions and biocriteria.
should have training and experience commensu-
rate with the needs of the program. At least one
staff member should be familiar with establish-
ing a quality assurance framework.
Laboratory and field facilities and services should
be in place and operationally consistent with the
designed purposes of the program so that high-
quality environmental data can be generated
and processed in an efficient and cost-effective
manner (USEPA 1992b). Adequate taxonomic
10-7
-------
Chapter 10
references and scientific literature should
support data processing and interpretation.
Quality management Is an important planning
aspect that focuses attention on establishing
and improving quality in all aspects of the
biocriteria development process. Quality
management requires that all personnel in-
volved in a biocriteria project (from senior
management to field and laboratory technicians)
be aware of and responsive to data needs and
expectations.
10-8
-------
Appendix A
Glossary of Terms
a posteriori classification: a classification
made based upon the results of experimenta-
tion.
o priori classification: a classification made
prior to experimentation.
alkali lakes: also referred to as "soda lakes;"
characterized by high pH (>=pH 10) and a high
concentration of salts.
antidegradation statement: statement that
protects existing designated uses and prevents
high-quality waterbodies from deteriorating
below the water quality necessary to maintain
existing or anticipated designated beneficial
uses.
aquatic assemblage: an association of interact-
ing populations of organisms in a given
waterbody, for example, fish assemblage or a
benthic macroinvertebrate assemblage,
aquatic community: an association of interact-
ing assemblages in a given waterbody, the biotic
component of an ecosystem.
aquatic life use: a beneficial use designation in
which the waterbody provides suitable habitat
for survival and reproduction of desirable fish,
shellfish, and other aquatic organisms.
assemblage structure: the make-up or
composition of the taxonomic grouping such
as fish, algae, or macroinvertebrates relating
primarily to the kinds and number of organ-
isms in the group.
beneficial uses: desirable uses that water
quality should support. Examples are drinking
water supply, primary contact recreation (such
as swimming), and aquatic life support.
Best Management Practice (BMP): an
engineered structure or management activity,
or combination of these, that eliminates or
reduces an adverse environmental effect of a
pollutant.
biological assessment: an evaluation of the
biological condition of a waterbody that uses
biological surveys and other direct measure-
ments of resident biota in surface waters.
biological criteria: numeric values or narra-
tive expressions that describe the reference
biological condition of aquatic communities
inhabiting waters that have been given a
designated aquatic life use.
biological indicators: plant or animal species
or communities with a narrow range of ecologi-
cal tolerance that may be selected for empha-
sis and monitored because their presence and
A-1
-------
Appendix A
relative abundance serve as a barometer of
ecological conditions within a management unit.
biological Integrity; the condition of the
aquatic community inhabiting unimpaired
waterbodies of a specified habitat as measured
by an evaluation of multiple attributes of the
aquatic biota. 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 characteristics such as
taxonomic richness and composition, trophic
structure, and (3) is found in the study biogeo-
graphic region.
biological monitoring: the use of a biological
entity as a detector and Its response as a
measure to determine environmental conditions.
Toxicity tests and biological surveys are com-
mon biological monitoring methods.
biological survey fbiosurvey): the process of
collecting, processing, and analyzing representa-
tive portions of a resident aquatic assemblage to
determine the assemblage structure and function.
biota: plants, animals and other living resources
of a region.
bisection scoring: used when metric value
distribution Is based upon data from unim-
paired reference sites. The 25th percentile
becomes the minimum value for the highest
score: the difference between the 25th percentile
and 0 is divided into two equal parts.
canonical correlation analysis (CC): a linear
multivariate ordination procedure using linear
canonical equations with multiple dependent
and independent variables.
canonical correspondence analysis: a non-
linear multivariate ordination procedure.
Carlson's Trophic State Index (TSI): a numeri-
cal index for estimating lake trophic state on a
scale of 0 to 100 with each increase of 10 in the
index representing a doubling of algal biomass.
coefficient of variation: standard deviation
(from the mean) expressed as a percentage of
the mean.
community component: any portion of a
biological community. The community compo-
nent may pertain to the taxonomic group (fish,
invertebrates, 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.
designated use classifications: classification of
a waterbody or segment based on the purposes
(beneficial uses) for which the waterbody may be
used as specified in water quality standards.
diatoms: any of a number of related micro-
scopic algae, one-celled or in colonies, whose
walls consist of two parts or valves and contain
silica.
discriminant analysis: a type of multivariate
analysis used to distinguish between two
groups.
ecological or environmental indicators:
measurable features of an ecosystem that
singularly or in combination with other features
provide manager!ally useful evidence of water
resource or ecosystem quality, or reliable
evidence of trends in quality. Indicators can be
biological, physical, or chemical measurements,
and can sometimes have elements of more than
one discipline: for instance, concentrations of
chemicals in fish tissue.
ecological integrity: the condition of the biotic
(aquatic community) and abiotic components
(water chemistry and habitat) of unimpaired
waterbodies as measured by assemblage struc-
ture and function, water chemistry, and habitat
measures.
ecological properties: biotic and habitat
attributes of a waterbody.
ecoregions: a relatively homogeneous area
defined by similarity of climate, landform, soil,
potential natural vegetation, hydrology, or other
ecologically relevant variable.
epifauna: benthic animals living on the sedi-
ment or among rocks and other structures.
epilimnion: the upper waters (above the
metalimnion) of a thermally stratified lake.
expert consensus: a method used to establish
reference condition when no candidate sites are
available based on the collective experience and
expertise of regional biologists.
A-2
-------
Glossary of Terms
flowage lakes: areas of a river system which are
sufficiently deep, slow moving and wide to have
lacustrine characteristics. Unlike reservoirs,
they typically have wide Inflow and outflow
regions.
forebay zone: same as the lacustrine zone of a
reservoir.
frustule: the hard shell of a diatom.
gradient analyses: a suite of statistical tech-
niques including principal component analysis,
canonical correlation's analysis, and canonical
correspondence analysis used to examine the
relationships between biotic and environmental
factors.
habitat: a place where the physical and biologi-
cal elements of ecosystems provide a suitable
environment including the food, cover, and
space resources needed for plant and animal
livelihood.
hypoxic: waters that have a very low oxygen
level.
infauna: animals that live within the sediments,
often in holes they have dug.
inflow zone: area where a river enters a reservoir.
interquartile coefficient: the ratio of the
interquartile range of a metric to its scope for
detection.
lacustrine zone: area of a reservoir which is
most lake-like: current velocities are much
slower than for riverine or transitional zones.
Little sediment deposition normally occurs since
most sediment load has been deposited in the
riverine or transitional zones; may thermally
stratify; primary productivity predominates.
lake: a body of fresh or salt water of consider-
able size, whose open-water and deep-bottom
zones (no light penetration to bottom) are large
compared to the shallow-water (shoreline) zone,
which has light penetration to its bottom.
limnology: the study of the functional relation-
ships and productivity of freshwater biotic
communities as they are affected by the dynam-
ics of physical, chemical and biotic environmen-
tal parameters.
littoral zone: the area of a lake near the shore
from the region of the highest seasonal water
level to the deepest point at which attached
submerged macrophytes occur.
log linear models: statistical modeling tech-
niques for dealing with categorical data.
marl lakes: lakes in which solid calcium car-
bonate precipitates during periods of high
photosynthesis forming a characteristic marl
bench in the euphotic zone.
metalimnion: the stratum of steep thermal
gradient that separates the epilimnion from the
hypolimnion in a thermally stratified lake.
morphoedaphic index (MEI): the ratio of
dissolved solids (measured as total dissolved
solids, alkalinity, or conductivity) to mean lake
depth; MEI has been used to predict the total
fish production, phytoplankton standing crop,
and total phosphorus concentration of lakes not
subject to cultural eutrophication.
multiple metric or multimetric approaches:
analysis techniques using several measurable
characteristics of a biological assemblage.
multiple use: when a water body has more than
one beneficial use designation.
multivariate community analysis: statistical
methods (e.g., ordination or discriminant
analysis) for analyzing physical and biological
community data using multiple variables.
ombrotrophic bog: an acidic wetland which
receives all of its nutrients from atmospheric
deposition.
ordination analysis: a set of techniques in
which sampling units are arranged in relation to
one or more coordinate axes such that their
relative positions to the axes and to each other
provide maximum information about their
ecological similarities.
oxycline depth: depth at which dissolved
oxygen levels fall below a threshold value.
paleolimnology: the study of the environmental
history of inland waters, based primarily on
analysis of biological, chemical, and physical
characteristics of sediment cores.
pelagic zone: the area of open water beyond the
littoral zone.
A-3
-------
Appendix A
predictive models: statistical models that can
be used to predict biological response based on
ecological (habitat) variables.
Principal Components Analysis (PCA): a linear
multivariate ordination technique that deter-
mines a reduced set of coordinate axes,
principal axes: new variables created by ordina-
tion analysis that account for variation in the data.
proftuidal zone: the sediments beyond the
littoriprofundal zone.
reference site: a site on a waterbody which
represents the best attainable physical habitat,
water chemistry, and biological parameters for
specific environmental conditions.
reference condition: The chemical, physical, or
biological quality or condition exhibited at either a
single site or an aggregation of sites that represent
the least impared or reasonably attainable
condition at the least impared reference sites.
regression: any of a number of statistical
techniques in which the relationship of one (or
more) variable(s) is (are) estimated as a function
of another variable or variables.
reservoir: a lake created for human use often
as the result of impoundment of a river system;
classified as lake type 73 by Hutchinson (1957).
risk assessment: a scientific process that
Includes hazard identification, receptor charac-
terization and endpoint selection, stress-re-
sponse assessment, and risk characterization.
riverine zone: the relatively narrow and well-
mixed area of a reservoir immediately down-
stream of the river inflow where current veloci-
ties decrease and significant sediment transport
still occurs.
robust: insensitive to assumption violations, i.e.,
holds even when the probability model is incorrect.
scope for detection: the range from 0 to the
lower quartile for metrics that have high values
under unimpaired conditions (e.g., EPT index) or
that range from the upper quartile to 100 for
metrics that have low values under unimpaired
conditions (e.g., percent Chironomidae).
spatial variability: variation in a biological
parameter due to different ecological conditions
among sites.
temporal variability: variation in a biological
parameter due to temporal fluctuations in
ecological condition such as changing water
chemistry or sunlight, e.g., diurnal and seasonal
variations.
Total Maximum Daily Load (TMDL): The total
allowable pollutant load to a receiving water
such that any additional loading will produce a
violation of water-quality standards.
transitional zone: area of a reservoir between
the riverine and lacustrine zones; current
velocities are intermediate, significant sedimen-
tation occurs, light penetration increases and
primary productivity increases.
trisection scoring: used when metric value
distribution is based upon data from reference
and impaired sites (population distribution). The
range of values from the 95th percentile to 0 is
divided into thirds with the top third receiving
the highest score, the middle third receiving the
intermediate score, and the bottom third receiv-
ing the lowest score.
trophic state index: any numerical index for
estimating trophic state of a lake.
tmimodal response: a response In which a
species has [a single) peak abundance at [an]
optimal value [or range] of an environmental
variable and its abundance is lower at higher or
lower values of [that] environmental variable.
univariate tests: statistical tests for comparing
two or more groups; techniques include t-test,
analysis of variance, sign test, Wilcoxon rank
test, and the Mann-Whitney U-test.
water quality standards: provisions of state or
federal law which consist of a designated use or
uses for the waters of the United States, water
quality criteria for such waters based upon such
uses. Water quality standards are to protect
public health or welfare, enhance the quality of
the water and serve the purposes of the Clean
Water Act (40 CFR 131.3) (USEPA 1983) a law or
regulation that consists of the beneficial desig-
nated use or uses of a waterbody, the numerical
and narrative water-quality criteria that are
necessary to protect the use or uses of that
particular waterbody, and an antidegradation
statement (ITPM 1994).
A-4
-------
Appendix B
Comparison of Existing Lakes
Protocols
Two research programs—the U.S. Geological
Survey's North American Water Quality Assess-
ment (NAWQA) and the USEPA Office of Re-
search and Development's Environmental
Monitoring and Assessment Program (EMAP)—
incorporate biological monitoring components
similar to those of USEPA's Biological Monitor-
ing Programs. The following discussion com-
pares these programs, as well as a program
implemented by the Tennessee Valley Authority
(TVA) and USEPA's Clean Lakes Program.
NAWQA, EMAP, TVA, and Clean
Lakes Program
NAWQA and EMAP are scientific research
programs with objectives quite different from
those of state-level biological assessment
programs (USEPA 1990a, Meador et al. 1993).
NAWQA and EMAP differ in the waterbodies they
address: NAWQA is designed for streams and
rivers and EMAP is designed for all environ-
ments. including streams and lakes (EMAP-
Surface Waters). The two programs are set apart
by the statistical and experimental designs used
to accomplish their program.
The NAWQA and EMAP programs make assess-
ments of the environment and trends in envi-
ronmental quality at the national or regional
scales. In these programs, USGS or USEPA
scientists and technicians are directly respon-
sible for collecting data from the field and
analyzing results. NAWQA and EMAP also use
different statistical designs to assess national
trends. The NAWQA approach is to sample
repeatedly at equal time intervals at the same
stations (Gurtz 1994) and to apply nonpara-
metric statistical algorithms to the data,
adjusted for seasonal variation (Meador et al.
1993). The EMAP approach is to overlay a
region or landscape with a grid and then to
monitor a randomly selected subsample of the
grid (USEPA 199Id). Both methods are useful
to evaluate trends in environmental quality.
The TVA program, which monitors ecological
condition of reservoirs,also covers a large
percentage of the country. The assessment is
based on sediment quality, dissolved oxygen
content (DO), chlorophyll a concentration, a
benthic macroinvertebrate index, and a
Reservoir Fish Assemblage Index (RFAI) (TVA
1994).
USEPA's Clean Lakes Program provides
support to water quality agencies for lake
assessment, protection, and restoration. Hie
program requires assessments of dissolved
oxygen (DO), chlorophyll a, Secchi disk
transparency, maerophytes, sediment chemi-
B-1
-------
Appendix B
cal characterization, and tributaiy streamflow
measurements. A complete description of the
Clean Lakes Program recommended sampling
methodology is found In the Clean Lakes
Program Guidance Manual (USEPA 1980a) .
State Mandates
USEPA's Biological Monitoring and Biological
Criteria Programs are targeted to state and
tribal water quality and environmental manage-
ment agencies. Monitoring data is collected at
the state or local watershed level by state agency
personnel or by suitably trained organizations to
identify problems and report on control effective-
ness. The Biological Monitoring Program accom-
plishes its mission through training and the
translation of research techniques to rapid,
efficient and defensible protocols these tech-
niques can be used on an operational basis by
state personnel and others to develop biological
criteria that can be used in management deci-
sion making.
The broad water quality goals of the Clean Water
Act are translated by each of the states into
water quality standards designed to protect
beneficial uses (USEPA 1996a). The USEPA
biological monitoring program therefore looks
to individual states and tribes rather than
federal agencies for the assessment of those
beneficial uses, as they have been designated
in various administrative codes. The protocols
presented in this document are designed to
assist states and tribes in making their
beneficial use assessments in biological terms.
The protocols and measurement techniques
will also help states fulfill their obligation to
report to Congress biennially on the attain-
ment of designated beneficial uses of their
surface waters (USEPA 1994f).
Monitoring Biotic Integrity
Although the statistical designs used by EMAP,
NAWQA, and USEPA's biological monitoring
programs differ, the monitoring components are
generally complementary. These programs
assess conditions of lakes, rather than produc-
ing comprehensive inventories of biological
resources. Some but not all major biological
assemblages—invertebrates, fish, plants, birds,
microorganisms—are sampled, identified, and
counted, using a standardized technique, during
a well defined season of the year.
In this kind of monitoring, the focus is on the
full array of species captured by the sampling
methods to provide cost-effective sampling and
standardization. The methods described in this
document do not target rare species. Issues of
seasonal variation and bias in the sampling
protocols are standardized or "indexed" to
simplify statistical analysis (Sokal and Rolfe
1969). These monitoring programs produce
sensitive and robust estimates of the biological
integrity of the aquatic system, as well as the
impacts that anthropogenic activities might have
on the environment in terms of degrading the
aquatic life designated use (Fausch et al. 1984,
Karr 1991, USEPA 1989b).
It is desirable for all agencies to use similar
water quality monitoring protocols so that data
can be shared and compared. The national
water quality monitoring council, (NWQMC)
which succeeded the Intergovernmental Task
Force on Monitoring (ITFM) is the forum for that
kind of information exchange, and several of the
recommendations of the task force address the
issue of data comparability (ITFM 1992). How-
ever, in the case of biological monitoring, species
characteristics differ regionally, and sampling
techniques are not expected to be the same
across ecoregions. In USEPA's Biological Moni-
toring Program and Biological Criteria Program,
sampling techniques and index period must be
identical for the reference condition and the test
locations, and the use of common techniques
across agencies within a region would signifi-
cantly improve the efficiency and power of
biological monitoring activities. Table B-l
compares lake habitat and biological monitoring
among EMAP, TVA, and the USEPA Biological
Monitoring Program. NAWQA has not published
specific protocols for lakes and reservoirs.
B-2
-------
Table B-1. Comparison of lakes protocols with EMAP, TVA Reservoirs, and Clean Lakes.
USEPA takes Blomonttoring/
Blocrlterla Program
EMAP
TVA Reservoirs
Clean Lakes Program
I. Habitat
Assessment
Single qualitative estimate to 10
stations, vegetation type, canopy layer,
understory, ground cover, shoreline
substrate, bank features, human
influence, bottom substrate, fish cover,
water quality
10 stations, observations are made 10m
from shore; plot dimensions are 15m
x 15m, vegetation type, canopy layer,
understory, ground cover, shoreline
substrate, bank features, human
influence, bottom substrate macrophyte
cover, fish cover, littoral microhabitat,
water quality.
Water quality.
Water quality.
II. Benthic
Invertebrates
Preferred sampling in the subltttoral
zone (profunda! zone Is an alternative)
(Tier 2) Sampling gear depends on
substrate:
Rocks or gravel - dome sampler
Sand - Peterson or Van Veen.
Mud - Ponar.
Ciay - Peterson or Van Veen.
Lake-wide composites of 2 to 3
casts at 3 to 10 stations.
100 individuals at lowest taxon*;
alternatives are more than 100
individuals or identification to family
<100 Individuals).
Sublittoral zone.
K-B cores.
ISO individuals.
No. 60 mesh used to sort organisms
larger than 250jim.
Line of sight transect across width of
reservoir; 10 samples collected at equal
intervals along transect
Ponar sampler for mud substrates.
Peterson sampler for rocky substrate.
Organisms identified to lowest level,
generally species or genus.
Organisms reported as no/m2.
* ,< vi" ».v'" 'V,
'-'v'.'., r ' V.-/i,.' c
III. Hsh
Eiectrofishing, gill nets, seines, fykes,
and minnow traps.
Identify, measure length and weight of
fish.
Inspect all fish for external anomalies
(supplemental).
Eiectrofishing, gill nets, seines, fykes,
and minnow traps.
Identify, measure length and weight of
fish.
Inspect all fish for external anomalies.
15 eiectrofishing runs (300m each) In
each location of reservoir (Inflow,
transition, forebay); all habitats (e.g.,
bluff, rip-rap, mud) sampled in
approximate proportion to Mr
occurrence in the sampling location.
Identify, measure length and weight of
fish. Young of the year fish counted
separately from adults. Inspect all fish
for external diseases, parasites,
anomalies.
Preproject monitoring (phase 1)
requires:
General discussion of fish populations
and ecological relationships.
Standard fish flesh analyses for organic
and heavy metal contamination if there
is significant public consumption.
No specific protocols recommended for
phase 2 and 3 monitoring.
IV.
Zooplankton
Vertical tow with 1Qjun mesh net
(Tier 3).
Identify 100 organisms to family or
species; measure Daphnia.
Vertical tow with dual (bongo) net; 48-
and 220pm mesh.
Identify to species; measure.
No protocols recommended.
No protocols recommended.
-------
Table B-1. (continued)
EPA Lakes Biomonitorfng/
Biocriterla Program
EMAP
TVA Rcstrvolrs
Clean Lakes Program
V. Sediment
Diatoms
Sediment grab for recent diatoms
(Tier 2).
Sediment cores for paieoilmnology
(supplemental).
Identify to species.
Single sediment core at index site with
modified K-B corer.
No protocols recommended.
No protocols recommended.
VI, Birds
No protocols recommended.
Special team of ornithologists visit lake.
May-July. Canoe shoreline transect;
record birds seen or heard for 5 minutes
at 200m Intervals.
No protocols recommended.
No protocols recommended.
VII,
Phytoplankton
(including
Chlorophyll)
Chlorophyll a (Tiers 1-3) Seasonal
composite from 6 to 12 spaced samples
(Tier 3).
Chlorophyll a only (1.5m).
Chlorophyll a; Mean of euphotlc zone
concentration.
Growing season mean of 6 to 10
sampling periods.
Chlorophyll a: Depth Integrated from
the top 6ft. of the water column
preferably from the deepest part of the
lake. Monthly sample September-April,
biweekly sample May-August.
VIII.
periphyton
Sample 3 to 5 substrates, composite
300-500 diatom frustules identified to
species (supplemental).
No protocols recommended.
No protocols recommended.
No protocols recommended.
IX
Submerged
Macrophytes
Estimate cover, identify dominant
species (Tier 1). Tiers 2-3: 5 to 10
samples. Sample with rake. Identify to
species, weigh each (net weight).
Estimate percent cover (habitat
component).
Macrophyte coverage determined by
color aerial photography.
Boat surveys to identify dominant
species.
Document species composition,
distribution, and depth during the
growing season. Include community
types and abundance along with
species list.
-------
Appendix D
Biological Assemblages
D. 1 ALGAE
Algae dominate the primary production of most
lake ecosystems, occurring as free-floating
phytoplankton or attached periphyton. Phy-
toplankton are the base of most lake food webs,
and fish production is linked to phytoplankton
primary production (e.g., Ryder et al. 1974).
Excessive nutrient and organic inputs from
human activities in lakes and their watersheds
lead to eutrophication, characterized by in-
creases in phytoplankton biomass, macrophyte
biomass, nuisance algae blooms, loss of water
clarity, and loss of oxygen from bottom waters.
From a human perspective, problems may also
include loss of aesthetic appeal, decreases in
desirable gamefish, loss of accessibility from
increased macrophyte production, and in-
creased cost of treating drinking water.
Measurements of algae include estimation of
total biomass with water column chlorophyll a
concentration, and identification and counts of
individual species within several subgroups,
including periphyton (attached forms), periph-
ytic diatoms, phytoplankton (free-living, float-
ing), phytoplankton diatoms, and sediment
diatoms. Diatoms are identifiable by their
frustules (valves) of silica and thus do not
require preservation and identification of soft
parts (Dixit et al. 1992). Analysis of diatoms can
be done on frustules preserved in lake sedi-
ments, attached assemblages on natural
substrates (rocks, macrophytes), or attached
assemblages that colonize artificial substrates.
Planktonic chrysophyte scales are also pre-
served in lake sediments, and their spatial
distribution is less variable than that of
diatom assemblages (Smol et al. 1984).
Algal assemblages respond rapidly (in days) to
changes in their environment with concomi-
tant changes in overall abundance, growth
rates, and species composition, and therefore
do not integrate conditions of the lake. Algal
species have characteristic optimal nutrient
and trace element requirements, and specific
tolerances for cations, salinity, pH, etc. (e.g.,
Tilman 1982). Changes in physical and
chemical water quality (nutrient concentra-
tions, loadings, salinity, temperature, turbid-
ity) can thus lead to a rapidly changed species
composition (Charles and Smol 1994, Dixit et
al. 1992). These mechanisms form the basis of
algal associations, seasonal assemblage
succession (Hutchinson 1967), and the well-
known responses to nutrient enrichment
(Reynolds 1984). Temporal variability is the
greatest disadvantage of indicators based on
measurements of the algal assemblage, and
either repeated sampling or temporally inte-
grated samples are required to obtain a
D-1
-------
Appendix D
seasonal or annual assemblage estimate. The algal
assemblage seasonal succession cycles are only
general and their exact timing and composition
are not predictable (Reynolds 1984). In addition,
assemblage composition and abundance are
influenced by grazing pressure from zooplankton.
Por example, lakes with abundant, large-bodied
zooplankton suspension feeders may have greater
water clarity than similar lakes without the
grazers (e.g., Edmondson and Litt 1982). Other
advantages and disadvantages of using algal
assemblages are listed in Table D- 1.
Primary Production
Nutrient enrichment from human activities
generally leads to increased biomass of algae in
lakes. Measurement of algal and other plant
biomass, or a surrogate (e.g.. chlorophyll a), is a
good indicator of eutrophication and is clearly
related to biological integrity, meeting the first
criterion for successful metrics. Trophic state is
an expression of the production of a lake
ecosystem. We use Carlson and Simpson's
(1996) definition of trophic state based solely on
biomass and operationally measured by vari-
ables that estimate biomass.
Carlson's Trophic State Index (TSI) (Carlson
1977) is the most widely used index for
eutrophication in lake monitoring programs,
and is based on epilimnetic chlorophyll a
concentration, total phosphorus concentration,
and Secchi depth. Index values range from
below 0 (ultraoligotrophic) to above 100
(hypereutrophic). Each of three measures
(chlorophyll, total P, Secchi) is used to calculate
an independent index, and the indices can be
compared to identify whether algal growth is
limited by phosphorus, light, or other nutrients.
Chlorophyll is the most accurate and is the
preferred indicator of trophic state. Total phos-
phorus and Secchi depth indices are also used
as surrogates in the absence of chlorophyll data,
and can be used to identify factors contributing
to algal growth when all three are measured
(Carlson and Simpson 1996). Measurement of
primary productivity is not recommended
because it is expensive to measure and fre-
quently difficult to interpret.
Other indices have been developed and might be
appropriate for different lake ecoregions in the
country. A nitrogen index can be included to
identify nitrogen limitation (Carlson 1992,
Kratzer and Brezonik 1981). Other trophic state
models (e.g., Dillon and Rigler 1974, Larsen and
Mereier 1976, Vollenweider 1975) use annual
phosphorus loading rates or retention fractions,
and rely on measurements of nutrient concen-
trations rather than the biological response to
nutrient loading. See Carlson and Simpson
(1996) for a complete discussion of the trophic
state concept.
Both algae and macrophytes contribute to a
lake's plant biomass, therefore, metrics for both
algal and macrophyte biomass are preferred for
whole-lake trophic state (Canfield et al. 1983,
Carlson and Simpson 1996) (see section 4.2 for
macrophyte biomass).
Advantages
Disadvantages
Alternatives
Species composition, abundance
respond to water quality:
- Nutrients (N, P, Si)
- pH, alkalinity
- Metals
- Temperature
Field sampling relatively easy.
Identification rapid to division and
family level.
Simple biomass indicators
(chlorophyll a or dry weight).
Historic and prehistoric record in
sediment diatoms.
Require taxonomic expertise for
species indentification.
Strong temporal variability; do not
integrate (except sediment
diatoms).
Quantitative inference of water
quality requires large calibration
data set.
Water quality measures (N, P),
pH, alkalinity measurement.
Metals analysis.
BOD, COD.
ATP
Table D-1. Advantages, disadvantages, and alternatives to using algal assemblages.
D-2
-------
Biological Assemblages
The Tennessee Valley Authority (TVA) uses
chlorophyll a concentrations for one of Its
reservoir assessment metrics. The metric is
based on the mean growing season water
column concentration and a single maximum
concentration (Dycus and Meinert 1994, TVA
1995). Reservoirs are considered mesotrophic or
oligotrophic, based on natural watershed
geochemistry and expected chlorophyll a con-
centrations. Low, moderate, and high mean
concentrations of chlorophyll a are rated
"good"," fair," and "poor," respectively, with
differing definition of these three categories for
mesotrophic and oligotrophic reservoirs. A very
high single-sample maximum (> 30 ng/L)
reduces the rating by one class. Thus, a good
rating implies chlorophyll concentrations within
the range expected and no extreme blooms. In
addition, for mesotrophic reservoirs unusually
low concentrations (< 3 Hg/L) are rated "fair."
Also, if this low concentration occurred despite
sufficient phosphorus, it was considered an
Indication of limitations other than nutrients
and resulted in a poor rating.
Phytoplanlcion Species
Composition
Many different levels of algal monitoring and
assessment exist. Metrics based on indicator
taxa can be quite simple, such as qualitative
estimates of relative dominance of algal divisions
(Table D-2). For example, dominance by diatoms
might be rated "good", and dominance by
cyanobacteria might be rated "poor," requiring
only a rapid, qualitative estimate of the relative
abundances of diatoms and cyanobacteria.
Indicator genera could also be used, For ex-
ample, abundant populations of the
cyanobacteria OsciUatoria or Anabaena indicate
eutrophication (e.g., Edmondson and Lehman
1981}. Certain diatoms and chrysophytes are
sensitive to pH and dissolved aluminum
(Charles and Whitehead 1986, Smol et al. 1984).
Algal assemblage data, consisting of taxonomic
identifications and abundance (relative or
absolute) of each taxa, can be analyzed in two
ways:(l) by determining assemblage metrics
based on species structure, or (2) by multivari-
ate assemblage analysis. Simplified field and
laboratory procedures are possible for some (but
not all) of the species structure metrics.
Due to the high temporal variability of plankton,
several samples during the growing season
might be needed for accurate assemblage
analysis.
Assemblage metrics, as defined and used in
assessment of biological integrity, rely on the
comparison of a metric to a reference value.
Assemblage metrics possible for use in algal
analysis include (Bahls 1993):
• Diversity metrics, such as number of taxa ,
percent contribution of dominant taxon, and
Shannon-Wiener diversity which incorpo-
rates both number of taxa and evenness,
and is less sensitive to sample size.
• Indicator taxa (e.g., bloom-forming
cyanobacteria) respond to acid, eutrophica-
tion, metals, organics, salinity, and climate.
Responses are reliable to water chemistry
and many responses of individual species
are unimodal.
• Indices and ratios.
• Pollution tolerance index, based on toler-
ance groups of Lange-Bertalot (Bahls 1993).
Metric
Optimal Condition
Impaired Condition
Trophic state.
No. of Taxa.
% dominance.
Indicator taxa, ecological
categories, and tolerance indices.
Similarity indices.
Similar to reference expectation.
High
Low
Similar assemblage to reference.
Similar to reference.
Substantially higher or lower.
Reduced number of taxa.
High dominance.
Abundance of indicator, taxa or
high index values (e.g.,
cyanobacteria acidophilic taxa,
tolerant species, etc.)
different.
D-3
Table D-2. Potential algal metrics.
-------
Appendix D
This Index is functionally similar to the
Hilsenhoff Biotic Index for invertebrates
(Hilsenhoff 1987).
• Similarity indices, comparing the similarity
in assemblage composition to reference
conditions (canonical correspondence, other
ordination; Jaccard's similarity, other).
• Ratios of algal divisions (e.g.,
cyanobacteriaitotal) or other functional
groupings (e.g., motile cellsitotal).
• Ecological categories.
• Trophic state categories (eutrophic, olig-
otrophic, acidophilic).
• Inferred water chemistry.
• Requires a large calibration data set to
develop predictive model of water chemistry.
Spatial Variability—Phytoplankton can be
patchily distributed in a lake, affecting the
variability of a sampling program. Most phy-
toplankton patchiness is the result of water
motion and identifiable water masses, such as
Langmuir circulation, vertical stratification, and
embayments with limited exchange to open water
due to morphometry or submerged vegetation.
Effects of these can be minimized by taking
vertically integrated samples in mid-lake, with a
vertical tow, a pump, or a series of bottle samples.
Temporal Variability—The largest single disad-
vantage of phytoplankton sampling, including
biomass and chlorophyll a measurement, is
temporal variability. The algal assemblage
seasonal succession cycles are only general, and
their exact timing and composition are not
predictable (Reynolds 1984). The variability is
best controlled with repeated sampling (typically
monthly or weekly) using a minimum of 10
samples to obtain either an annual average or
an index period average (e.g., growing season,
spring overturn, peak biomass) (Knowlton and
Jones 1989).
Sedlmented Diatoms
Diatoms and chrysophytes preserved in lake
sediments are integrators of lake history and
make it possible to infer changes in other biotic
assemblages (Charles et al. 1994, Dixit et al.
1992). Environmental variables, such as alka-
linity. aluminum, dissolved organic carbon,
salinity, nickel, conductivity, calcium, total
nitrogen, total phosphorus, Secchi transpar-
ency, and trophic state have been inferred using
diatom-based predictive models (Charles et al.
1994, Dixit et al. 1992, Fritz 1990).
The diatom fossil record can aid in establishing
reference conditions. See Appendix C for meth-
ods. Surface sediments represent recent or
current lake conditions and usually integrate
the assemblage over 1 or more years (Dixit et al.
1992). Presettlement conditions may be charac-
terized by sediment cores of 0.5 to 1.0m depths
(Charles et al. 1994). Dating sediment cores is
possible using pollen or radioactivity of 210Pb
(radon decay product).
Periphyton
Periphyton Spatial Variability—Feriphyton
abundance and species composition might be
variable around the periphery of a lake owing to
differences in water quality, local variation of
runoff from the shore, differences in substrate,
and other factors. Periphyton may be scraped
from natural substrates, or artificial substrates
may be deployed for periphyton colonization
(Kentucky DEP 1993, Bahls 1993, Florida DEP
1996, Oklahoma CC 1993). A composite sample
from several substrates at several sites should
remove most of the effects of local spatial
variability.
Temporal Variability—Like phytoplankton,
periphyton are subject to changing water
chemistry and seasonal succession. Sampling
during an index period in a time of relative
stability might remove most of the confounding
effects of time.
Response of Metrics—Although periphyton have
been used successfully in streams (e.g., Bahls
1993, Patrick 1949), their application as lake
indicators is relatively new. Metrics of periph-
ytic diatoms have shown promise for
bioassessment, based on investigation of
undisturbed reference lakes in Montana
(Gerritsen and Bowman 1994), but actual
response to disturbance or pollution is as yet
unknown. Periphyton are considered an experi-
mental assemblage for lake assessment because
of limited information on response to stressors.
D-4
-------
Biological Assemblages
D.2 SUBMERGED
MACROPHYTES
Aquatic plants respond to nutrients, light, toxic
contaminants, salt, and management. A lack of
macrophytes might indicate water quality-
problems due to herbicides, salinization, or
excessive turbidity. Submerged and floating
macrophytes respond to nutrients in the sedi-
ment (Barko et al. 1992), and an overabundance
of submerged or floating leaved plants can be an
indicator of excess nutrients. Exotic species
(e.g., Eurasian water milfoil) often become
dominant and cause weed problems under
eutrophic conditions. In addition, submerged
macrophytes are sensitive to shading by turbid-
ity and by dense periphyton growth. Many
species are sensitive to phytotoxins, such as
copper and herbicides.
Submerged macrophytes are extensively man-
aged. Exotic species frequently dominate
eutrophic lakes, and control attempts include
harvesting, herbicides, and grass carp. Natural
macrophytes are managed where they are
thought to interfere with recreation.
Extreme eutrophication in shallow lakes may
have alternate stable states: one dominated by
macrophytes, the other by phytoplankton
(Scheffer et al. 1992). Management of such lakes
to promote the macrophyte dominated state
includes removal of planktivorous fish and
introduction of macrophytes and piscivorous
gamefish (Hosper et al. 1992).
Macrophytes respond more slowly to environ-
mental changes than do phytoplankton or
zooplankton and might be better integrators of
overall environmental conditions (Table D-3).
This would allow a single sampling event per
year, during the time of maximum abundance of
macrophytes. Both floating leaved and emergent
plants are easily assessed from aerial photo-
graphs, which permit estimates of total area
covered and percent cover (density) within
stands. For the purposes of lake assessment,
emergent vegetation (i.e., semi-terrestrial) is lake
habitat, but floating and submerged vegetation
are lake biota.
Macrophyte Indicators
Extent and Percent Cover— Extent and percent
cover of rooted vegetation are easily obtained
from rapid surveys or remote sensing (aerial or
satellite imaging). These methods have been
used successfully to monitor the status and
trends of submerged vegetation in estuaries
(e.g., Orth and Moore 1983). Extent of both
floating-leaved and emergent vegetation can be
estimated from aerial photos or from shorezone
surveys. Wetlands can also be estimated from
maps developed by the National Wetlands
Inventory (NWI), although these would not
indicate the extent of littoral emergent vegeta-
tion in most lakes. When compared to expected
or reference values, the extent and percent cover
of macrophytes and emergents provide an
assessment of the overall integrity of the lake
system. Loss of emergents and wetlands on a
lake margin indicates lost wildlife habitat and
possibly increased nutrient and sediment input.
Nuisance weed problems might indicate
eutrophication, and loss of native macrophytes
(compared to reference) might indicate excess
turbidity or toxic contamination.
Technical Issues
Spatial Variability—With suitable substrate and
sufficient light, macrophytes colonize the littoral
areas in lakes and reservoirs. Spatial variability
of cover and extent within these areas can be a
result of one or more of the factors listed below:
• Substrate type—bedrock would be colonized
by periphyton instead of macrophytes.
• Topography.
• Disease.
• Insect infestation.
• Local sources of nutrients and turbidity.
Vegetation functional measurements such as
net growth, primary productivity, etc., are time
consuming and require repeated monitoring at
different times in the growing season. It is not
clear that the information gained from func-
tional measurements is any better for assess-
ment and management purposes than remote,
wide-scale measurements.
Assemblage Metrics—Identification of taxa and
relative abundance counts or biomass estimates
of each allow calculation of similar assemblage
metrics described for the algae assemblages
(Table D-4).
D-5
-------
Appendix D
To minimize effects of variability, several sites
are sampled in a lake and combined Into a
composite sample.
Temporal Variability—Aquatic macrophyte
assemblages on the whole are usually at maxi-
mum cover and extent in midsummer. Temporal
variability is avoided by sampling the macro-
phyte assemblage at approximately the same
time every year. Interannual variability of
macrophyte cover can be high (Scheffer et al.
1992); if so, total vegetated area may not be an
effective metric.
Research Needs—It is generally accepted that
macrophytes respond to nutrients by expanding
their extent and cover. Research is needed to
determine which species respond to contami-
nants such as acid, metals, organics, and
salinity. Macrophytes might respond to indi-
vidual contaminants or only a combination of
contaminants. They might respond to contami-
nants only at extreme levels or conditions.
D.3 BENTHIC
MACROINVERTEBRATES
Benthic invertebrate assemblages in lakes
correspond to particular habitat types and can
be classified according to the three basic habi-
tats of lake bottom: littoral, sublittoral, and
profunda!. The littoral habitat of lakes usually
supports larger and more diverse populations of
benthic invertebrates than do the sublittoral
and profundal habitats (Moore 1981,
Wiederholm 1984). The vegetation and substrate
heterogeneity of the littoral habitat provide an
abundance of microhabitats occupied by a
varied fauna, which in turn enhances inverte-
brate production. The littoral habitat is also
highly variable due to seasonal influences, land
use patterns, riparian variation, and direct
climatic effects producing high-energy areas.
The epifauna species composition, number of
individuals, areal extent, and growth form vaiy
with the species composition of the macrophyte
beds, making it difficult to determine the
benthic status accurately.
The sublittoral habitat, below the area of dense
macrophyte beds, but above typical ther-
moclines, lacks the heterogeneity of the littoral
habitat; However, it is also less subject to littoral
habitat variables and influences. The sublittoral
habitat is rarely exposed to severe hypoxia but
might also lack the sensitivity to toxic effects
that is found in the profundal habitat. The
sublittoral habitat supports diverse infaunal
populations, and standardized sampling is easy
to implement because a constant depth and
Advantages
Disadvantages
Alternatives
Respond to:
- Nutrients
Subject to management (planted,
removed, poisoned).
TSI
Secchi
- Metals
Not important in some regions.
Nutrient analysis.
- Herbicides
Metals analysis.
- Turbidity
Herbicide analysis.
- Water level change
Structural component; littoral
habitat for fauna.
Sampling relatively easy (aerial
photography or transects); simple
abundance metrics.
Integrators of environmental
conditions.
Endpoints of concern (weeds,
wetlands, SAV loss).
Table D-3. Advantages, disadvantages, and alternatives to using macrophyte assemblages.
D-6
-------
Biological Assemblages
substrate can be selected for sampling. There-
fore, the sublittoral habitat is the preferred
habitat for surveying the benthic assemblage in
most regions.
The profundal habitat, in the hypolimnion of
stratified lakes, is more homogeneous due to a
lack of habitat and food heterogeneity, and
hypoxia and anoxia in moderately to highly
productive lakes are common. The profundal
habitat is usually dominated by three main
groups of benthic organisms including chirono-
mid larvae, oligochaete worms, and phantom
midge larvae (Chaoborus) (Wiederholm 1984).
Many species of chironomids and tubificid
oligochaetes are tolerant to low dissolved
oxygen, such that these become the dominant
profundal invertebrates in lakes with hypoxic
hypolimnia. As hypoxia becomes more severe
tubiflcids can become dominant over chirono-
mids (Hergenrader and Lessig 1980). In cases of
prolonged anoxia, the profundal assemblage
might disappear entirely. If hypoxia is rare in
reference lakes of the region, and if toxic sedi-
ments are suspected to occur in some lakes,
then the profundal habitat might be preferred
for the region.
Benthic macroinvertebrates are moderately long-
lived and are in constant contact with lake
sediments. Contamination and toxicity of
sediments will therefore affect those benthic
organisms which are sensitive to them
(Wiederholm 1984). Acidification of lakes is
accompanied by shifts in the composition of
benthic assemblages to dominance by species
tolerant of acidic conditions (Perry and
Troelstrup 1988, Schindler et al. 1989). Effects
of rapid sedimentation are less well-known but
Table D-4. Potential macrophyte metrics.
appear to cause shifts toward lower abundances
and oligotrophie species assemblages as well as
more motile species (Masters 1992, Wiederholm
1984).
Benthic macroinvertebrates are present year-
round and are often abundant, yet not very
motile. However, the benthos integrate environ-
mental conditions at the sampling point (Table
D-5). To date, TVA, EMAP, and several states
(Florida, Oklahoma, North Dakota) have sur-
veyed benthos as part of lake bioassessment in
the United States. Developmental work by TVA,
USEPA, and several states is likely to refine
metrics based on macroinvertebrates.
Invertebrate Indicators
Primary emphasis in the past has been placed
on chironomids and oligochaetes as indicators
of lake trophic status. Several indices and
classification systems have been developed for
lake trophic state using chironomid and oli-
gochaete assemblages as indicators (e.g.,
Naumann 1932). The trophic indices, most of
which were developed for lakes of northern
Europe, rely on relative abundances of chirono-
mid species, the ratio of tolerant to intolerant
tubificid oligochaetes, or the ratio of oligocha-
etes to chironomids (reviewed in Wiederholm
1980). Ratios are unstable metrics because
numerator and denominator are independent
(Barbour et al. 1992); proportions or percentage
metrics work better.
TVA is using benthic macroinvertebrate compo-
sition as one of five assessment indicators in
reservoirs (Dycus and Meinert 1992, Dycus and
Metric
Optimal Condition
impaired Condition
% cover or biomass In available
habitat colonized.
% cover, biomass in vegetated
areas.
No. of taxa.
% cover, biomass of dominant
species.
No. of exotic species.
% cover, biomass of exotics.
Similar to reference.
Similar to reference.
High
Low
Zero
Zero
Substantially more or less than
reference.
Substantially more or less than
reference.
Low
High
£1
High
D-7
-------
Appendix D
Melnert 1993, Dycus and Meinert 1994). TVA
bcnthlc composition metrics evaluate richness,
composition, abundance, and indicator taxa.
The condition of macroinvertebrate assemblages
in TVA reservoirs is strongly associated with
hypoxia in the reservoirs (after Dycus and
Meinert 1992). The EMAP surface waters pilot
project is also using benthic macroinvertebrates
for assessing the biological condition of lakes
and has found that number of taxa among
benthic macroinvertebrates corresponds to level
of disturbance in a watershed (USEPA 1993a).
Lake benthic metrics that are responsive to
stresses, are in general, similar to stream
invertebrate metrics (Table D-6). Metrics used
successfully by TVA in assessing reservoirs
include (TVA 1994, TVA 1995):
• Number of taxa.
• Number of long-lived taxa (Corbicula,
Hexagenla, mussel, snails),
• Number of EPT taxa (Ephemeroptera,
Plecoptera, Trichoptera).
• Proportion as Tubijlctdae.
• Proportion as dominant taxon.
• Total abundance excluding Chironomidae
and Tiibljlcidae.
• Percentage of samples on a transect with no
organisms present.
Invertebrate metrics demonstrated to respond to
stresses in Florida lakes include ( FDEP 1994,
Gerritsen and White 1997):
• Number of taxa.
• Shannon-Wiener diversity.
• Percent oligochaetes.
• Number of ETO taxa (Ephemeroptera,
Trichoptera, Odonata).
• A tolerance index similar to HBI.
Biological assessment using benthic
macroinvertebrates must focus on a subset of
assemblages (defined by habitat and season) to
avoid costly sampling of all assemblages.
Assemblage composition is affected by sub-
strate, macrophytes, depth, and season. The
optimal assemblage for reasons of cost, variabil-
ity, and interpretation appears to be the sublit-
toral assemblage of epifauna and infauna. The
littoral assemblage is highly variable and costly
to sample, and the profundal assemblage might
be uniformly impacted by hypoxia in many
regions of the country. Hypoxia might be natural
in deep, mesotrophic lakes or in warm water
lakes. If hypoxia is an expected profundal
condition, sublittoral benthos is the preferred
assemblage. If hypoxia is rare or not expected in
the reference condition, profundal benthic
sampling might be preferred.
Advantages
Disadvantages
Alternatives
Respond to:
- DO
- Sediment metals
- Other sediment toxins
- Organic enrichment
- Fish
Integrators of environmental
conditions.
Low mobility.
Moderate temporal variability.
Trophic link to fish, birds.
High spatial variability due to
habitat dependence.
Littoral habitat sampling difficult.
Metrics not well developed or
tested in lakes.
Laboratory identification and
count can be time-consuming,
requires expertise.
DO
Sediment TOC.
Toxicity bioassays.
Fish assemblage.
Table D-5. Advantages, disadvantages, and alternatives to using macroinvertebrate assemblages.
D-8
-------
Biological Assemblages
Technical Issues
Spatial Variability—To account for spatial
variability within the sampling area of a lake, at
least three grabs must be taken. The grabs can
be combined into a composite sample to save
money, but valuable information is lost. For
example, data on spatial variability is lost, but
more importantly effects of one sample with a
veiy large density of a single taxon will be more
significant in a composite sample than in the
average of individual samples. In large lakes or
lakes with heterogenous bottom substrate, five
or more sites might need to be sampled. Selec-
tion of the epifauna or infauna for sampling will
depend on the major substrate type present and
the overall objectives of the biosurvey. For
example, with sediment problem the benthic
infauna would be the appropriate part of the
assemblage to sample. If the major substrate
type present is hard substrate or vegetation, the
epifauna should be sampled.
Toxic or contaminated sediments are more likely
to be a stress on profundal invertebrates be-
cause sediments accumulate in the deep,
depositional areas and infaunal oligochaetes
might be more sensitive to toxicity than are
other invertebrates. However, the sublittoral
habitat has certain advantages for sampling
macrobenthos because it is subject to hypoxia
less frequently than the profundal habitat and
because the sublittoral area typically has
greater number of taxa, including some mayflies
and caddisflies than the profundal area.
Temporal Variability
The issue of seasonality needs further investiga-
tion to determine the most effective index period
for sampling or the sampling frequency. Sam-
pling period can be either during the most
stressful period or during a time after recruit-
ment when the populations have stabilized. The
selected period should be of the least conse-
quence to the identification and sampling
process, especially if the sampling is designed
for volunteer monitoring groups. For example,
samples taken right after recruitment will have
early instars that are difficult to identify- If more
than one period is designated, the appropriate
sampling frequency needs to be established.
Sampling Strategies
The sampling area should focus on the most
predominant substrate available and the metrics
Table D-6. Potential benthic metrics.
Metric
Response to stress
No. of taxa.
Reduced
Shannon-Weiner diversity.
Reduced
Mean no. of individuals per taxon.
Variable
% contribution of dominant taxon.
Elevated
% intolerant species.
Reduced
% oligochaetes.
Elevated under organic enrichment.
ETO taxa (ephemeroptera, trichoptera, odonates).
Reduced under enrichment or DO stress.
% non-insects.
Reduced
Crustacean + mollusc taxa.
Reduced under acid stress.
% crustaceans and molluscs.
Reduced under acid stress.
Tolerance indices (e.g., HB1 [Hilsenoff 1987]; Hulberfs
Lake Condition Index [LCI]).
Reduced
% suspension feeders.
Reduced
% shredders.
Reduced under enrichment (not useful in very large lakes).
Abundance (exclude Chlronomidae and Tubificidae).
Reduced
No. of samples with no organisms present.
Increased
D-9
-------
Appendix D
should be developed independent of microhabi-
tat variation. The type of sampling gear will
depend on the substrate being sampled as each
substrate has its own optimal sampling gear.
Standardized sampling techniques for each gear
type should be implemented to allow for the
comparison of data. Processing of samples
should be standardized by using a standard net
size of 595 |im (No. 30 mesh}.
The objective is to adequately characterize the
sampling unit which is a single lake,
embayment, or lake basin. Heterogeneity within
a sampling unit (lake) is not of interest in
bioassessment. Samples from several sites are
combined into a single composite for analysis
and characterization of the lake. To get a
representative sample of benthic invertebrates.
It is necessaiy to sample at several locations,
such as. three to five areas of the sublittoral
zone around the lake. Sampling at each site
might also consist of several grab, which can be
composited to save money.
Research Needs—Six recommendations for
further study were identified during the develop-
ment of this document by the Benthic
Workgroup:
1. Metric development and calibration must
allow for regional modifications. The need
for regional modification of metrics must be
clear so that states do not discount the
program as "not working" if the metrics
being used are not suitable for their region.
2. Sampling methodology must be based on
regional characteristics and must be appro-
priate to the needs of states. Regional
adaptations will be based on substrate,
habitat, lake type, and other environmental
characteristics. Design strategies should
also include ways to evaluate the design,
and identify specific problems/ characteris-
tics so that states can easily identify
whether or not a specific design is working
In their region. One suggestion was that a
questionnaire accompany the guidance to
specifically identify whether the sampling
methods were found to be suitable for the
region after use for a predetermined period
of time.
3. The appropriate number of replicate
samples needs to be investigated in order to
tighten confidence intervals and to resolve
the best returns on the data for the invest-
ment. A determination must also be made
as to whether multiple sampling efforts
should be conducted on the same day or on
separate days.
4. Investigate seasonality so that the best
index period(s) is selected for sampling.
Ideally,sampling should occur during the
period that will least affect the field identifi-
cations and yield the most valuable informa-
tion. More than one index period might be
needed to address specific objectives. Cost-
effective strategies will focus on reduced
frequency of sampling.
5. Investigate the applicability of vertical
stratification of biomass as in estuarine
sediments, and develop a surrogate infaunal
trophic index for lakes that might have
universal application.
6. Investigate the occurrence and causes of
morphological deformities of benthic organ-
isms in response to stressors. This type of
metric would provide information on indi-
vidual health or sublethal effects.
7. Evaluate potential for field identification as
cost saving measure.
D.4 ZOOPLANKTON
Lake zooplankton consist primarily of crusta-
ceans, rotifers, and, to a lesser extent, semi-
planktonic insect larvae of the genus Chaoborus.
Many zooplankton species found in north
temperate lakes are cosmopolitan or wide-
ranging in their distribution (Hutchinson 1967).
There is a strong positive relationship between
the number of crustacean zooplankton species
and lake surface area (Dodson 1992, Fryer
1985), and weaker positive relationships be-
tween number of species and lake productivity,
and the number of neighboring lakes (Dodson
1992).
More than any other assemblage, zooplankton
structure and function are controlled externally
by both higher and lower trophic levels (fish
predators and algal food) and internally by
planktonic predators (Lewis 1979, Zaret 1980,
Carpenter et al. 1987) (Table D-7). Zooplankton
composition and abundance are variable in time
D-10
-------
Biological Assemblages
with numbers changing one to three orders of
magnitude within weeks. The complexity of open
water zooplankton dynamics is in part due to
trophic interactions taking place in a three-
dimensional environment of reduced structure
(Gerritsen 1980).
The trophic cascade can be modified by nutrient
enrichment and internal interactions (Carpenter
et al. 1987) and can in turn affect physical
characteristics such as light penetration and
temperature (Mazumder et al. 1990).
Zooplankton Indicators
Zooplankton indicators that have been investi-
gated rely on measurement of plankton size
structure, and trophic categories (Stemberger
and Lazorehak 1994) (Table D-8). From the
ecological interactions listed above, zooplankton
body size is a potential indicator of the presence
or absence of planktivorous forage fish, and of
the absence or presence, respectively, of large
piscivores. Use of zooplankton body size as an
indicator (Mills and Schiavone 1982, Mills et al.
1987, O'Gorman et al. 1991) showed that mean
zooplankton body size can predict populations
of yellow perch and migration of ale wives.
Furthermore, dominance by large, visible
Daphnta species (e.g., D. pulex, D. galeata)
indicates the presence of large piscivores,
eireumneutral pH, and the absence of blue-
green algal blooms (Edmondson and Litt 1982,
Mills et al. 1987).
Several zooplankton species, especially some of
the larger predators and Daphnids, are sensitive
to acidification, and acidic lakes have fewer
zooplankton taxa than eireumneutral lakes
(Baker and Christensen 1991). Large Daphnia (>
1 mm) are used as an indicator of trophic
balance in operational biomanipulation in
Europe (Hosper et al. 1992, Hosper and Meijer
1993), and lakes with large Daphnia have lower
chlorophyll concentrations than comparable
lakes without (Mazumder 1994).
The EMAP Surface Waters program is testing
selected zooplankton metrics in New England
lakes. EMAP zooplankton sampling consists of a
single vertical tow at the deepest point of a lake,
using a dual (bongo) net, with a fine (48nm) net
and a coarse (202|im) net (USEPA 1994a,
USEPA 1994b).
Technical issues
Spatial Variability—Zooplankton are subject to
many of the same water movements that affect
phytoplankton. In addition, many species
perform diurnal vertical migration. Integrated
sampling of the mid-lake water column with a
vertical or oblique tow is usually sufficient for
relative abundances of zooplankton species. To
avoid possible effects of vertical migration,
samples should not be taken near dawn or
dusk.
Advantages
Disadvantages
Alternatives
Respond to:
- Fish
- Phytoplankton
- Thermal loading
- Acidity
- Pesticides
Field sampling and counting
relatively easy (but does require
taxon. expertise).
Trophic link to fish.
Sedimentary record for some
groups.
Response to human stressors and
impacts not well documented.
Interpretation difficult: respond to
both higher and lower trophic
levels.
Do not integrate well (high
temporal variability).
Fish assemblage.
Trophic state (Secchi depth,
chlorophyll, phosphorus).
Algae
D-11
Table D-7. Advantages, disadvantages, and alternatives to using zooplankton assemblages.
-------
Appendix D
Index Period—Zooplankton assemblages are not
stable In time undergoing seasonal succession.
To the extent that assemblages are seasonally
predictable, they can be sampled within an
index period. Mid-summer or mid-winter are
relatively stable periods. Midsummer is pre-
ferred to coincide with other assemblages.
Research Needs—Although preliminary results
from EMAP are encouraging, the responsiveness
and reliability of many zooplankton-based
metrics are not yet well known. Response of
zooplankton metrics to stressors, needs to be
tested in different regions of the country.
Seasonal variability and predictability of zoo-
plankton assemblages needs to be analyzed to
determine optimal index periods and the mini-
mum number of samples required to character-
ize a lake.
D.S FISH
Fish populations are powerful structuring forces
on other lake assemblages through feeding
Interactions (trophic cascades). Abundant
populations of piscivorous fish reduce
planktivorous forage fish species, releasing
predatory zooplankton from predation, resulting
in dominance by large-bodied, suspension-
feeding zooplankton (e.g., Brooks and Dodson
1965, O'Brien 1979). The large suspension-
feeding zooplankton can in turn reduce phy-
toplankton abundance. Increasing water clarity
and altering the thermal structure of the lake
(Mazumder et al. 1990). The trophic cascade
also Influences, and is influenced by, nutrient
dynamics (Carpenter et al. 1987).
Table 0-8. Potential zooplankton metrics.
It is well known that fish production is tied to
lake primary production (e.g., Oglesby 1977,
Ryder et al. 1974). In fact, oligotrophic lakes
are often fertilized by fishery agencies to en-
hance sport fish production. In addition, there
are regional, geographic differences in fish
abundance that are not explained by trophic
state (Nurnberg 1996). Moderate to severe
eutrophicatlon reduces and might eliminate
desirable sport fish due to loss of habitat, poor
water quality, and food web simplification (NRC
1992). Fish are highly dependent on habitat for
spawning and for refuge. Some species (e.g.,
yellow perch, most salmonids) spawn in
streams; others require clean rock or gravel
habitat in the lake. Submerged vegetation
provides cover for both forage fish and
piscivores, and recolonization of littoral areas by
macrophytes increases sportfish abundance, as
well as improving water quality.
More than any other assemblage, fish are subject
to management which can confound assessment
efforts (Table D-9). Exotic piscivorous sport fish
(e.g., striped bass, Pacific salmon) are wide-
spread in lakes throughout the United States,
and many of these populations are maintained
by regular stocking. Stocking to maintain a
population results in an artificially large popula-
tion—especially juveniles—with resultant trophic
cascade effects on zooplankton and phytoplank-
ton. This problem is especially pronounced in
"put and take" fisheries, where large numbers of
hatchery-reared adults are released for a fishing
season and decimate invertebrate assemblages
during the season. In general, if exotic piscivo-
rous species reproduce naturally, biological
integrity is less likely to be affected.
Metric
Response to stress
% large Daphnla (> 1 mm).
Low
No. of taxa.
Reduced under contamination or stress.
% dominance.
High
Size structure (% of large animals or % of small animals).
Dominated by small species (e.g., rotifers).
Trophic structure metrics
- No. of trophic links
- Complexity measures
- % large predators
- No. of predator species
Simplified trophic structure.
D-12
-------
Biological Assemblages
FisH Indicators
Assemblage Composition arid Abundance—Mea-
surements of fish assemblage composition and
relative abundance can be incorporated into several
metrics, including the Index of Biotic Integrity (IBI),
an index of several assemblage-level metrics and
their variations, and multivariate assemblage
analysis. Field measurements for these are the
relative abundances of species in the habitat.
Index of Biotic Integrity (IBI)—The Index of Biotic
Integrity (IBI) incorporates attributes of fish
assemblages to evaluate human effects on a
stream and its watershed (Karr 1991, Karr et al.
1986). Those attributes cover the range of
ecological levels from the individual through
population, community, and ecosystem. IBI
consists of 6 to 12 measures, or metrics, in 4
broad categories: species composition, trophic
composition, fish abundance, and condition
(Karr 1991). A site is assigned scores for the
resemblance of each metric to the reference
(unimpacted or least impacted) condition
expected for that area. Total scores of all metrics
result in an overall score for the site.
As with other multimetric indices, component
metrics of IBI require adaptation and calibration
to the geographic regions in which they will be
applied, thus incorporating biogeographic
variation of assemblages and systems into the
assessment (Karr 1991). This may include
deletion or replacement of selected IBI metrics
and is done with the development of a reference
site data base. Local adaptations of IBI for
streams have been developed for several regions
of the United States (Karr 1991, Leonard and
Orth 1986, Miller et al. 1988, Steedman 1988).
Although lakes and reservoirs differ in physical
attributes from rivers and streams (the former
being more homogenous), the valued attributes,
or biological integrity, of fish assemblages apply
equally. These attributes include species compo-
sition, trophic composition, abundance, and
condition. Differences between lake and stream
habitats lie in the expectations for the attributes
and will be reflected in reference site data. An
index used by TVA on its reservoirs is based on
12 metrics and is called the Reservoir Fish
Assemblage Index, or RFAI (Jennings et al.
1995, Hickman and McDonough 1996) (Table D-
10). The status of this index is discussed under
Research Needs.
The major problem in applying IBI to lakes is
obtaining representative samples of fish assem-
blages in lakes. Quantitative sampling in lakes
is not as reliable as that in streams because of
lake morphology, bottom types, and gear
efficiency. Modification of IBI for lakes may
include use of relative abundances based on
subsamples from constant-effort sampling.
Sampling gear and protocols for different
habitats of lakes will need to be standardized.
Advantages
Disadvantages
Alternatives
Respond to:
- DO
- Pesticides
- Metals
- Organic enrichment
- Eutrophication
- Acidification
- Thermal loading
Tolerances to stress known.
Integrators of environmental
conditions.
Easy taxonomic ID; expertise
widespread.
Universal endpoirit.
Filed sampling is time consuming
arid expensive, with high spatial
variance and gear problems.
Intensively managed.
- Stocking
- Angling Impact of sampling
The only index which has been
developed (RFAI), has only been
tested regionally.
DO
Trophic state.
Toxicity bioassays.
Contaminant analysis.
pH, alkalinity measurement.
D-13
Table D-9. Advantages, disadvantages, and alternatives to using fish assemblages.
-------
Appendix D
Qualitative Screening—Widespread familiarity with
the condition of sport and forage fish in natural
resource agencies permits qualitative screening
assessment using expert knowledge of local and
state fisheries experts (USEPA 1989b). The intent
is to serve as a screening tool and to maximize the
use of existing knowledge of fish assemblages with
a questionnaire polling state fish biologists and
university ichthyologists believed knowledgeable
about the fish assemblages in lakes of concern.
Unlike field surveys, questionnaires can provide
information about tainting or fish tissue contami-
nation and historical trends and conditions.
Disadvantages of questionnaires include inaccu-
racy caused by hasty responses, a desire to report
conditions as better or worse than they are, and
insufficient knowledge.
Contaminants in Ftsh Tissue—Contaminant
concentrations in fish tissue have been monitored
to assess the extent of environmental contamina-
tion and to estimate risks to human health from
consuming fish. Contaminant concentration is an
excellent indicator of health risk, but it is not an
indicator of biological integrity.
Pathology—Pathological abnormalities (lesions,
tumors, growth anomalies) of fish are monitored
as overall indicators of environmental degrada-
Table D-10. Fish assemblage metrics under investig
Hickman and McDonough (1996).
tion, including effects of severe eutrophication,
sediment contamination, and acidification.
Significant rates of pathology typically occur
only in the most severely polluted habitats and
in populations of nonmigratoiy, bottom-feeding
fish. Pathology can be incorporated into
multimetric indices, such as IBI (Dionne and
Karr 1992).
Technical issues
The major problem in developing fish indices for
lakes is obtaining representative samples of fish
assemblages in lakes. Quantitative sampling in
lakes is not as reliable as in streams because of
lake morphology, bottom types, and variable
gear efficiency. Modification of IBI for lakes can
include use of relative abundances based on
subsamples from constant-effort sampling.
Sampling gear and protocols for different lake
habitats will need to be standardized.
Spatial and Temporal Variability—Fish are highly
mobile and respond rapidly to gradients in
physical habitat and water chemistry. They
actively avoid harmful conditions. Physical and
chemical parameters that affect fish spatial
distribution include:
lion by TVA. After Dycus and Meinert (1994) and
Metric
Optimal Condition
Impaired Condition
Species Richness and Composition
- No. oftaxa.
High
Reduced
- No. of Lepomis sunfish species.
High
Reduced
- No. of sucker species.
High
Reduced
- No. of intolerant species.
High
Reduced
- % tolerant individuals.
Low
High
- % dominance by one species.
Low
High
Trophic Composition
- No. of pisclvore species.
High
Reduced
- % omnivores.
Low
High
- % invertivores.
High
Low
Reproduction Composition
- No. of lipophilic spawning species.
High
Low
Abundance
- Total individuals.
Similar to reference.
Reduced
Fish Health
- % individuals with anomalies.
Low
Increased
D-14
-------
Biological Assemblages
• Habitat.
• Cover.
• Dissolved oxygen.
• pH.
• Temperature.
• Turbidity.
• Light.
Many fish seek specific habitats for activities
such as feeding, resting, and spawning. Their
movement between habitats is dependent on
time of day and season. Although fish popula-
tions are relatively stable compared to smaller,
shorter-lived plankton and benthos, fish mobil-
ity and behavior make fish difficult to sample.
Index Period—Sampling during the spring
coincides with optimal biological conditions and
may show recovery from environmental stress
periods. However, to avoid spring spawning,
sampling is usually conducted in late summer
and early fall. Seasonal changes in the relative
abundances of the fish assemblage occur
primarily during reproductive periods and (for
some species) the spring and fall migratory
periods. If fish sampling is required during this
period, then changes in relative abundance will
be important. Mid to late summer is often a time
of oxygen stress and should show the greatest
effects from environmental stress.
Sampling Gear— Obtaining both qualitative and
quantitative data on fish populations is limited
by gear selectivity and the fish mobility (USEPA
1992b). All sampling gear is selective. The
habitat or portion of habitat sampled and
efficiency of gear for a particular species in one
area does not necessarily apply to different
species nor to the same species in another area.
Temporal and spatial changes in relative abun-
dance of a species can be assessed under a
given set of conditions if those species are
readily collected with a particular kind of gear.
Electrofishing is the technique used most often
by agencies that monitor fish assemblages. The
EMAP Surface Water Northeast Lake Pilot
Survey found electrofishing the most effective
single-gear technique (USEPA 1994a, USEPA
1994b). The RFAI for TVA reservoirs includes
electrofishing as a collection technique
(Hickman and McDonough 1996). Other consid-
erations with respect to electrofishing are:
• Many agencies already have the equipment
available.
• Electrofishing is easy to use and produces
quick results.
• It is depth- and species-selective and does
not effectively sample catfish or any fish in
deeper water. It can be difficult to get close
to some fish (e.g., northern pike).
• It can be difficult to get equipment into
remote areas.
Seining, an active sampling technique, can be
used in the littoral areas (straight seines). Haul
seines and trawls are used in deeper open water
areas. Seining or trawling is not effective in
areas with bottom obstructions that can tear or
foul the net. Although the results are expressed
as number of fish captured per unit effort,
quantitative seining is very difficult. This
method is more useful in determining the variety
of fish rather than the number of fish inhabiting
the water.
Athough gill nets are a passive technique with
several disadvantages, they might be the most
appropriate gear type for sampling deep sublit-
toral habitats. Gill nets are size-selective,
depending on mesh size and do not obtain
representative samples of the total population.
They are most effective on lake herring, trout,
lake whitefish, yellow perch, walleyes, and
northern pike (USEPA 1992b). There is a high
mortality rate of fish caught in gill nets and
occasional mortality of nontarget species such
as turtles, muskrats, beavers, and diving
waterfowl. Trap and fyke nets are effective in
shallow areas. Like gill nets, they are also
passive and do not obtain a representative
sample of the total population. Meador et al.
(1993) and Weaver (1993) recommended a
multi-gear approach that takes advantage of
differences in gear selectivity and efficiency to
achieve a more accurate representation of the
fish assemblage structure. TVA uses shoreline
electrofishing for the shallow littoral zone and
experimental gill nets for the sublittoral/limnetic
zone (Hickman and McDonough 1996).
D-15
-------
Appendix D
Research Needs—TVA has been actively develop-
ing assessment tools for Its reservoirs for several
years. The move to a multimetric approach for
reservoir fish began In 1990. Successive steps in
this development process have brought contin-
ued Improvement to the REM. Potential im-
provements in the fish indices include using a
simple random sampling design rather than a
fixed station design to enhance statistical
validity with little increase in variability. Use of
the index in reservoirs or other river systems is
necessary to test its performance under a wider
range of conditions than is available in the
Tennessee River. Correlation with known
human-induced impacts remains a critical need
before general acceptance of the fish index as a
reliable method to address reservoir environ-
mental qualify.
A related issue is the effect of game fish man-
agement on IBI or other fish assemblage metric
scores. Nearly all lakes are stocked or have been
stocked in the past, and these practices can
affect the biological assemblages in a lake.
Stocking lakes with large piscivores is also used
in biomanipulation to improve water clarity of
eutrophic lakes (e.g., Hosper et al. 1992).
D-16
-------
Appendix E
Statistical Analysis Methods for
Biological Assessment
INTRODUCTION
A central premise of biological assessment is
comparison of the biological resources of a
waterbody to an expected reference condition
(Figure E-l). Impairment of the waterbody is
judged by its departure from the expected condi-
tion. This approach presumes that the purpose of
management is to prevent, identify, and subse-
quently repair anthropogenic damage to natural
resources. Biological assessment of waterbodies
is predicated on our ability to define, measure,
and compare an assessment endpoint between
similar systems. This guidance outlines analytic
methodologies to perform two tasks shown in
Figure E-l:
• Characterization of the biological expectation.
• Determining whether a site deviates from
that expectation.
All of the methods considered here use the same
general approach: sites are assessed by compar-
ing the assemblage of organisms found at a site
to an expectation derived from observations of
many relatively undisturbed reference sites
(Figure E-1) The expectations are modified by
classifying the reference sites to account for
natural variability, and each assessment site is
classified using non-biological (physical, chemi-
cal, geographic) information. Biological vari-
ables are tested for response to stressors by
comparison of reference unimpaired sites and
known impaired sites. A set of "rules" are
developed from this information, which are
then used to determine if the biota of a site
deviate from the expectation, indicating that
the site is impaired.
Several analytic methods have been developed to
assess the condition of water resources from
biological data, beginning with the saprobien
system in the early 20th century to present-day
development of biological markers. This appen-
dix outlines three methods for analyzing and
assessing waterbody condition from assemblage
and community-level biological information:
1. Multimetric assessment using an index
that is the sum of several metrics. This is
the basis of the Index of Biotic Integrity
(IBI) (Karr et al. 1986), the Invertebrate
Community Index (ICI) (Ohio EPA 1990):
the Rapid Bioassessment Protocol (USEPA
1989b); and state indices developed from
these (e.g., Southerland and Stribling 1995,
Barbour et al. 1996a, Barbour et al. 1996b).
2. Multimetric assessment using an index
that is developed from a multivariate
discriminant model to discriminate refer-
ence from impaired sites. This is the basis
E-1
-------
Appendix E
4-
CM 3-
H2
sua a
m 'Biological Expectation" ||
V Values at undisturbed site^J
1-
+ Siteb
0 12 3 4
Biological Indicator t
Figure E-1 Graphical representation of
bloassossment. Assessment sites a and b are
compared to an ideal biological expectation. Site
a is near to its expectation; Site b deviates from it
and is considered to be impaired.
of stream bioasscssment in Maine (Davies et
al. 1993), and of the estuarine invertebrate
indices developed by the EMAP-Estuaries
program (USEPA 1993e, USEPA
1994h,USEPA 1994i, Engle et al.1994).
3. Assessment using multivariate ordination of
species abundances. This methodology has
been used widely in assessment of streams
in Britain (e.g., Wright et al. 1984); assess-
ment of marine macroinvertebrates in the
North Sea (e.g., Warwick and Clarke 1991);
and in assessment of benthic
macroinvertebrates in the Great Lakes (e.g.,
Reynoldson et al. 1995).
Many other methods are possible, as well as
permutations of the three methods above, all of
which are beyond the scope of this document.
The three approaches were selected because:
• They use community and assemblage data.
• The methods are not restricted to any one
assemblage. The examples all use benthic
macroinvertebrates and fish (freshwater and
estuarine), but any other assemblage could
also be used, such as phytoplankton,
zooplankton or macrophytes.
• The methods are general, and have been
used by many agencies in many areas. Hie
examples used to illustrate the methods
have also been carried out over wide geo-
graphic areas with many sites, demonstrat-
ing the generality of the methods.
* The examples used to illustrate the methods
are concise. Methods were fully docu-
mented, and have been carried to comple-
tion, that is, assessment of biological
impairment and non-impairment.
The optimal analysis methodology should also
be cost-effective and easy to communicate to
managers and the public. Both the multimetric
index and the discriminant model index (ap-
proaches 1 and 2) are easy to apply in a con-
tinuing operational monitoring program because
data from an individual site are entered into a
formula, and the site's deviation from reference
conditions can be known immediately (Gerritsen
1995). The ordination approach (3) requires
reanalysis of a the entire reference data set for
each new batch of monitoring sites. The
multimetric index (approach 1) is the easiest to
explain to managers and the public because it
does not rely on specialized concepts such as
multivariate statistics. The ordination approach
(3) may be most cost-effective if the biological
survey is a single event—a large number of sites
are surveyed once, and there is no plan to
continue monitoring or to survey new sites.
Characterization off Reference
Conditions
Reference conditions establish the basis for
comparison and for detecting impairment of
waterbodies. They should be applicable to an
individual waterbody, such as a stream or lake
and also to similar waterbodies on a regional
scale (USEPA 1996a).
Classification Tools
The objective of classification is to group similar
waterbodies together, so that reference condi-
tions will reflect reasonable expectations for
assessing waterbodies. There are two funda-
mental approaches to classifications; a priori or
rule based, where known rules are applied to
classifying objects; and a posteriori, or data-
based, where rules for classifying objects are
derived from data obtained from the objects
(waterbodies) themselves (Conquest et al. 1994).
E-2
-------
Biological Assemblages
For example, a rule-based classification may
divide mountain and lowland streams by
elevation or stream gradient. An a posteriori
classification would examine data from all
streams, and determine if there is a basis for
separating them into two or more classes (not
necessarily including elevation or gradient). The
a posteriori approach requires a relatively large
sample of reference sites to derive the classes
and rules, with both biological and physical-
chemical data from each site.
The basic assumption of classification is that
physical habitat and water quality largely
determine the composition of biological commu-
nities in waterbodies. Therefore, if waterbodies
are classified adequately, reference biological
community types should correspond to the
classification. Classification is often an iterative
process of refining the classification scheme as
new data are obtained, until a satisfactory
classification emerges that accounts for varia-
tion in the reference site biological data.
Several statistical tools can assist in site classifi-
cation, but there is no set procedure. If a priori
classification is based on well-developed prior
knowledge, then graphical analysis of biological
data, followed by any necessary modifications
and tests of the resultant classification, may be
sufficient.
If a rule-based classification is not self-evident,
then it may be necessary to develop an alterna-
tive classification from the data using one or
more analytical classification approaches.
These methods include several cluster analysis
methods, and several approaches to ordination
analysis, including principal components
analysis (PCA), correspondence analysis (CA)
and its variants, and non-metric multidimen-
sional scaling (NMDS).
In statistical terminology, each site is a sample
unit (SU) (Ludwig and Reynolds 1988). Ideally,
sample units should be independent, which is
generally achievable in small streams and in
many lakes, where each waterbody can be a
separate sample unit. Large lakes and reser-
voirs, large rivers, and estuaries may include
several sample units within the same waterbody.
For large and complex lakes and estuaries, it
may be necessary to define a site as a contigu-
ous basin or embayment. Any portion of the
waterbody that is partially isolated from the rest
by bottom topography or water motion should
be considered a separate site and sampled
accordingly. This also applies to the three zones
of large reservoirs (riverine, transition, and
forebay) and to salinity zones of estuaries (e.g.,
fresh, mesohaline polyhaline), which have
different biological communities and dynamics
even though they are not hydrologically isolated
(Thornton 1990b). Thus, large waterbodies
(including large reservoirs) may comprise several
sites or SUs. Sites (SUs) are considered inde-
pendent and are kept separate in analysis; no
"average" is estimated for a multiple-site
waterbody. Multiple sites are not strictly
independent and will need to be considered
carefully in reference condition characterization
and in metric response evaluation.
Large rivers may be more problematic in that
sites on a river are serially linked by water flow.
Sites are defined as river reaches of some
minimum length that exhibit some (but not
complete) independence. Sample units (reaches)
may be defined by length (e.g., a set length or a
multiple of stream widths), as the reach between
major tributaries, or as segments downstream of
major impacts and discharges (e.g., urban
areas).
Graphical Analysis
A key graphical display is box-and-whisker plots
(Figure E-2). These show population attributes
of the data: central tendency, spread, and
outliers. In the display used here, the central
point is the median value of the variable; the
box shows the 25th and 75th percentiles
(interquartile range); and the whiskers show
values within the inner fences (Figure E-2).
Points beyond the fences may be considered
outliers or extreme values. Box-and-whisker
plots are simple, straightforward, powerful, and
the interquartile ranges are used to evaluate
whether there is a real difference between two
areas and whether a metric is a good candidate
for use in assessment. Graphing the data
should always be a first step in data analysis.
Statistical methods used by biologists are
frequently tests of whether two or more popula-
tions have different means using t-tests, analy-
sis of variance, or various nonparametric
methods. However, the fundamental problem of
biological assessment is not to determine
whether two populations (or samples) have a
different mean, but to determine whether an
E-3
-------
Appendix E
*
extreme value
I
C
3
outlier
Inner fenca
whisker; largest value within Inner fence
1
75th psroenfllB (upper quarttle)
i
1
1
median (50th percentile)
25th percentile (lower quarffle)
Inner fenca
Figure E-2. Box and whisker diagram (after Tukey
1977), The box is the interquartile range (25th -
75th percentile). Inner fences are the quartiles ±
1,5 x interquartile range; outer fence is 3 x
interquartile range. Ends of whiskers are the most
extreme observations within the inner fences.
individual site is a member of the least-impaired
reference population (Figure E-l). If it is not,
then a second question is how far it has devi-
ated from that reference. Therefore, biological
assessment requires the entire distribution of a
metric, which is effectively displayed with a box-
and-whisker plot.
Ordination Methods
The purpose of ordination analysis is to reduce
the complexity of many variables (for example,
abundance of 200 species from 50 sites) into
fewer variables, such that the sites and the
species are ordered on the new variables (Figure
E- 3). The new variables are called the principal
axes of the analysis; the first axis accounts for
the most variation in the original data, the
second accounts for somewhat less variation,
and so on. Typically, only the first two to four
axes of the analysis are presented because higher
axes contribute little to the variance explained
and because one cannot present or conceptualize
more than three axes simultaneously.
Principal Components Analysts—One of the most
commonly used ordinations is principal compo-
nents analysis (PCA). In PCA, the new variables
(principal axes) are linear combinations of the
original data; that is, the relationship between
each principal axis and the abundance of each
species can be expressed as a straight line, as in
simple linear regression (Jongman et al. 1987).
Thus, PCA is a multivariate extension of linear
regression (Figure E-3), making the assumption
that a variable will have a maximum value at
one end of a principal axis and minimum value
at the other. Because the principal axes can be
seen as environmental gradients to which the
species respond, ordination is also called
gradient analysis (Jongman et al. 1987).
The procedure of PCA is an eigenanalysis of the
correlation matrix among variables in the
original data matrix. The variables may be
species abundance, calculated assemblage
metrics, or environmental (chemical and habitat)
variables. Eigenanalysis results in as many
eigenvalues as there are rows (or columns) in
the correlation matrix, and each eigenvalue and
corresponding eigenvector describes an axis of
the ordination. The eigenvalue of an axis is the
variance accounted for by that axis. Often, only
the first two or three axes explain significantly
more variance in the original data than a
random axis. Rules for determining the number
of significant axes are explained in Jackson
(1993b). Details of formulas and calculations
for PCA, as well as variations of PCA, are in
Ludwig and Reynolds (1988).
CM
0
o°
Variable 1
Figure E-3. Ordination. The relationship of
species 1 and species 2 can be described by
translating and rotating the axes, so that most of
the variance is on the first axis. In this 2-
dimensional example, the observations have been
reduced to a single dimension, the first axis, which
is a linear combination of species 1 and species 2.
E-4
-------
Biological Assemblages
Because PCA is linear, and assumes multivariate
normal distributions, data transformations are
often necessary. Species abundance data usually
have many zeros in the data matrix, and no
transformation will normalize them. PCA is not
useful for species abundance data, although it
can be made to work well for data that are normal
or can be transformed to a normal distribution
(e.g., environmental variables, assemblage at-
tributes such as number of taxa, etc.).
Correspondence Analysis Family—A problem
with linear ordinations such as PCA is that
species do not always respond linearly to
gradients; in fact, a unimodal response to
environmental gradients is much more common
(Jongman et al. 1987). A unimodal response is
one in which a species has peak abundances at
certain optimal values of an environmental
variable (for example, pH or nutrient concentra-
tion) and abundances are lower at both higher
and lower values of the environmental variable.
There are many examples of environmental
optima for aquatic organisms; optima are
supported by uptake kinetics, and they form the
basis for resource-based competition and
seasonal succession (e.g., Hlman 1982).
Multivariate ordination based on unimodal
responses to environmental gradients is called
correspondence analysis. As in PCA, correspon-
dence analysis also seeks new variables to
explain the species abundances on fewer axes
and is frequently "detrended" to eliminate a
mathematical artifact from its calculation
(Jongman et al. 1987).
Ordination can also be done to develop associa-
tions between the species abundances and
measured environmental variables. In this case,
both species abundance and the environmental
variables are related to the principal axes and
the whole procedure can be regarded as a
multivariate, multiple regression. The linear
form is called canonical correlation (CC); the
unimodal form is called canonical correspon-
dence analysis (CCA). Because it assumes
unimodal responses, CCA is thought to be a
realistic and robust multivariate ordination (ter
Braak 1986, Palmer 1993).
In CCA, each species, site, and environmental
variable has a score on each of the principal
axes. Results of CCA are presented graphically
by plotting the scores on two of the axes (usually
the first two) (Figure E-4). Plotting site scores
with environmental
variable scores shows the
relationship between the
sites and the environ-
mental variables and can
also show clustering of
sites.
Nonmetric Multidimen-
sional Scaling—
Nonmetric multidimen-
sional scaling (NMDS) is
increasing in use in
ecological application
because it offers several
advantages over other
ordination methods.
Because the ordination
works on a matrix of
distance ranks, it is
distribution-free and
hence unaffected by non-
normality and
nonlinearity in the data
(Ludwig and Reynolds
1988). It is robust and
produces interpretable
1.0 -
Dam
0.5 -
Ca
0.0 -
a a
E Anaconda
• Cabinet
a Selway-Bitterroot
-0.5 -
*
-1.0
¦1
¦2
1
0
2
First Axis
Figure E-4. Canonical corresponence analysis of periphytic diatom
assemblages from Rocky Mountain lakes. Site scores (points) and
environmental variables (arrows) on the first two axes. Points within ovals
are lakes with dams at their outlet; single point inside diamond is dammed
by a natural glacial moraine.
E-5
-------
Appendix E
ordinations from different ecological data sets.
The disadvantages of NMDS are that it is
iterative and subject to local minima (SYSTAT
1992) and that no canonical form has yet been
developed. It is possible, however, to estimate
correlations of environmental (explanatory)
variables with the axes of NMDS.
Like cluster analysis, NMDS uses a distance
metric among sample units (sites), and results
can be sensitive to the choice of the distance
metric (Jackson 1993a). Bray-Curtis distance
and the relative distance metrics (relative
Euclidean distance and chord distance) tend to
work best (Kenkel and Orloci 1986, Ludwig and
Reynolds 1988).
The objective of NMDS is to obtain a "best fit"
between the dissimilarity measures and the
distances calculated in ordination space. The
dissimilarities have as many dimensions as
there are sites, but the ordination reduces these
to a smaller number, usually 2 or 3. The
procedure is to rank the distances in the simi-
larity matrix from smallest to largest, then to
calculate an initial starting ordination (termed
the initial configuration) directly from the
dissimilarity matrix. Intersite distances are
calculated from the initial configuration, ranked,
and compared to the ranked dissimilarities. A
best solution is sought iteratively, changing the
configuration so that the two rankings (dissimi-
larities and configuration) become more similar.
Goodness of fit of the configuration to the
dissimilarities is measured by Kruskal's stress
coefficient (Ludwig and Reynolds 1988) or
Guttman's coefficient of alienation (SYSTAT
1992). Iterations stop when stress or alienation
reaches a minimum value.
NMDS is available on many commercial statisti-
cal software packages. Distance measures used
by ecologists, especially Bray-Curtis distance
and chord distance, are not usually available in
these packages and must be calculated sepa-
rately. Relative Euclidian distance is also only
rarely available; however, if an input matrix of
percent abundances of species is used, then
Euclidean distance will yield relative Euclidean
distance.
Results from NMDS are a final configuration,
consisting of coordinates for each site in the 2 or
3 dimensional ordination. As in other ordina-
tions, points close to each other In the ordina-
tion space (Figures E-3 and E-4) represent sites
with similar species composition.
Classification Analysis—Classification, or the
placement of objects into categories, is an innate
human activity. A wide variety of formal classifi-
cation procedures have been developed (see
Gauch 1982 for a review). Only two will be
discussed here, cluster analysis and two-way
indicator species analysis (TWINSPAN).
Cluster Analysis—Cluster analysis is known as
an agglomeratlve classification, that is, It
successively builds clusters until all objects
have been joined in a single cluster. Cluster
analysis begins with a matrix of intersite dis-
similarities. The smallest dissimilarity in the
matrix Is selected and those two sites are joined
in a cluster. The algorithm then calculates the
dissimilarities between the new cluster and all
other sites or clusters. Again, the smallest
dissimilarity is selected, the two objects are
joined, and the process repeats itself until all
objects are joined. Results can be shown in a
dendrogram (Figure E-5), where the bars con-
necting clusters represent the dissimilarity
between them. Final clusters are identified by
choosing a cutoff dissimilarity value. The cutoff
dissimilarity value clearly affects the number of
clusters (Figure E-5): it may range from one to
the number of sites. The number of clusters
should be small, and should explain as much
variance of the biological data as possible.
Classification with cluster analysis is not as
straightforward and objective as is implied by a
dendrogram produced by a mathematical algo-
rithm. First, several algorithms may be used for
recalculating dissimilarities among agglomerated
clusters of sites, and each algorithm may pro-
duce different results. A favored algorithm for
ecological data is the unweighted pair-group
method (UPGMA) (Ludwig and Reynolds 1988,
Reynoldson et al. 1995). Second, the dissimilarity
measure affects results. As in NMDS analysis,
relative dissimilarity measures (relative Euclid-
ean, chord distance) and Bray-Curtis distance
work best for species-abundance data (Ludwig
and Reynolds 1988). Finally, as noted above,
selection of a distance cut point for defining
clusters is subjective (Figure E-5).
Two-way Indicator Species Analysis
(TWINSPAN)—TWINSPAN was developed by Hill
(1979), and is a divisive technique. Instead of
E-6
-------
Biological Assemblages
Distance 0 0.5 1.0 1.5
i i 1
SAMP72 j~| | |
SAMP30
SAMP209
SAMP204
SAMP208
SAMP206
SAMP6
SAMP1
l l L
a b c
Figure E-S. Dendrogram from cluster analysis.
Outpoints a, b, c are at distances 0.7S, 1.0,1.25,
respectively, and result in 5,4, and 2 clusters,
respectively.
building up clusters from individual sites,
divisive methods start with the entire data set
and divide it into two. The division process Is
repeated until a specified number of clusters are
obtained (Gauch 1982). TWINSPAN first ordi-
nates the data, then divides the sample into two
clusters near the middle of the first ordination
axis. Ordination is by reciprocal averaging,
which is a variation of correspondence analysis.
New ordinations are repeated on each daughter
cluster, and the daughters are in turn divided
on their first ordination axis. TWINSPAN is only
available in specialized software packages.
Discriminant Model—The objective of a discrimi-
nant model is to predict community type, or
community composition, from non-biological
data. Development of such a model requires a
data set with both biological and non-biological
data, and testing of the model requires a sec-
ond, similar data set. Discriminant analysis is
best illustrated with a simple example (e.g.,
Ludwig and Reynolds 1988, Johnson and
Wichern 1992). Suppose that abundances of
two species are examined in riffle and pool sites
of streams (Figure E-6) and we wish to develop a
model that will discriminate between riffle and
pool sites, using only the biological data. As
shown in the figure, pool sites tend to have
greater abundances of both species. Using
either species alone to form the rule would lead
to frequent errors. Discriminant analysis finds
a best fit straight line to separate the groups;
the heavy line of Figure 6 is the border and the
hatched line perpendicular to it is the discrimi-
nant function. Sites with positive scores are
more likely to be pools, and sites with negative
scores are more likely to be riffles.
Discriminant function analysis involves compu-
tation of a pooled variance-covariance matrix of
the groups, and solving for the coefficients of the
discriminant function. Formulae and computa-
tions are shown in Ludwig and Reynolds (1988),
Johnson and Wichern (1992), Pielou (1977), and
other multivariate statistics textbooks. Dis-
criminant analysis also allows calculation of
multivariate distance (Mahalanobis Dz) between
groups, and an F-test for group differences. A
limitation of discriminant function analysis is
that it is linear: strong nonlinearity of the data
will reduce its power to separate groups.
Rule-Based Classification;
Characterization of Reference
Conditions
The objective of reference characterization is to
describe (characterize) each of the reference
classes in terms of biological indicators and
other descriptive variables. The first step is to
support or reject the a priori classification,
followed by modifying it to arrive at a parsimoni-
ous and robust classification; that is, one with
the fewest classes that explains the most
variance in the reference data set.
There is no single "best" classification nor are
resources generally available to determine all
possible differences between all waterbodies in a
region. The key to classification is practicality
within the region or state in which it will be
applied; local conditions determine the classes.
Classification will depend on regional experts
familiar with the range of conditions in a region
as well as biological similarities and differences
among waterbodies. Ultimately, classification
can be used to develop a predictive model of
those chemical and physical characteristics that
affect the values of the biological metrics and
indices in reference sites.
A useful classification scheme is hierarchical,
beginning at the highest (regional) level and
stratifying as far down as necessary (Conquest
et al. 1994). The procedure is to classify
waterbodies according to region and then to
increase the stratification in the classification
hierarchy to a reasonable point for the given
region. Although several classification levels are
possible, in practice, only one, or at most two,
E-7
-------
Appendix E
Discriminant
function
Species A
O Riffle
¦ Poo!
Figure E-6. Illustration of discriminant function
analysis. Neither species A nor species B can be
used alone to distinguish riffle from pool sites.
Discriminant analysis estimates a linear border
between the two site classes (heavy line), and a
discriminant function (graduated line). The
discriminant function is a linear combination of
the input variables (species A and species B), and
yields a probability that a site belongs to the riffle
or pool class.
relevant levels would typically be used. Classifi-
cation should avoid a proliferation of classes
that do not contribute to assessment. One or
two relevant levels of the hierarchy will yield the
best classification scheme. Potential hierarchical
classifications for streams, lakes, and estuaries,
respectively, are given in Gerritsen
(1995),USEPA (1996a), and USEPA (1997a).
Confirmation of a priori
Classification
Uniuartate Tests—Univariate tests of classifica-
tions include all the standard statistical tests for
comparing two or more groups: t-test, analysis
of variance, sign test, Wilcoxon rank test, and
Mann-Whitney U-test (USEPA 1996b, Ludwig
and Reynolds 1988). These methods are used to
test for significant differences between groups
(classes) to confirm or reject the classes. They
are univariate, with a single dependent (re-
sponse) variable. Biological variables (metrics)
may require transformation to meet assump-
tions of t-tests and ANOVA, or non-parametric
tests (e.g., rank tests, Mann-Whitney) may be
used. See USEPA (1996b) for discussions on
the use of these and other univariate tests for
biocriteria. Failure to confirm the classification
for any single response variable does not mean
that it will fail for other response variables.
Because assessment is based on multiple
variables (metrics or species composition),
multivariate tests might be more convenient
than a succession of individual tests.
Discriminant Analysis—Discriminant analysis
can be used as a form of multivariate, one-way
analysis of variance that tests differences
between a set of groups based on several
response variables. It is used as a test of
classifications (Conquest et al. 1994), provided
that the assumptions of linearity and normality
are met. Many statistical software packages
provide discriminant analysis.
Gradients—On occasion, environmental gradi-
ents might not allow formation of discrete site
classes. For example, the number of zooplank-
ton taxa in lakes is usually related to lake size
(e.g., Dodson 1992). Similarly, fish and inverte-
brate number of taxa in streams is typically
related to stream size (order, discharge or
watershed area) (e.g., Ohio EPA 1987, DeShoji
1995)
Ordination—The a priori classification may also
be confirmed with one of the ordination meth-
ods. Sites are plotted in ordination space using
different symbols for the a priori classes. If
classes overlap completely in ordination space,
then there is no apparent difference in their
species composition (or other variables used in
the ordination), and it may be appropriate to
aggregate the coinciding classes. Species or
variable scores can be plotted in ordination
space to determine which contribute most to
separation among classes. Correlation coeffi-
cients of environmental variables with the site
scores will show if there are environmental
gradients that are associated with the ordination
and with the site classes. Examples and de-
tailed methods for ordinations are given in
Jongman et al. (1987) and Ludwig and Reynolds
(1988).
A Posteriori Classification
This method of classification determines classes
from the structure of the data, rather than from
pre-existing knowledge or hypotheses. Because
E-8
-------
Biological Assemblages
the principal goal of classification in biocriteria
programs is to account for biological variation,
the biological data (typically species composition
data) are used for classifying. As with a priori
methods, only data from reference sites are used
to develop the classification (e.g., Moss et al. 1987,
Wright et al. 1984, Reynoldson et al. 1995).
Test sites must also be assigned to appropriate
classes, so that they can be compared to refer-
ence sites. Because anthropogenic degradation
affects the biota of the waterbodies, assigning
test sites to classes using their biological data
may lead to incorrect classification (Figure E- 7).
Therefore, the classification also requires a
method to assign test sites to classes, using
non-biological measures that are not affected by
anthropogenic degradation. Following an a
posteriori classification, this is typically a
discriminant function model that is constructed
from the reference data set (Norris 1995).
Identifying Classes
Classification is a subjective activity even when it
is done with seemingly objective quantitative
methods. The subjectivity is due in part to the
information that will be used to decide if objects
are similar or not, and in part to the methods
and their variations that will be used to classify
the objects. For example, we may say that Miami
is similar to Havana. We may also say that,
during the Cold War era, Havana was similar to
Moscow. Does it then follow that Miami is similar
to Moscow (SYSTAT 1992)? This example illus-
trates that the variables used to determine
similarity (climate, economic system) profoundly
affect the resultant classification.
There are several different quantitative methods
to classify objects, each of which may result in
different classification. Furthermore, each
classification method requires subjective deci-
sions on the similarity measure to use in the
classification and on the number of classes to
identify. Thus, classification remains subjective,
even when done with seemingly quantitative
algorithms. Classifications developed from
biological data should make sense In the physi-
cal and chemical context of the habitats. A
posteriori classification is developed from the
biological data set. Species abundance data are
examined, and groups of sites are identified that
are similar to each other. Usually, this is done
with a similarity (or dissimilarity) measure and a
form of cluster analysis. Subjective decisions
Species 2
/'Assemblage
VType ky
\ /Assemblage
Test Site X
Species 1
Figure E-7. Misclassification of test sites. A test
site (X) that was originally in assemblage Type A
has been degraded (arrow). If biological data are
used to classify the test site, then it would be
classified as Type B because it is now more similar
to Type B. If, on the other hand, non-biological
measures that are not affected by degradation are
used to classify the test site, then it would be
correctly identified as Type A and the degree of
biological degradation could be assessed.
are required to select the classification method-
ology, the similarity measure, and the number
of groups to identify.
As was stated above, the general objective of
classification is parsimony of classes (few classes)
to obtain a large partitioning of variance among
the classes. Too few classes results in large
variability within each class, and too many classes
results in trivial differences among classes.
Assigning Test Sites to Classes
After reference site classes have been deter-
mined, using cluster analysis or some other a
posteriori classification a model is developed to
enable test sites to be assigned to one of the
reference classes. This is typically a discrimi-
nant model developed from non-biological data
of the reference sites. Data for the discriminant
model should be measurements that are not
affected by anthropogenic degradation, such as
stream gradient, sinuosity, natural water
chemistry, lake depth, watershed soil type, etc.
(Norris 1995). The output of a discriminant
model is a discriminant function that assigns
sites to one of the classes. It is developed from
reference site data, and should be tested with an
independent reference site data set.
E-9
-------
Appendix E
Multimetrlc index Method
The indices currently used are variations of the
Index of Biotic Integrity (IBI) for fish assem-
blages in streams, developed by Karr and his co-
workers (e.g., Karr 1981, Karr et al. 1986). The
concept was extended to benthic invertebrate
assemblages (Ohio EPA 1987, USEPA 1989b,
Barbour et al. 1992, Kerans and Karr 1994).
Each index is the sum of several (up to 12)
standardized component metric scores. Metric
scores are usually on an ordinal scale of 1 to 5
(Karr et al. 1986), or 0 to 6 (USEPA 1989b) or as
a percentage of the reference metric value
(Maxted et al. 1994). Component metrics consist
of measures such as total number of taxa,
percent abundance of the dominant taxon,
number of species and percent abundance of
intolerant groups, and percent abundance of
functional feeding groups such as planktivorous
fish or invertebrate shredders.
Metric Variability
Metrics that are too highly variable within the
reference sites are unlikely to be effective for
assessment. Relative variability is often mea-
sured with the coefficient of variability, defined
as the standard deviation divided by the mean
(expressed as percent):
ev = =xioo
X
The CV is a measure of how large the variability
is compared to the mean. Ideally the CV should
be small, which can be achieved with a small
variance or with a large mean value. However,
some metrics might have low values under
reference conditions (e.g., number of exotic
species), and CV will always be large for such
metrics. For example, if a sample of 10 refer-
ence sites, each with 10 taxa, includes a single
site with a single exotic species, then the CV of
the number of exotic species is over 300 per-
cent. Furthermore, the multimetric approach
calls for comparison of metric values to a
percentile of the reference population values and
is thus a distribution-free approach. Because
the CV is the ratio of the sample standard
deviation to the mean, it might not adequately
express variability for non-normal distributions.
An alternative measure to the CV is the
"interquartile coefficient," which is based on
quartiles of the reference distribution and the
expected change of the metric rather than its
parameters (Gerritsen and Bowman 1994). In
operational bioassessment, metric values below
the lower quartile of reference conditions are
typically judged as not meeting reference
expectations (e.g., Ohio EPA 1990). The range
from 0 to the lower quartile can be termed a
"scope for detection." For those metrics with low
values under reference conditions and high
values under impaired conditions, the scope for
detection is the range from the 75th percentile
to the maximum possible value (e.g., 100
percent) (Figure E- 8).
The larger the scope for detection, compared to
the interquartile range, the easier it will be to
detect deviation from the reference condition.
The "interquartile coefficient" is thus defined
here as the ratio of the interquartile range to the
scope for detection:
'19
_IQ
Dc
where IQ
Ds=<
interquartile range
25th percentile (for metrics that
decrease with impairment); or
maximum possible value - 75th
percentile (for metrics that
increase with impairment
The interquartile coefficient is analogous to the
CV and is used similarly, but it is bidirectional
and is calculated from percentiles in the same
way that assessment uses percentiles. In
general, an interquartile coefficient greater than
1 indicates excessive variability of a metric.
Metric Response
Response of metrics to stresses is evaluated by
comparison of reference sites to test sites. The
simplest comparison is using box-and-whisker
plots of the metric distribution in reference and
test sites (Figure E-8) or by univariate tests of
metrics in reference and test sites. Alterna-
tively, it may be possible to develop an empiri-
cal model of metric response to stressors.
E-10
-------
Biological Assemblages
a.
Ill
3
£
o
DC
I
~
Interquartile
range
scope for
detecting
impairment
X
o
p
X
Reference
Sites
Impaired
Sites
Mixed
Non-Reference
(Impaired and
unfr
b.
lu
=>
_l
$
o
cc
tn
2
X
~
scope for
detecting
impairment
Interquartile
range
T
T
Reference
Sites
Impaired
Sites
Mixed
Non-Reference
(impaired and
unimpaired)
Figure E- 8. Assessing candidate metrics that have (a) high
values under reference conditions, and (b) low values under
reference conditions.
variance in the test sites should be
larger than that in the reference
sites (Figure E-8). If possible, it is
advisable to separate test sites
according to the stressors or types
of impairment (e.g., habitat degra-
dation, toxic substances, organic
enrichment) so that response to
each stressor can be determined.
When selecting metrics, it is
important to visually examine the
distribution of metrics in reference
sites and in impacted sites.
Metrics are selected for inclusion
based on their responsiveness,
typically by visual examination of
box and whisker plots (e.g., Fig. E-
8} or scatterplots (Barbour et al.
1996a, Fore et al. 1996). If there
is no overlap of the data points, or
if the overlap is restricted only to
the whiskers of the box plots, then
the metric responds strongly to the
impairment, A strong response
here implies that at least 75% of
affected sites have no overlap with
at least 50% of the reference sites.
A minimum response strength
might be defined as no overlap of
the median of one site type with
the quartile of the other; implying
that at least 50% of affected sites
are below the 25th percentile of
reference sites.
Several approaches are available including
multiple regression, canonical correlation,
canonical correspondence analysis, and log-
linear models (Ludwig and Reynolds 1988,
Jongman et al. 1987).
Metrics are judged responsive if there are
significant differences in central tendency or in
variance between reference and test sites (Figure
E- 8). If the test sites are known to be affected
by anthropogenic pollution or disturbance, then
mean or median values of responsive metrics
should be substantially different between
reference and test sites (Figure E-8). If the test
sites simply do not meet reference criteria (i.e.,
they might be a mix of impaired and unimpaired
sites, or sites with different stressors), then the
Many biologists may be tempted to
use statistical significance tests to
select metrics, but slavish reliance on signifi-
cance tests does not contribute to biological
understanding (Yoccoz 1991) and may weaken a
multimetric index. If sample size is small (say,
n = 6 in both reference and impact sites), then
significance tests (at oc = 0.05) will have low
power and responsive metrics may be rejected.
On the other hand, if sample size is large (say,
n = 30 in both site categories), then it would be
possible to detect a statistically significant
difference that is biologically meaningless. In
this case, metrics that do not contribute to
meaningful assessment could be selected,
simply because statistical significance was
detected. A better measure is the expected
frequency with which a metric will fall below a
threshold to register impairment. Frequency
E-11
-------
Appendix E
can be estimated with a box and whisker plot,
but not with a significance test. For example, if
the median of impaired sites is below the quar-
tile of reference sites (Figure E-4), then we
estimate that impaired test sites will be below
the reference quartile in at least 50% of all
observations.
Metrics that are responsive to known or un-
known stresses are retained for index develop-
ment, Finally, responsive metrics are evaluated
for redundancy, where redundancy means a
tight correlation (r>0.9) and a linear relation-
ship. A metric that is linearly correlated with
another might not contribute new information to
the assessment. Pairs of metrics with correla-
tion coefficients greater than 0.9 should be
examined carefully to determine whether they
are linear and if both metrics are necessary.
Often, strongly correlated metrics are calculated
from the same raw data, or their method of
calculation ensures correlation. For example,
Shannon-Wiener diversity and percent abun-
dance of the dominant taxon are linearly corre-
lated in any data set. A scatterplot of the
strongly (>0.9) correlated metrics should be
examined; if there is an apparent nonlinear or
curved relationship, then both should be
retained. If all the points fall very close to a
straight line, then one of the metrics can be
safely eliminated.
Mtalllmetrlc Index
Development
likewise compared to index values at reference
sites. Index values at reference sites are then
used to establish biocriteria. Socio-political
decisions must then determine the numerical
values of biocriteria corresponding to aquatic life
use categories.
Metric Scoring
Several methods may be used for scoring
metrics, all of which are based on the metric
distribution in reference sites. Metrics may be
given ordinal scores (most often 1, 3, or 5);
corresponding to impaired, intermediate, or
unimpaired biota, respectively, or may be given
a score which is the metric's percentage of the
reference value (Figure E-9).
All of these require comparison to some measure
of the reference value distribution: an upper
percentile, a lower percentile, or a central
tendency (Figure E-9). Although a central
tendency of the reference sites (e.g., the mean
value) may be Intuitively attractive as a basis of
comparison, there are two important reasons for
using percentiles instead:
• An assessment methodology must be able to
take into account natural variability of
ecological systems. We know that aquatic
biota may differ from riffle to riffle in the same
unimpacted stream. Central tendency does
not take into account the natural variability,
and scoring criteria based on central tendency
Multimetric indices are typi-
cally developed by summing
the metrics that proved respon-
sive to disturbance. The first
step is to standardize the
different numerical scales of
metrics (e.g„ number of taxa;
% of individuals that are
predators) into unitless scores
(e.g., Karr et al, 1986,
Gerritsen 1995). The scores
may be ordinal, or they may be
a percentage of a reference
value. Ordinal scores are more
commonly used, and corre-
spond to categories such as
"impaired" and "unimpaired."
The index is the sum (or mean)
of the metric scores, and is
Ordinal Percentage Ordinal Percentage
Score Score Score Score
Score
CJ
- 95 %ile-
• 100
median
mean—
T
25%lle-
SO
100
50
100
SO
Reference
distribution
Upper
percentile
Lower
percentile
Central
tendency
Scoring Methods
Figure E-9. Illustration of alternative scoring methods, using an
upper percentile, a lower percentile, or a central tendency. Most
common score breakdowns (5-3-1 ordinal, or percentage) are shown
for each, but other ordinal scores have also been used (e.g., 6-4-2-0).
E-12
-------
Biological Assemblages
may result to lowered scores for many sites
that are within the expected variability of
natural, undisturbed sites.
• A second disadvantage of central tendency
measures occurs when reference sites, upon
which the reference condition is empirically
based, are known to be affected to some
degree by human activities. Reference sites
are often selected to be the least
anthropogenically affected in a region, but
may still be subject to regional and wide-
ranging impacts (e.g., USEPA 1996a).
Examples include estuaries and large rivers
which receive inputs from their entire
watersheds, and small streams and lakes in
extensively altered agricultural ecoregions
(e.g., the Cornbelt Plains). Use of central
tendency then reflects the general (and
unquantifiable) degradation of the region
and will not result in reference conditions
that represent the biological potential.
Two approaches are used to develop metric
expectations and scoring criteria (Simon and
Lyons 1995). The first approach uses defined
reference sites that meet criteria for representa-
tive reference sites. Data from the reference
sites are used to define expectations and de-
velop metric scoring criteria (Simon and Lyons
1995). The principal scoring criterion (between
meeting and not meeting reference expectations)
is typically based on a lower percentile of the
reference distribution; for example, the 25th
percentile (Ohio EPA 1990, Barbour et al.
1996a, Barbour et al. 1996b). In this method,
§
N
c
3
S to
a
,o
Log Lake Area (m!)
Figure E-10. Total crustacean zooplankton taxa in North American
lakes (redrawn after Dodson 1992).
values above the 25th percentile are considered
unimpaired (similar to reference conditions) and
values below the 25th percentile are considered
impaired to some degree. The range from 0 to
the 25th percentile is bisected, with values in
the top half receiving a score of 3 and those in
the bottom half receiving a score of 1 (FigureE-
9). This approach also lends itself to scores
using percent of reference value (Figure E-9).
The second approach does not include definition
of reference criteria, but uses information from
the entire range of sites, from the most to the
least affected by anthropogenic pollution and
disturbance. A large and representative survey
data set is required to develop the reference
criteria. Reference expectations and scoring
criteria are based on the best values observed
for each metric, even if the best values do not
occur in the least affected sites (Simon and
Lyons 1995). The most common scoring method
is trisection (Karr et al. 1986) using the 95th
percentile of the metric distribution. Metric
values from 0 (or the lowest possible value) to
the 95th percentile are trisected; values in the
top one-third receive a 5, values in the middle
third receive a 3, and values in the bottom third
receive a 1 (most impaired).
Choice of scoring method should be based on the
approach used for defining reference sites, rather
than on the method that will produce the most
conservative or most liberal scoring. If reference
sites are representative of relatively unimpaired
conditions, then the lower percentile cutoff and
bisection is preferred. If reference sites are not
definable, then scoring criteria
based on the "best" values are
the only alternative.
To account for eovariables such
as size, the data are plotted, a
locally weighted estimate is made
of the appropriate percentile
(95th or 25th), and the range
below it is trisected or bisected
(Figure E-10).
Additive Index
The index is the sum of the
scores of the selected metrics
that prove responsive to distur-
bance. Criteria for index values
are also generated from the
reference sites, just as with
E-13
-------
Appendix E
Individual metrics. A perfect index score is
unlikely in the reference sites, therefore, a
reference expectation is developed for the total
score. Because the index is the sum of several
metrics, the Central Limit Theorem predicts that
it will have a lower coefficient of variation than
Individual metrics, and can be approximated
better by a normal distribution than can indi-
vidual metrics (Fore et al. 1994). Because of
these properties, multimetric indices can
usually distinguish 3 to 5 statistically significant
gradations of impairment, based on comparison
of a single sample to the reference distribution
(Fore et al. 1994, Gerritsen 1995).
Discriminant Model Index
Discriminant analysis may be used to develop a
model that will divide, or discriminate, observa-
tions among two or more predetermined classes.
Output of discriminant analysis is a function
that is a linear combination of the input vari-
ables, and that obtains the maximum separa-
tion (discrimination) among the defined classes.
The model may then be used to determine class
membership of new observations. Thus, given a
set of unaffected reference sites, and a set of
degraded sites (due to toxicity, low DO, or
habitat degradation), a discriminant function
model can identify variables that will discrimi-
nate reference from degraded sites.
Developing biocriteria with a discriminant model
requires a training data set to develop the
discriminant model, and a confirmation data set
to test the model. The training and confirmation
data may be from the same biosurvey, randomly
divided Into two, or they may be two consecutive
years of survey data, etc. All sites in each data
set are identified by degradation class (e.g.,
reference vg impaired) or by designated aquatic
life use class. To avoid circularity, identification
of reference and Impaired, or of designated use
classes, should be made from non-biological
information such as riparian zone modification;
known discharges, known contamination,
toxicity, nonpoint sources, impervious surface in
the watershed, land use practices, etc.
One or more discriminant function models are
developed from the training set, to predict class
membership from biological data. After develop-
ment, the model is applied to the confirmation
data set to determine its performance: The test
determines how well the model can assign sites
to classes, using independent data that were not
used to develop the model. More information on
discriminant analysis is in any textbook on
multivariate statistics (e.g., Ludwig and
Reynolds 1988, Jongman et al. 1987, Johnson
and Wichern 1992).
Classification of Estuaries with a
Discriminant Model Index
A straightforward a priori classification for
estuaries was that used by EMAP-Estuaries:
first, regionalization into the biogeographic
provinces used by NOAA, the U.S. Fish and
Wildlife Service and USEPA; second, stratifica-
tion of estuaries by physical characteristics of
shape and size; and third, measurement of
physical eovariates that affect assemblage
composition, principally salinity, depth, and
sediment attributes (for benthos) (USEPA
1993e).
Estuaries were classified as large estuaries,
large tidal rivers, and small estuaries. Data
collected in 1990 showed that large estuaries
had the greatest number of taxa, and tidal rivers
the fewest taxa. On that basis, the three estuary
classes were retained for further analysis.
It has long been known that estuarine faunal
diversity is highest at the seaward end of estuar-
ies, in full-salinity seawater. The lowest number
of taxa are found in brackish waters that are too
saline for freshwater organisms, and too fresh for
marine-adapted organisms. To characterize
reference conditions in EMAP estuaries, it was
therefore necessary to predict the number of taxa
that could be found at any given salinity. Figure
E-ll shows effect of the covariate, salinity, on
the number of taxa captured in benthic grabs. In
the example shown, the reference expectation for
EMAP estuaries in the Virginian Province was a
third order polynomial regression of a running
average 90th percentile of the data shown, given
by the line in the figure.
Of five possible eovariates considered in EMAP-
estuaries, only salinity was deemed to have a
strong enough effect on benthic macroinver-
tebrates to justify adjusting reference expecta-
tions for it (USEPA 1993e, USEPA 1994h). The
observed number of taxa per site was corrected
by dividing by the expected number of taxa for
the site, obtained from the regression model, to
yield percent of expected number of taxa (USEPA
1993e, Engle et al. 1994, USEPA 1994h).
E-14
-------
Biological Assemblages
60
* Base Stations
ft Reference Stations
¦ Degraded Stations
50
CO
fc
Q.
»
M
§
*
40
20
c
<0
«
2
'«?
0
5
10
25
30
35
Sainity (ppt)
Index
From the set of sampled sites in the EMAP data
set, reference and degraded sites were identified
based on predetermined criteria. Criteria for
reference sites were:
• Summertime bottom DO never less than 1
ppm.
• No contaminants observed in the sediment
(1980).
• No toxicity observed in the sediment.
Reference sites had to meet all three criteria.
Reference site salinities ranged from < 5 ppt to >
18 ppt.
Sites were rated as degraded if they met either
of the two following criteria;
• One or more hypoxic events with bottom DO
< 0.3 ppm.
• The concentration of at least one sediment
contaminant exceeding the ER-M value, and
Ampelisca bioassay indicated toxicity (<
75% survival and significantly different from
control) (USEPA 1993e).
Stepwise discriminant analysis was used to
determine which metrics could best discriminate
between reference and degraded sites. Because
number of taxa was deemed an important
indicator by itself, it was "forced" into the
discriminant model. The eventual discriminant
model had five variables:
• % expected number of taxa.
• No. of amphipods.
• % of abundance as bivalves.
• Mean weight per polychaete.
• No. of capitellid polychaetes.
E-15
-------
Appendix E
The model correctly classified 89% of degraded
sites, and 86% of reference sites, using the
learning data set to test its performance. Dis-
criminant scores were normalized to a range of 0
to 10, for ease In communication of index scores
(US EPA 1993e).
The original discriminant model was developed
from the 1990 EMAP-Virginlan Province sam-
pling effort. The model was subsequently tested
with the EMAP-Virginian Province data set
collected in 1991, an independent test (USEPA
1994h). The 1990 model failed to discriminate
the 1991 data correctly, so a new discriminant
model was developed using both 1990 and 1991
data sets. The revised model used the following
3 variables;
• Mean abundance of opportunistic species.
• Biomass/abundance ratio for all species.
• Mean number of infaunal species per grab.
The revised model was subsequently tested with
another independent data set, the 1992 Virginian
Province data. The revised model correctly dis-
criminated the 1992 reference and degraded sites,
and the model was not further modified (USEPA
19941). The model correctly identified 83% of
reference sites and 100% of degraded sites.
Designated Aquatic Life Use Classes
An alternative to the above methodology is to
develop biocriteria directly for administrative
aquatic life use classes (Davies et al. 1993). In
this approach, data from a set of sites (the train-
ing set) are assigned to predetermined aquatic life
use classes. The classes are determined by
regulation and might be (for example): (a) pristine;
(b) altered habitats, but native species maintained;
(c) discharges and vegetation permitted, native
communities altered, but fishable-swimmable
goals met; or (d) nonattalnment. Experts assign
sites to one of the four classes based on the
narrative descriptions of the aquatic life use
classes (above) and biological data from the
training set sites (Davies et al. 1993).
One or more discriminant models to predict class
membership are developed from the training set.
The purpose of the discriminant analysis here is
not to test the classification (the classification Is
administrative rather than scientific), but to
assign test sites to one of the classes.
An example of this approach is the biocriteria
adopted by Maine for streams (Davies et al.
1993). Stream biologists assigned a training set
of streams to four life use classes. A two-stage
discriminant modeling process was used to
develop discriminant models for assigning test
streams to use classes. The first stage was a
model to predict membership in each of the four
classes, expressed as a probability for each. The
second stage was a set of three discriminant
models that predict two-way class membership
(i.e., nonattainment (NA) versus A or B or C; NA
or C versus A or B; and NA or C or B versus A).
A selection procedure was used to select predic-
tive variables for the models, and the second-
stage models were constrained to exclude predic-
tive variables used in the first-stage model. This
approach is detailed by Davies et al. (1993).
Multivariate Ordination Model
The third approach that has been successfully
used for development of biocriteria uses multi-
variate ordination to determine if test sites are
different from reference sites. The comparisons
are usually made graphically, in ordination
space (c.f.. Figures E-l, E-3, and E-9); such that
if a site is outside of the area on an ordination
diagram defined by reference sites, It is judged
to be degraded.
Classification of reference sites is often a poste-
riori, using one of several clustering methods on
the biological data. Following definition of
biological clusters (reference classes), a discrimi-
nant model is developed using physical-chemi-
cal data to allow classification of test sites (Moss
et al. 1987, Wright et al. 1984, Reynoldson et al.
1995, Norris 1995).
Cluster analysis must be done with great
caution because there are many similarity
measures and many clustering algorithms,
many of which may produce different, and often
unintelligible, results (Ludwig and Reynolds
1988, Jackson 1993a). In general, the best
results for bioassessment purposes have been
achieved with UPGMA, and with TWINSPAN, a
divisive technique (Reynoldson et al. 1995, Moss
et al. 1987, Gauch 1982). The most successful
similarity measures have been Bray-Curtis
similarity, chord distance, and relative Euclid-
ean distance (Kenkel and Orloci 1986, Ludwig
and Reynolds 1988).
E-16
-------
Biological Assemblages
Ordination analysis is often done after determina-
tion of clusters, to see whether the identified
clusters also separate in ordination space. Hie
clusters are now treated the same as an a priori
classification. Ordination methods most often
used at this stage include correspondence analy-
sis and non-metric multidimensional scaling (e.g.,
Moss et al. 1987, Reynoldson et al. 1995).
Test sites are then assigned to one of the
reference classes with the discriminant model
(using physical-chemical data), and test sites
are compared to their respective reference
population in ordination space.
The ordination approach is illustrated with the
classification of benthic macroinvertebrate
assemblages from Great Lakes reference sites
(Reynoldson et al. 1995). Cluster analysis of
benthic macroinvertebrates from 96 reference
sites revealed five groups of sites. Cluster
analysis used the Bray-Curtis distance measure,
and the clustering algorithm was unweighted
pair group mean averages (UPGMA) (Reynoldson
et al. 1995). The sites were subsequently
visually depicted with ordination by NMDS,
which showed each cluster occupying a unique
area in ordination space.
Following classification, a discriminant model
was developed to identify class membership of
sites using physical and chemical (i.e., non-
biological) data. The model was developed with
the reference sites as the calibration data set,
for subsequent use with test sites to identify the
class of benthic assemblage the test site should
belong to. Sites were selected for uniform
characteristics (< 2 km from shore, < 30 m
depth, fine-grained sediment), and the criterion
for reference sites was large distance (> 10 km)
from known discharges (Reynoldson et al. 1995).
First, explanatory variables for input into the
discriminant function analysis were identified
with correlation analysis of physical-chemical
variables. Variables that were significantly
correlated with any of the three ordination axes
were used for the discriminant analysis. Of 25
variables examined, 18 were strongly correlated
with the ordination and were input into the
stepwise discriminant analysis. Of these, nine
produced the best model to predict class mem-
bership. Biological assemblage group member-
ship was correctly predicted by the discriminant
model for 87% of the sites, ranging from 64%-
100% for each of the five assemblage groups
(Reynoldson et al. 1995).
Biological integrity of test sites was assessed by
first assigning a test site to one of the five
assemblage groups, based on the discriminant
model applied to the site's physical-chemical
data. The biological assemblage structure of the
test site was then compared to assemblage
structure of the reference sites of that group, by
plotting the positions of reference sites and the
test site in ordination space, and determining if
the test site was within the region defined by the
reference sites (e.g., Figure E-12). The approach
was used for an assessment of benthic sites In
Collingwood Harbour, Ontario, which had been
contaminated with metals. Benthic
macroinvertebrate assemblages were different
from reference sites within two boat slips, and
the authors concluded that sediment
remediation was justified in the boat slips.
Outer reaches of the harbor exceeded Ontario
sediment metals criteria but benthic assem-
blages in the outer harbor were similar to
reference sites of their respective classes.
Because there were no discernible biological
differences, the authors concluded that sedi-
ment remediation could not be justified in the
outer harbor (Reynoldson et al. 1995).
T> 0 ¦
c
o
a aQ
T~
-2
O Reference
~ Unknown
A Impacted
First Axis
Figure E-12. Assessment by ordination. Solid circles
are reference sites, known impacted sites (triangles)
deviate from the reference group, primarily on the
first axis. Impairment may be judged by whether a
site is outside the region bounding reference sites
(ellipse), or by the distance between a site and the
reference centroid (arrow).
E-17
-------
Appendix F
Executive Summaries of Stale Pilot
Studies
The following includes executive summaries
from three states (Maine, Vermont and Wiscon-
sin) that performed and completed pilot studies
based upon the Lake and Reservoir Bioassess-
ment and Biocriteria Technical Guidance
Document. These pilot studies are distinct from
the case studies incorporated throughout this
document and, thus, are presented separately.
The purpose of each pilot study was not to
evaluate individual state sampling field methods
but rather to test the utility of the data analy-
sis and biocriteria development guidance in
this manual. Each state adapted its study to
incorporate its available data. All three states
reported favorably on the usefulness of this
guidance manual and indicated the various
stages of their own biocriteria development.
For more information regarding these studies,
please contact the state's project leader as
listed in each executive summary.
Geographic Analysis and Categorization of Maine Lakes: A Trial of
the Draft Lake Bioassessment and Biocriteria Technical Guidance
Maine's test of the Draft Lake Bioassessment
and Biocriteria Technical Guidance focused on
lake classification, selection of reference sites
and metric evaluation from extant data. Dataset
limitations prevented the creation of a lake
biocriteria program for the entire state; instead,
various aspects of the guidance were 'spot
tested.'
Classification of Lakes
Although Maine lakes are quite diverse, it was
decided that delineating 10 or fewer lake classes
would be most practical. The classification
effort focused on 451 lakes which had complete
datasets, utilizing cluster analysis on a combi-
nation of morphometric and chemical variables
including surface area, flushing rate, maximum
depth, mean depth, drainage area, elevation,
color, alkalinity and specific conductance. The
ecoregion approach suggested in the guidance
resulted in 3 'modified* ecoregions (based on
Omernik 1987) each having two lake classes.
Clustering was primarily based on surface area
and depth variables.
In this portion of the test, the biggest con-
straints were related to the difficulties of doing
statistical analysis with available software.
Three other concerns were evident: (1) how to
handle outlier lakes (i.e., very large lakes), ( 2)
F-1
-------
Appendix F
how to Incorporate additional lakes as data
becomes available, and (3) the selection bias
represented by monitored lakes in the datasets.
Selection of Reference Lakes
Each of the lakes in the dataset had a develop-
ment ranking assigned from 1990 Census Data.
Rankings from a subset of these lakes were
found to be similar to rankings derived from the
examination of United States Geological Service
(USGS) topographic maps. Lakes having low
development rankings were screened by profes-
sionals to eliminate lakes impacted by activities
unrelated to population. Box and whisker plots
were used to compare trophic status of reference
lakes to non-reference lakes in each lake class.
These comparisons often showed little separa-
tion between reference and non-reference lakes.
This may be partially explained by the dispro-
portionate number of reference and non-
reference lakes. The technique used to choose
reference lakes may need refinement but ap-
pears to be compatible with the guidance
emphasis on using readily available information.
Biological Parameters
Maine examined metrics from Tier 2B level
biological data (phytoplankton and zooplankton)
to evaluate their potential utility. Some metrics
were suggested in the guidance, others had
basis in current scientific literature and a few
were suggested by the nature of the dataset.
Forty-six phytoplankton metrics were examined;
the thirteen metrics showing potential utility
were reduced to four after the elimination of
redundant metrics: total cell volume, % volume
Cyanophyta, % volume chrysophytes and the
ratio of volume of Cyanophyta to desmids.
Seven out of nineteen zooplankton metrics
showed potential utility in screening for trophic
Increases. Metrics were eliminated with inher-
ent redundancy and scoring was developed for
two: total abundance and the ratio of cladocer-
ans to copepods. Cumulative distribution plots
for reference sites and non-reference sites were
utilized to determine scoring levels for the two
metrics. A multimetric index was not developed
due to the low number of lakes and the overlap
of the Tier 2B biological data.
The guidance provided a reasonable framework
for the development of the biological metrics.
However, the literature suggests that rotifers
respond to trophic changes as well as crusta-
cean zooplankton, a point previously overlooked
in the guidance. Potential phytoplankton
metrics should not only be based on count data,
but also total cross-sectional area and/or
volume measures. The variation in phytoplank-
ton cell size is so great that cell numbers do not
approximate standing crop as well as volume or
area. Although a multimetric index was not
developed, it appears that reasonable guidance
is provided in the draft to accomplish this.
General Comments and Applicability within
Maine's Lake Management Strategy
There are some concerns that this experience will
be extrapolated to the rest of the country. For
example, Maine has a large number of lakes and
most are of glacial origin. Maine has only 3 lakes
receiving point source discharges: Maine's lake
management focus continues to be on trophic
status and NPS pollution control. The test of this
document was from this perspective, and other
states may have other or additional management
priorities. Another concern is whether or not the
gain is worth the additional cost of obtaining
biological (phytoplankton and zooplankton) data.
One aspect that may be beyond the scope of the
guidance document is the lack of reference lakes
for atmospheric deposition impacts (in particular,
Hg accumulation). If one assumes that atmo-
spheric deposition is somewhat uniform over
regions, it becomes nearly impossible to select
unimpacted reference lakes and strong reliance
must be placed on the term "minimally impaired"
as used in the guidance.
Overall, this test has been considered a success
despite some limitations of the dataset, and
Maine has shifted from 'test' mode to 'develop-
ment' mode. Biological samples have been
collected from 100 potential reference lakes
(1996) which are currently being analyzed and
similar samples from 100 lakes of unknown
status (1997) which will be analyzed and used
as a test dataset in the future. The development
of biocriteria for Maine's lakes will be an ongo-
ing process over the next few years as time and
funding permits. It is anticipated that the
results will be useful to concerned citizens at the
local level as well as biologists at the state level.
Project Contacts:
Linda Bacon and Roy Bouchard
Maine Department of Environmental Protection
Division of Environmental Assessment
Bureau of Land and Water Quality
phone (207) 287-3901
fax (207) 287-7191
F-2
-------
Executive Summaries of State1 Pilot Studies
Biocrlteria Development for Vermont Lakes-Pilot and Field Phases
Project Summary, April, 1998
The development of test biocriteria by the Ver-
mont Department of Environmental Conservation
(VTDEC) was conducted in order to evaluate
methods presented in this Lake and Reservoir
Bioassessment and Biocriteria Technical Guid-
ance Document, Methods for conducting each
phase of the project were taken directly from the
document, and a comprehensive Pilot Phase final
report is available. The Pilot Phase of this project
was completed during 1995, using existing
information contained in the Vermont Lakes and
Ponds Database. Following this effort, a field
program was initiated to develop reference level
biocriteria beginning in 1996. Results and
lessons learned from the Pilot Phase and results
from the Field Phase (1996-1997) are presented
in this summary. Movement towards implemen-
tation of fully developed lake biocriteria for
Vermont is discussed.
Field Phase
Taking the lessons learned from the Pilot Phase,
a comprehensive bioassessment project was
designed using this Guidance. To avoid the
difficulties of classification, a regional
approach to definition of lake biological
reference conditions was adopted by
planning assessments of both Vermont
and New Hampshire lakes. To date,
this cooperative Field Phase has
evaluated 29 lakes. Ten additional
lakes are scheduled for assessment
during 1998. Data results from 1996
and 1997 are available, and an over-
view of analyses conducted with these
data to date is presented below. The
reader should note that trial criteria
presented below are provisional, and
should be considered in development.
The present study lake set contains 23
candidate reference lakes, and six
known impaired or test lakes. The test
lakes are either culturally eutrophied,
anthropogenically sedimented, or have
perpetually anoxic hypolimnia. The
geographic distribution of 1996-1997
study lakes is presented in Figure F-1.
Bioassessment Assemblages, Metrics,
and Methods
Following recommendations from the
Pilot Phase, the trophic state, phy-
toplankton, benthic macroinvertebrate, and
macrophyte assemblages were selected for
assessment. Trophic state parameters (Secchi
disk transparency, chlorophyll a, algal bio-
volume) were collected bi-weekly at a central
location in the lake. Phytoplankton were enum-
erated from a season-wide, whole lake com-
posite, consisting of composited, bi-weekly,
depth-integrated samples of the photic zone
acquired from a fixed station network. Discrete
bi-weekly composites were retained in archive
for future analysis if necessary. Profundal and
sublittoral benthic macroinvertebrate samples
were collected as triplicate composites using an
Ekman dredge, from a fixed station network on
each lake. Triplicate composited samples of
benthic macroinvertebrates from rocky-cobbled,
littoral-mud, and macrophyte bed habitats were
collected using a sweep net. A timed collection
period of 20 minutes total per composite sample
per habitat was employed to ensure quantitative
data comparability among lakes. The entire lit-
toral zone was surveyed for macrophytes,
whereby species were identified and abun-
dances classified using the Braun-Blanquet
Long (Greens)
Caspian ^
MeCon/xHI
BakJH
Nathan Pond
Wolcott* Cryrtai •
Long (She!)
Little El mora
MarshfieJd
Russell
»
Interval®
Hfnkura
* High (Sud)
• Spring (Shrews)
Hatch
Giiman
* Buttamut
Smith * m French
Wfflard • Dudley
Beaver Lake %
Figure F-1. Location of 1996-1997 study lakes In the
Bioassessment and Paleolimnology of Vermont and New
Hampshire Lakes Project
F-3
-------
Appendix F
scale. Benthic macroinvertcbrates and macro -
phytes were assessed during the mld-Iate
summer index period (approximately August 1-
August 31). Habitat quality was assessed at the
time of the macrophyte survey. A quality
assurance program was employed to ensure the
precision, accuracy, comparability and repre-
sentativeness of data collected. Table F-l
presents selected metrics under evaluation for
the Field Phase of this bioassessment project.
Preliminary Lake Classification
In selecting candidate reference and test lakes,
the same classification metrics were used as for
the Pilot Phase. An a-priori classification was
adopted using alkalinity as a classifying vari-
able. A cutoff of approximately 15mg/l was
used to classify lakes as poorly buffered ( 15
mg/1 as CaC03), or well-buffered (> 15mg/l as
GaC03). Existing lake assessment data suggests
that these two lake classes correspond to tannic
and clear water lakes (one exception being the
clear, but lower-alkalinity Hatch Pond, NH).
Thus for ease of presentation, low-alkalinity,
poorly buffered lakes are called tannic, while
higher-alkalinity, well buffered lakes are called
clear. Table F-2 provides the range of physico-
chemical attributes for each of these classes.
This proposed classification was validated with
phytoplankton data using canonical correspon-
dence analysis (Figure F-2). The position of
clear and tannic lakes is well separated along
the second axis, as are the relative positions of
the algal orders. Test lakes with increased blue-
green algae in the community separate along the
first axis. This ordination suggests that there is
variability in the phytoplankton assemblage
biometrics which can be explained by the
proposed classification.
Criteria Development, Phytoplankton
All phytoplankton data were examined using
Tukey box plots to identify metrics which discrimi-
nate between reference and test lakes. Metrics
thus 'appearing* discriminatory were tested by
calculating interquartile coefficients. To avoid
'double counting' of impairments, the selected
metrics were examined for covariance. A high
degree of covariance was noted between the
percent composition of cyanobacteria and percent
composition of Aphanizomenonjlos-aquae, Ana-
baena jlos-aqiiae, and Anacystis marina (r=0.85,
p<0.05). The latter metric was retained as it more
accurately defines occurrences of undesirable
blue-green algal blooms in test lakes.
A total of five metrics were selected for each lake
class to construct a phytoplankton index. Trial
criteria were developed by scoring the metric
ranges using the 'bi-section' method presented
In this Guidance. Trial criteria are presented in
Table F-3. Lakes were scored using these
criteria, and the distribution of scores is pre-
sented in Figure F-3. For tannic lakes, both test
lakes met reference criteria. None of the clear
test lakes met reference criteria.
Table F-1. Selected tier two metrics evaluated for
1996-1997 Bioassessment arid Paleolimnology of
Vermont and New Hampshire Lakes Project study
lakes.
Trophic State and Physico-chemical;
Alkalinity
Conductivity
Dissolved oxygen
Algal blovolume - Sweet TSI
Chlorophyll a - Carlson TSI
Secchl transparency - Carlson TSI
Benthic Macrolnvertebrate:
No. of taxa
% dominants
Shannon-Welner index of diversity
% intolerant species
COTE index (Coleoptera, Odonata, Ephemeroptera,
Trichoptera)
% Intolerant chironomids
No. of Crustacea - Mollusca taxa
Functionality (ie. shredder, scrapers...)
Macrophytes:
% cover - littoral zone
% cover - littoral zone, nuisance species
No. of species
Relative species dominance
No. of rare species
No. of Potamogeton spp.
No. of Utricularia spp.
% occurrence by structural morphology
Phytoplankton:
Total density
Total blovolume
Shannon-Welner diversity
% Anabena spp., Aphanizomenon spp., Anacystis spp.
% cyanobacteria (density and biovolume)
% diatoms (density and biovolume)
% chlorophytes (density and blovolume)
% euglenophytes (density and blovolume)
% phyrrophytes (density and biovolume)
% cryptophytes (density and biovolume)
F-4
-------
Executive Summaries of State Pilot Studies
Table F-2. Ranges of selected attributes of candidate
reference lakes falling into two classes evaluated in
conjunction with the Bloassessment and Paleolimnology of
Vermont and New Hampshire Lakes Project.
Tannic lakes
Clear lakes
Size (ac)
20-96
20-789
Depth (m)
3-17
7-43
Alkalinity
6-14
9.6-100*
Mean Secchi disk
transparency
1.7-11.5
2.5-7.8
CM
-2
¦1
0 OCAAxis1(p^43 1
2
Figure F-2. Canonical correspondence ordination triplot for unclassified reference and test lakes evaluated
in conjunction with the Bioassessment and Paleolimnology of Vermont and New Hampshire Lakes Project.
Lakes are denoted as clear (C) or tannic (T). Eigenvalues (Y) are provided for each axis. For simplification,
physical variables are grouped and presented by their relative position to the ordination axes. Relative
percent composition by algal orders are scaled by a factor of 2 for ease of interpretation.
F-5
-------
Appendix F
Table F-3. Trial phytoplankton assemblage biocriteria for Bioassessment and Paleolimnology of Vermont
and New Hampshire Lakes Project study lakes.
Tannic lake metrics
Interquartile
coefficient
Score attributed:
1 3 5
Total density
0.40
>980
862-980
<862
Total blovolume
0.89
>361K
287K-361K
<287K
% cryptophytes
0.66
>47
28-47
<28
% dlatoms-biovolume
0.99
<11
11-13.4
>13.4
% APHA-ANFA-ANMA*
0.56
>10
1-10
<1
Clear lake metrics
Score attributed:
1 3 5
Total density
0.15
>780
620-780
<620
Total blovolume
0.09
>580K
389K-580K
<389K
% cryptophytes
0.76
>9
9-19
<19
% dlatoms-biovolume
0.67
<42
42-63.5
>63.5
% APHA-ANFA-ANMA*
0.06
>5
13-5
<3
X 8
•§£
c <=
o to
32 c
c o
CO Q.
II
30
25 -
20-
15
— 10
Meets Reference
Colored Lakes
Meets Reference
Clear Lakes
Reference
Test
Reference
Test
Figure F-3. Distribution of phytoplankton index scores for 29 classified study lakes. The dotted line
corresponds to the lower quartile of the reference distribution. Lakes which score above this value are
considered meeting reference conditions.
F-6
-------
Executive Summaries of State Pilot Studies
Criteria Development, Littoral Zone
Macrophytes
The methods by which macrophyte data were
collected do not permit calculation of a Shan-
non-Weiner index of diversity, yet this is an
important measure describing the macrophyte
assemblage. To provide an alternate measure, a
relative species dominance metric is proposed
where impairment is indicated by metric values
increasing above reference. This metric Is
calculated as:
% cover - littoral zone
No. of species
An alternate way of assessing macrophyte
communities is to determine the relative contri-
bution by different structural groups. Analo-
gous to relative percent composition by algal
divisions in the phytoplankton, or by function In
the macrolnvertebrates, relative percent occur-
rence by structural grouping can affect other
biological assemblages and vary with impair-
ments to lake water quality. To evaluate this for
the study lakes, seven structural groupings were
proposed (Table F- 4).
Interquartile coefficients for macrophyte metrics
were calculated, and many metrics were found
to be insensitive. The most discriminating
metrics were nevertheless retained for trial
criteria development In the interest of assessing
test lakes against a reference condition. The
criteria presented In Table F-5 are at best draft,
and should be considered in development
pending the acquisition of additional data.
Reference and test lake scoring is presented in
Figure F-4.
Table F-4. Proposed structural macrophyte groupings for use in
bioassessment of Vermont and New Hampshire lakes.
Proposed structural group
Representative example species
Emergent erect
Pontederia spp.
Emergent pronate
Sparganium minimum
Floating leaved
Brasenia schreberi
Submerged narrow-leafed
Najas spp.
Submerged broad-leafed
Potamogeton amplifolius
Submerged whorled
Ceratophyllum spp.
Submerged mat-like
Eriocaulon spp.
Table F-5. Trial macrophyte assemblage biocriteria for Bioassessment and Paleolimnoiogy of Vermont
and New Hampshire Lakes Project study lakes.
Metrics
Tannic Lakes
IC 1 3 5
Clear Lakes
IC 1 3 5
Percent cover-littoral
zone
0.55
>26
21-26
<21
1.19
>37
22-37
<22
Relative species
dominance
0.28
>1.5
1.2-1.5
<1.2
1.87
>1.4
1-1.4
<1
% occurrence floating
leaved
1.60
>25
20.1-25
<20.1
0.37
>18
15-18
<15
% occurrence
submerged narrow
leaved
1.05
<6
6-10
>10
3.76
>35
32-35
>32
% occurrence
submerged whorled
2.06
>12
6-12
<6
0.52
>12
7-12
<7
F-7
-------
Appendix F
Figure F-4. Distribution of macrophyte index scores for 29 classified lakes. The dotted line corresponds
to tha lower quartite of the reference distribution. Lakes which score above this value are considered
meeting reference conditions.
S
8
x 8
•81
£ e
5
-------
Executive Summaries of State Pilot Studies
Table F-6. Interquartile coefficients (IC) and trial biocriteria
scoring for trophic state indices - Bioassessment and
Paleolimnology of Vermont and New Hampshire Lakes Project
study lakes.
Biocriteria scoring
Tannic lake metrics
IC 1 3 5
TSI Chlorophyll-a
0.35
>52
49-52
<49
TSI algal biovolume
0.27
>52
49-52
<49
TSI Secchi disk
transparency
1,82
>52
50-52
<50
Clear lake metrics
TSI Chlorophyll-a
1.50
>48
44-48
<44
TSI algal biovolume
1.18
<46
42-46
<42
TSI Secchi disk
transparency
0.93
>40
39-40
<39
evaluated. Based upon review of
the 1996 taxonomy, and observa-
tions from the 1997 samples
currently in taxonomy, it is be-
lieved that profundal zone samples
may not provide useful data.
Indeed, profundal zone samples are
comprised almost entirely of
Chtronomidae, Chaoboridae, and
Oligochaeta, and variation in overall
density is dependent on
hypolimnetic oxygen conditions. If
these observations hold true,
profundal zone macroinvertebrate
assessments will likely be dropped
from Vermont's bioassessment
protocols.
Moving Toward Implementation
16
Meets Reference
Impairment to
assemblage
k'
10
J?
4?
Meets Reference
Impairment to
assemblage
cr
Figure F-S. Distribution of trophic state index scores for 29 classified lakes. The dotted lines correspond
to the lower quartile and lower 5th percentile of the reference distribution. Lakes which score above the
lower quartile value are considered meeting reference conditions. Lakes which score below the 5th
percentile are considered impaired.
F-9
-------
Appendix F
Table F-7. Potentially robust macroinvertabrate metrics for five lakes habitats evaluated in
conjunction with the Paleolimnology and Bioassessment of Vermont and New Hampshire
Lakes Project,
Habitat
Macroinvertebrate metric
Profunda!
VT-BI*; % other; % intolerant chironomids
Sublfttoral
% collector-Merer, % predator, % shredder-detritivore, %
shredder-herbivore
Littoral rocky-cobbled
VT-BI, Shannon-Wiener diversity, % collector-filterer, %
predator, % shredder-detritivore, % shredder-herbivore, %
coleopterans, % tn'chopterans, % oligochaetes
Littoral macrophyte-beds
VT-BI, Shannon-Wiener diversity, % collector-gatherer, %
predator, % coleopterans, % trichopterans, % oligochaetes
Littoral fine-muds
VT-BI, Shannon-Wiener diversity, % collector-filterer, %
predator, % coleopterans, % trichopterans, % oligochaetes
'VT-Bt is the Vermont stream biotic index
Vermont's efforts toward developing useful
biocriteria are by no means complete. The Field
Phase outlined above was designed to provide the
States of Vermont and New Hampshire with the
baseline experience and information needed to
move forward with a long-term sustainable
bioassessment program. The trial criteria pre-
sented herein should be reevaluated and further
refined in conjunction with that longer term
program. The present Field Phase provides a large
volume of useful data, but it may not be sustain-
able over the long-term. Using results from the
1996 to 1998 study lakes, the need for robust
data to develop (and assess compliance with)
criteria will be balanced with the fixed personnel
and operating expenses of a small State agency.
Presently, the classification of study lakes is
provisional. The reference condition for poorly-
buffered (tannic) lakes is well characterized,
though a group of poorly buffered and clear
lakes might need to be characterized sepa-
rately. Also there exists the need to assess the
reference condition for a variety of well buffered
(clear) lake types. Progress on this will be made
during the 1998 field season.
This cooperative Vermont/New Hampshire
initiative carries with it a paleolimnological
component designed specifically to determine
the historical condition of candidate reference
lakes. Application of paleolimnological models to
the sediments of selected candidate reference
lakes will ensure that the underlying biological
information used to develop criteria is indeed of
reference qualify.
Vermont has already seen the benefits of
biological assessment as a tool for evaluating
lakes. Data from this Field Phase have been
used to refine and update Aquatic Life Use
Support in Vermont's 305(b) inventory for every
Vermont study lake bioassessed to date. While
numeric criteria are not yet ready for inclusion
into Vermont's Water Quality Standards, it is
anticipated that subsequent revisions to Stan-
dards will contain lake biological criteria.
Readers of this Guidance are encouraged to
communicate with the Project Contact directly
for information regarding Vermont's
bioassessment program.
Project Contact;
Neil Kamman, Aquatic Biologist
Vermont Department of Environmental
Conservation
103 S. Main St., 10N
Waterbury, VT 05671-0108
phone (802) 241-3777
fax (802)241-3287
neilk@dec. anr.state. vt. us
http://www.anr.state.vt.us/water1.htm
F-10
-------
Executive Summaries of State Pilot Studies
An Evaluation of the Draft Lake and Reservoir Bloassessment and
Biocrlteria Technical Guidance Document Using Wisconsin Lakes
Wisconsin's test of the Draft Lake
Bioassessment and Biocrlteria Technical Guid-
ance focused on the development of a
multimetrie index for Wisconsin lakes. Such an
index would provide a more accurate determina-
tion of use impairment for the biennial 305(b)
water quality assessment, and changes as a
result of watershed best management practices
under the nonpoint priority watershed program.
This index would also allow more informed
permitting decisions, as well as proactive
management by rapidly detecting emerging
pollutant threats to lakes.
Background
Although bioassessment could detect changes
from a broad range of anthropogenic sources, this
study only involved lakes that have been impaired
by eutrophication. While it would be beneficial to
include other pollutants, e.g., acid precipitation
and mercury, there was not sufficient information
in our data set from lakes impacted by such
pollutants. All of the lakes in our data set have
experienced some degree of impairment from
anthropogenic sources. As such, the reference
lakes would be classified as "least impaired."
Better reference sites could be selected if they had
been chosen prior to data collection.
This study examines the trophic variables: total
phosphorus, chlorophyll, Secchi depth and the
biotic communities of phytoplankton, zooplank-
ton, sedimented diatoms, and macrophytes.
Analysis involved a comparison of index devel-
opment using Tier 1A (single visit) Tier IB
(multiple visits) as well as Tier 2A and 2B.
Macroinvertebrates and fish were not used in
the analysis.
Data used for this analysis was largely collected
under the Long Term Lakes program of the
Wisconsin Department of Natural Resources.
Collection began in 1986 and continues through
the present. Samples are collected five times
annually: late winter, June, July, August, and
during fall turnover. Parameters that are
analyzed from these collections include Secchi
depth, chlorophyll, total phosphorus, phy-
toplankton, and zooplankton. In addition, the
macrophyte assemblage was surveyed occasion-
ally during this time period. Not all of the
samples are available for this analysis. Trophic
variables (Secchi, chlorophyll, and P) were
available for the years 1986-1994 while phy-
toplankton from 1986 were used and zooplank-
ton from 1986 and 1988 were used. As part of
another project, sediment samples were col-
lected from the main basin of most of these
lakes in 1991 for diatom analysis. Seven
reservoirs of varying trophic characteristics were
sampled in 1994 for trophic variables, phy-
toplankton, and macrophytes.
Classification
Hie lakes sampled covered three ecoregions, but
not all regions contain reference quality lakes.
Therefore, lakes were not separated into three
ecoregions. In fact, all but two of the reference
lakes were found in the northern lakes and forests
ecoregion. The reference lakes were chosen based
upon low levels of development in their water-
sheds. All of the lakes had some development on
the shoreline but most were summer homes and
the density was relatively low. A criteria that was
not used in the selection of the reference lakes
was their known trophic status. It was felt that
lakes that may have naturally had higher nutrient
levels but low development should be included in
the reference lakes. Since all of the lakes had
some degree of disturbance in their watersheds
the reference condition was calculated using the
bisection method.
The reference lakes, test lakes, and reservoirs
exhibited a wide range of morphological condi-
tions and watershed size (Table F-8). These
data were used to determine how robust various
metrics were in assessing unknown lakes.
When possible, metrics were constructed based
upon multiple visits as well as single visits
during an index period. August was chosen as
the index period.
Metric Development
At least 4 metrics were developed for each
biological entity (Table F-9). Trophic variables of
the lakes were described using Carlson's
Trophic Status Index fTSI), modified for Wiscon-
sin lakes. Trophic status for chlorophyll,
phosphorus, and Secchi depth was described
using the equations:
WTSIsd = 60 -(32.2 Log SD)
WTSItp = 60 -(33.2(0.96 - .054 log TP)
F-11
-------
Appendix F
WTSIchI » 60 - (33.2(0.76 - 0.52 log Chi)
Reference metrics are unusable if they are
excessively variable. Variability of the metrics
was measured by determining the ratio of the
interquartile range to the scope of detection.
The scope of detection was defined as the
distance from the lower quartile to the minimum
value possible when reference values were
higher than the test cases. When the reference
values were lower than the test cases, the scope
of detection was defined as the distance from
the upper quartile to the maximum value
possible. An interquartile coefficient greater
than one generally indicates the metric is too
variable to detect impairments.
Some of the metrics had an interquartile coeffi-
cient greater than one, but each of the biological
units had at least one metric with a coefficient less
than one (Table F-9). For metrics to be useful
there must also be good separation between the
reference and test lakes. Each biological unit also
had at least one metric that fulfilled this condition.
The list of metrics that were judged to be robust
and useful are listed in Table F-10.
Discussion
This study has formulated a draft bioassessment
index for Wisconsin lakes using biocriteria
metrics. Different metrics were not formulated
for individual lake classes or for separate
ecoregions largely because of the lack of suffi-
cient reference lakes. Instead, the lakes for all
the regions were combined. The Wisconsin Lake
Index was tested on 13 lakes using all of the
metrics and tested independently for Hers 1A,
IB, 2A, and 2B. A comparison was made of each
lake's classification at the four different levels.
All the lakes received the same classification
("departing from reference conditions") under Her
1A and IB. Tier 2 appeared to be more descrip-
tive of lake condition than Tier 1. In fact, two
lakes classified as "departing from reference
conditions" under Tier I, were categorized as
"impaired" using Tier 2 sampling techniques.
This analysis has allowed us to make some
recommendations concerning which metrics are
useful for developing a Wisconsin Lakes Index.
Since all the lakes received the same score
under Tier 1A and IB it is suggested that only a
single visit during the index period (August) is
necessary if lakes are to be classified using Tier
1 only. In addition, we suggest that the macro-
phyte metrics be expanded under Tier 1. All of
the macrophyte metrics, with the exception of
density, can be determined with little or no extra
effort under the suggested Tier 1 metrics in the
draft document. Density should be included as
a Tier 2 metric for macrophytes.
Although there was not complete agreement
between Tier 2A and 2B they were similar
enough to suggest that only one sampling trip is
sufficient. In addition, it is suggested that the
zooplankton metrics be eliminated until it is
better understood how these metrics relate to
the lake's impairment.
The diatom metrics tested were not as useful as
expected. The only metric that proved robust
enough to use was the percentage of Stephano-
discus. Although this metric was useful there
was a great deal of variability across the test
lakes. In the 13 lakes where the index was
tested, the diatom metric tended to indicate that
the lake was less impaired compared with most
of the other metrics. This metric may not be as
robust as some others. We recommend its usage
but with reservations. The diatom metrics likely
would be more useful if TSI values were calcu-
lated using the entire diatom assemblage but this
would entail considerably more work, including
detailed taxonomic knowledge.
A summary of the recommendations are:
Tier 1 Only one sampling trip (during August).
Metrics: trophic state variables, macro-
phyte metrics except density.
Tier 2 Only one sampling trip (during August).
Metrics: trophic state variables, macro-
phytes, phytoplankton, diatoms.
It is evident from this analysis that lake assess-
ment using biocriteria is a more robust tech-
nique than using the traditional indices: phos-
phorus, Secchi depth, and chlorophyll by
themselves. The additional information from the
biota gives a much more accurate picture of a
lake's health, especially its biological integrity.
Improvements could be made in developing a
Wisconsin Lake Index if better reference condi-
tions were used to define the metrics. Since
most of the reference lakes used in this study
had some iakeshore development they were not
ideal choices. Another ongoing study has
identified sufficient reference lakes in each of
the major ecoregions in the state, and will
develop metrics for sedimented diatoms.
F-12
-------
Executive Summaries of State Pilot Studies
Table F-8. Morphological data for the study lakes and reservoirs.
MORPHOLOGY
Lake Name
County
lake area
(Hectares)
mix,
depth im)
mean
depth (m)
like vol.
Cm3)
watershed
{hectares)
watershed
lake area
hydrology
description
S.0.F.
REFERENCE LAKES
Bear Piw Lake
Oconto
19.8
6 J
414
20.9
seepage
1.88
Eau Claire Lake {Lower)
Douglas
324.6
12.5
6.7
21763552
2813
8.7
drainage
1.73
Eau Oiire Lake (Upper)
Bayfield
403.1
28.0
8.8
35627887
2046
5.1
drainage
2.30
Eseanaba Lake
Vilas
U8.6
7.9
4.3
5059742
518
4.4
drainage
2.10
Franklin Lake
Oneida
65.2
7.6
259
4.0
seepage
1.46
Keyes Lake
Florence
81.7
23.5
570
7.0
drainage
1.61
Lac Courte Oreilles
Sawyer
2039.2
27.4
10.4
211327493
17042
8.4
drainage
2.55
Lost Lake
Florence
37.2
13.7
5.2
1929(65
52
1.4
seepage
1.09
Fatten Lake
Florence
103.2
15.8
5.5
5661681
2098
20.3
drainage
t.78
Round Lake
Chippewa
87.4
5.5
3.0
2664321
298
3.4
seepage
1.28
Silver Lake
Barron
136.4
27.7
11.6
15795967
1935
14.2
seepage
2.25
TEST LAKES
Amnicon Lake
Douglas
172.4
9.4
3.0
5254632
1251
7.3
drainage
2.70
Bass Lake
St. Croix
168.8
10.7
seepage
Big Cedar Lake
Washington
377.2
32.0
10.4
39086570
drainage
2.25
Big Green Lake
Green Lake
2972.8
71.9
31.7
942360318
958
0.3
drainage
1.39
Big Long Lake
Manitowoc
48 6
H.6
518
. 10.7
seepage
2.22
Big McKcnzie
Burnett
479.6
21.6
5.8
27771841
1709
3.6
drainage
1.47
Browns Lake
Racine
160.3
152
2.4
3907670
drainage
1.82
Butternut Lake
Price
407.1
9.8
4.3
17372357
11139
27.4
drainage
2.52
Cedar Lake
Polk/SL Croix
448.0
8.5
8614
19.2
drainage
1.42
Clark Lake
Door
351.3
7.6
2.1
7494635
4817
13.7
drainage
1.53
Crystal Lake
Sheboygan
61.5
18.6
6.1
3749785
seepage
1.87
Fish Lake
Dane
87.4
18.9
seepage
1.26
Fox Lake
Dodge
1062.3
5.8
2.1
22665227
15022
14.1
drainage
2.19
Fries® Lake
Washington
47.3
14.6
8.2
3896569
drainage
1.51
Kentuck Lake
Vilas
387.3
12.2
4.0
15345746
777
2.0
drainage
1,30
Lac La Belle
Waukesha
471.1
13.7
3.4
15793500
drainage
1.69
Long Lake
Chippewa
425.7
30.8
6.1
25952456
1746
4.1
drainage
3.08
Long Lake
Fond du Lac
168.8
14.3
6.7
i1315962
drainage
1.76
Mason Lake
Adams
346.0
2.7
2.1
7382388
9453
27.3
drainage
t.92
Minocqua Lake
Oneida
550.4
18.3
7.0
38583309
20720
37.6
drainage
3.68
Nagawlcka Lake
Waukesha
371.1
27.4
S1.0
40719699
drainage
1.98
Pelican Lake
Oneida
1450.8
11.9
2590
I.S
drainage
1.91
Pewaukee Lake
Waukesha
1008.9
13.7
4.6
46126050
drainage
1.94
Pike Lake
Marathon
83.0
10.4
4.0
3287229
829
10.0
drainage
1.32
Pike Lake
Washington
211.2
13.7
drainage
1.19
Ripley Lake
Jefferson
169.2
13,4
5.5
9280717
seepage
1.40
Keck Lake
Jefferson
554.8
17.1
4.9
27057656
drainage
1.43
Jfollingstone Lake
Langlade
271.9
3.7
2512
9.2
drainage
1.32
Sand Lake
Rusk/Chippewa
106.0
30.5
8.8
937)994
251
2.4
seepage
1.98
School Section Lake
Waupaca
15,8
11.6
8.5
1346962
drainage
1.76
Shell Lake
Washburn
1044,1
11.0
7.0
73194807
4159
4.0
seepage
1.43
Silver Lake
Waupaca
27.5
5.2
2.1
587137
seepage
1.05
Squaw Lake
Si. Croix
52.2
9.8
4.0
2068549
259
5.0
seepage
2.91
Thunder Lake
Oneida
742.6
2.7
2590
3.5
drainage
1.80
White Clay Lake
Shawano
94.7
14.0
4.3
4040886
1036
10.9
drainage
1.56
Whitewater Lake
Walworth
259.0
11.6
drainage
2,80
Wilson Lake
Iron
65.6
6.4
4.3
2797537
207
3.2
drainage
1.89
RESERVOIRS
Big Eau Pteine
Marathon
2764.0
14.0
4.9
134794885
8536
3.1
drainage
5.48
Brute
Florence
120.2
19.5
6.!
7326882
271949
2262.6
drainage
2.51
Caldron Falls
Marinette
412.0
12.2
4.6
18835266
124796
302.9
drainage
4.70
Dutch Hollow Lake
Sauk
83.0
12.2
1313
15.5
drainage
2.31
Gilc
Iron
1369.5
7.6
18130
13.2
drainage
3.19
Minong
Washburn
632.9
6.4
2.7
17362489
60507
95.6
drainage
4.48
Rainbow
Oneida
823.5
8.5
194249
235.9
drainage
3.53
Redstone Lake
Sauk
247.7
11.0
4.3
10568472
7677
31.0
drainage
4.69
St, Craix
Douglas
774.2
8.5
2.1
16517554
32437
41,9
drainage
3.38
Willow
Oneida
2551.9
9.1
3.0
77783359
84693
33.2
drainage
7.29
F-13
-------
Appendix F
Table F-9. Summary of degree of separation and the interquartile coefficient for all of the metrics.
Metric
Separation
Interquartile Coefficient
Trophic Variables
WTSlc,
yes
0.08
WTSIjo
yes
0.05
WTSIjt
yes
0.06
' ' "" ¦
Phyioplanktort
Richness
no
0.8
Ant, Apb, Micro
no
4.2
% Blue-green density
yea
1.0
% Blue-green biomass
yes
0.4
Zooplanlaon
Richness
yes
0.8
Daphniasizc
yes
1.0
Herbivore/Predator
yes
0.6
Copepod/Cladocera
no
Large Predator
QO
Chydorus
no
0.3
Diatoms
Richness
maybe
0.15
Diversity
no
0.5
% Planktonic taxa
no
2.3
% Aulacoscirs spp.
no
0.3
% Stephanodiscus spp.
yes
0.02
% Cyclotella
yes
7.3
1
Macrophytes
Richness
no
0.9
% Coverage of littoral zone
no
1.5
Max depth of growth
no
0.8
% Exotic taxa
yes
1.0
% Sensitive taxa
yes
1.0
F-14
-------
Executive Summaries of State Pilot Studies
Table F-10. Metrics that possess good separation
between reference and test lakes as well as an
interquartile coefficient less than or equal to 1.0.
Trophic Variables
WTSIc,
WTSIsd
WTSItp
Phytoplankton
% blue-green density
% blue-green biomass
Zooplankton
Herbivore/predator
Daphnia size
No. of taxa
Diatoms
No. of taxa
% Stephanodiscus
% Cyclotella
Macrophytes
% exotic species
% sensitive species
Project Contacts:
Paul Garrison and James Johnson
Wisconsin Department of Natural Resources
Bureau of Research
1350 Femrite Drive
Monona, Wl 53716
F-15
-------
Appendix G
Literature Cited
Alcaraz, M., G.A, Paffenhofer, and J.R. StricMer.
1980. Catching the algae: A first account of
visual observations on filter-feeding calanoids.
Pages 241-248 In ASLO Special Symposium 3:
The evolution and ecology of zooplankton commu-
nities. University Press of New England,
Hanover, NH.
APHA. 1992. Standard methods for the exami-
nation of water and. wastewater. 16th ed.
American Public Health Association, Washing-
ton, DC.
ASTM. 1990. Standard guide for collection,
storage, characterization, and manipulation of
sediments for toxfcological testing. ASTM Desig-
nation E 1391 - 90. American Society of Testing
and Materials, West Conshohocken, PA,
. 1998 (draft). Standard test methods for
measuring the toxicity of sediment-associated
contaminants with fresh water invertebrates.
ASTM Designation E 1706 - 95b. American
Society of Testing and Materials, West
Conshohocken, PA.
Bahls, L. 1993. Periphyton bioassessment
methods for Montana streams. Montana Water
Quality Bureau, Department of Health and
Environmental Science, Helena, MT.
Baker, J.P. and S.W. Christensen. 1991.
Effects of acidification on biological communi-
ties in aquatic ecosystems. Pages 83-106 In
D.F. Charles, ed. Acidic deposition and aquatic
ecosystems: Regional case studies. Springer-
Verlag, New York, NY.
Barbour, M.T. and J. Gerritsen. 1996.
Subsampling of benthic samples: A defense of
the fixed-count method. Journal of the North
American Benthological Society 15(3):386-391.
Barbour, M.T., J. Gerritsen, G.O. Griffith, R.
Frydenborg, E. McCarron, J.S. White, and
M.L. Bastian. 1996a. A framework for biologi-
cal criteria for Florida streams using benthic
macroinvertebrates. Journal of the North
American Benthological Society 15:185-211.
Barbour, M.T., J. Gerritsen, and J.S. White.
1996b. Development of the Stream Condition
Index (SCI) for Florida. Prepared for Florida
Department of Environmental Protection,
Tallahassee, FL.
Barbour, M.T., J.L. Plafkin, B.P. Bradley, C.G.
Graves, and R.W. Wisseman. 1992. Evalua-
tion of EPA's rapid bioassessment benthic
metrics: Metric redundancy and variability
among reference stream sites. Environmental
Toxicology and Chemistry 11:437-449.
Barbour, M.T., J.B. Stribling and J.R. Karr.
G-1
-------
Appendix G
1995. Multimetric approach for establishing
biocriteria and measuring biological condition.
Pages 63-77 In W.S. Davis and T.P. Simon, eds.
Biological Assessment and Criteria: Tools for
water resource planning and decision making.
Lewis Publishers, Boca Raton, FL.
Barbour, M.T. and S. Thomley. 1990. Quality
assurance issues related to field assessment of
aquatic ecosystems. Pages 167-183 In D. Hart,
ed, Proceedings of the Third Annual Ecological
Quality Assurance Workshop. Burlington,
Ontario.
Barko, J.W., H.L. Eakin, and D.G. McFarland.
1992. Interrelationships between macrophyte
growth and sediment nutrient availability.
Abstracts, NALMS 12th Annual International
Symposium. North American Lake Management
Society, Madison, WI.
Battarbee, R.W. 1973. A new method for the
estimation of absolute microfossil numbers, with
reference especially to diatoms. Limnology and
Oceanography 18:647-633.
Birks, H.J.B., S. Juggins, and J.M. Line. 1990.
Lake surface-water chemistry reconstructions
from paleolimnological data. Pages 301-313 In
B. J. Mason, ed. The Surface Waters Acidifica-
tion Programme. Cambridge University Press,
Cambridge, MA.
Blalock, H.M., Jr. 1979. Social Statistics.
Revised 2nd edition. McGraw-Hill Book Com-
pany, New York, NY.
Brooks, J.L. and S.I. Dodson. 1965. Predation,
body size and composition of plankton. Science
150:28-35.
Camburn, K.E., J.C. Kingston, and D.F.
Charles. 1984-1986. PIRLA Diatom Iconograph.
PIRLA unpublished report series, Report no. 3,
Indiana University, Bloomington, IN.
Canfield, D.E., Jr. 1981. Chemical and trophic
state characteristics of Florida lakes in relation to
regional geology. Final report to Florida Coop-
erative Fish and Wildlife Unit, University of
Florida, Gainesville, FL.
Canfield. D.E., Jr., K.A. Langeland. M.J.
Maceina, W.T. Haller, J.V. Shireman. and J.R.
Jones. 1983. Trophic state classification of
lakes with aquatic macrophytes. Canadian
Journal of Fisheries arid Aquatic Sciences
40:1713-1718.
Caper, J.. G. Power, and F.R. Shivers, Jr. 1983.
Chesapeake waters, pollution, public health and
public opinion. Tidewater Publishers,
Centreville, MD.
Carlson, R.E. 1977. A trophic state index for
lakes. Limnology and Oceanography 22:361-
369.
Carlson, RE. 1992. Expanding the trophic
state concept to identify non-nutrient limited
lakes and reservoirs. Pages 59-71 In Enhancing
the Stale's Lake Management Programs: Proceed-
ings of a conference, Chicago, 1991. North
American Lake Management Society, Madison,
WI.
Carlson, R.E. and J. Simpson. 1996. Volunteer
Lake Monitoring in the Upper Mid-west Programs,
Techniques, and Technical Recommendations.
North American Lake Management Society,
Madison, WI.
Carpenter, S.R., J.F. Kitchell, J.R. Hodgson, P.A.
Cochran, J.A. Elser, M.M. Elser, D.M. Lodge, D.
Kretchmer, X. He, and C.N. von Ende. 1987.
Regulation of lake primary productivity by food
web structure. Ecology 68:1863-1876.
Charles, D.F. and J.P. Smol. 1994. Long-term
chemical changes in lakes: Quantitative infer-
ences using biotic remains in the sediment
record. Advances in Chemistry 237:1-51.
Charles, D.F., J.P. Smol, and D.R Engstrom.
1994. Paleolimnological approaches to biologi-
cal monitoring. Pages 233-293 In L. L. Loeb,
and A. Spacle, eds. Biological monitoring of
aquatic systems. CRC Press, Boca Raton, FL.
Charles, D.F. and D.R Whitehead. 1986. The
PIRLA project: Paleoecological investigation of
recent lake acidification. Hydrobiologia 143:13-
20.
Cochran, W.G. 1963. Sampling Techniques.
John Wiley and Sons, Inc., New York, NY.
Conover, W.J. 1980. Practical Nonparametric
Statistics. 2nd edition. John Wiley and Sons,
Inc., New York, NY.
Conquest, L.L., S.C. Ralph, and R.J. Naiman.
1994. Implementation of large-scale stream
8-2
-------
Literature Cited
monitoring efforts: Sampling design and data
analysis issues. Pages 69-90 In L. Loeb and A.
Spacie, eds. Biological Monitoring of Aquatic
Systems, Lewis Publishers, Boca Raton, FL.
Courtemanch, D.L., S.P. Davies, and E.B.
Laverty. 1989. Incorporation of biological
information in water quality planning. Environ-
mental Management 13:35-41.
Crowder, A. and D.S. Painter. 1991. Sub-
merged macrophytes in Lake Ontario: Current
knowledge, importance, threats to stability, and
needed studies. Canadian Journal of Fisheries
and Aquatic Sciences 48:1539-1545.
Cumming, B.F., J.P. Smol, J.C. Kingston, D.F.
Charles, H.J.B. Birks, K.E. Cambum, S.S. Dixit,
A.J. Uutala, and A.R. Selle. 1992. How much
acidification has occurred in Adirondack
regionlakes (New York, USA) since preindustrial
times? Canadian Journal of Fisheries and
Aquatic Sciences 49:128-141.
Davie, R.D. and J.E. DeShon. 1989. The Ohio
Lake Condition Index: A new multiparameter
approach to lake classification. Lake Research
Management 5:1-8.
Davies, S.P., L.T Somides, D.L. Courtemanch,
and F. Drummond. 1993. Maine biological
monitoring and biocriteria development program.
Maine Department of Environmental Protection,
Bureau of Water Quality Control, Division of
Environmental Evaluation and Lake Studies,
Augusta, ME.
DeBemardi, R. 1984. Methods for the estima-
tion of zooplankton abundance. Pages 59-86 In
J.A. Downing and F.H. Rigler, eds. A manual on
methods for the assessment of secondary produc-
tivity in fresh waters. 2nd edition. Blackwell
Scientific Publications, Oxford, England.
DeShon, J.E. 1995. Development and applica-
tion of the Invertebrate Community Index (ICI).
Pages 217-244 In W.S. Davis and T.P. Simon,
eds. Biological Assessment and Criteria. Tools
for Water Resource Planning and Decision
Making. CRC Press, Boca Raton, FL.
Dillon, P.J. and F.H. Rigler. 1974. The phos-
phorus-chlorophyll relationship in lakes.
Limnology and Oceanography 19:767-773.
Dillon, P.J., W.A. Scheider. R.A. Reid, and D.S.
Jeffries. 1994. Lakeshore capacity study: Part
1-Test of effects of shoreline development on the
trophic status of lakes. Lake Reservation
Management 8(2): 121-129.
Dionne, M. and J.R. Karr. 1992. Ecological
monitoring of fish assemblages in Tennessee
River reservoirs. Pages 259-281 In D.H.
McKenzie, D.E. Hyatt, and V.J. McDonald, eds.
Ecological indicators. Elsevier, New York, NY.
Dixit, S.S., B.F. Cumming, H.J.B. Birks, J.P.
Smol, J.C. Kingston, A.J. Uutala, D.F. Charles,
and K.E. Cambum. 1993. Diatom assemblages
from Adirondack lakes (New York, USA) and the
development of inference models for retrospec-
tive environmental assessment. Journal of
Palaeolimnology 8:27-47.
Dixit, S.S., J.P. Smol, J.C. Kingston, and D.F.
Charles. 1992. Diatoms: Powerful indicators of
environmental change. Environmental Science
Technology 26:23-33.
Dodson, S.I. 1992. Predicting crustacean
zooplankton richness. Limnology and
Oceanography 37:848-856.
Donigian, A.S., Jr. and W.C. Huber. 1991.
Modeling of nonpoint source water quality in
urbanand non-urban areas. EPA 68-03-
3513(#29). U.S. Environmental Protection
Agency, Athens, GA.
Dycus, D.L. and D.L. Meinert. 1992. Reservoir
vital signs monitoring - 1991: Summary of vital
signs and use impairment monitoring on Tennes-
see Valley Reservoirs. TVA/WR—92/8. Water
Resources Division, Tennessee Valley Authority,
Chattanooga, TN.
. 1993. Monitoring and evaluation of
aquatic resource health and use suitability in
Tennessee Valley Authority reservoirs. Draft.
Water Management. Tennessee Valley Authority,
Chattanooga, TN.
. 1994. Tennessee Valley reservoir and
stream quality 1993 summary of vital signs and
use suitability monitoring - Volume 1. Water
Mangement Tennessee Valley Authority,
Chattanooga, TN.
Eakins, J.D. and R.I. Morrison. 1978. A new
procedure for the determination of 210Pb in lake
and marine sediments. International Journal of
G-3
-------
Appendix G
Radio-Isotopes 29:531-536.
Edmondson, W.T. and J.T. Lehman. 1981. The
effect of changes in the nutrient income on the
condition of Lake Washington. Limnology and
Oceanography 26:1-29.
Edmondson, W.T. and A.H. Litt. 1982. Daph-
nia in Lake Washington. Limnology and Ocean-
ography 27:272-293.
Engle, V.D., J.K. Summers, and G.R. Gaston.
1994. A benthic index of environmental condi-
tion of Gulf of Mexico estuaries. Estuaries
17:372-384.
Environment Canada. 1994. Development of
methodologies for whole-sediment toxicity tests
with benthic invertebrates: a. Optimum age for
initiation and duration of Hyallela azteca survival
and growth test, fa. Formulation of artificial
sedimentfs) for use in reference toxicity tests with
benthic invertebrates. Environment Canada
Method Development and Application Division
Technology Development Directorate. TS - 27.
Fairweather, P.G. 1991. Statistical power and
design requirements for environmental monitor-
ing. Australian Journal of Marine Freshwater
Research 42:555-67.
Fausch, D.D., J.R. Karr, and P.R. Yant. 1984.
Regional application of an index of biotic integ-
rity based on stream fish communities. Trans-
actions of the American Fisheries Society
113:39-55.
FDEP. 1994. Lake bioassessments for the
determination of nonpoint source impairment in
Florida. Draft. Florida Department of Environ-
mental Protection, Biology Section, Division of
Administrative and Technical Services, Tallahas-
see, FL.
. 1996. Standard operating procedures for
biological assessment Florida Department of
Environmental Protection, Biology Section,
Tallahassee, FL.
Fore, L.S., J.R. Karr, and L.L. Conquest. 1994.
Statistical properties of an index of biotic
integrity used to evaluate water resources.
Canadian Journal of Fisheries and Aquatic
Sciences 51:1077-1087.
Fore, L.S., J.R. Karr, and R.W. Wisseman.
1996. Assessing invertebrate responses to
human activities: Evaluating alternative ap-
proaches. Journal of the North American
Benthological Society 15:212-231.
Frey, D.G. 1975. Biological integrity of water -
An historical approach. Pages 127-140 In R.K.
Ballentine and L.J. Guarrai, eds. The Integrity
of Water, U.S. Environmental Protection Agency,
Washington, DC.
Fritz, S.C. 1990. Twentieth-centuiy salinity
and water-level fluctuations in Devils Lake,
North Dakota: Test of a diatom-based transfer
function. Limnology and Oceanography
35:1771-1780.
Fiyer, G. 1985. Crustacean diversity in relation
to the size of waterbodies: Some facts and
problems. Freshwater Biology 15:347-361.
Gauch, H.G. 1982. Multivariate analysis in
community ecology. Cambridge University Press,
Cambridge, MA.
Gerritsen, J. 1980. Adaptive responses to
encounter problems. Pages 52-62 In W.C.
Kerfoot, ed. The evolution and ecology of zoop-
lankton communities. University Press of New
England, Hanover, NH.
Gerritsen, J. 1995. Additive biological indices
for resource management. Journal of the North
American Benthological Society 14:451-457.
Gerritsen, J., and M.L. Bowman. 1994. Periph-
ytic diatom assemblages on high elevation Rocky
Mountain lakes: Characterization of reference
conditions. Prepared for the Montana Depart-
ment of Health and Environmental Quality,
Helena, MT and the U.S. Environmental Protec-
tion Agency, Washington, D.C.
Gerritsen, J., and J.S. White. 1997. Develop-
ment of a biological index for Florida Lakes.
Prepared for Florida Department of Environmen-
tal Protection, Tallahassee, FL.
Glew, J.R. 1988. A portable extruding device
for close interval sectioning of unconsolidated
core samples. Journal of Palaeolimnology
1:235-239.
. 1989. A new trigger mechanism for
sediment samplers. Journal of Palaeolimnology
2:241-243.
Gray, J.S. 1989. Effects of environmental
G-4
-------
Literature Cited
stress on species fish assemblages. Biological
Journal of the Linriean Society 37:19-32.
Green, G.R. and B.A. Beckman. 1992. The
composite index of coincident indicators and
alternative coincident indexes. Survey of
Current Business. U.S. Department of Com-
merce, Bureau of Economic Analysis, Washing-
ton, DC.
Gurtz, M.E. 1994. Design of biological compo-
nents of the National Water-Quality Assessment
(NAWQA) Program. Pages 323-354 (Chapter 15)
In S.L. Loeb and A. Spacie, eds. Biological
Monitoring of Aquatic Systems. Lewis Publish-
ers, Boca Raton, FL.
Hall, R.I. and J.P. Smol. 1992. A weighted-
averaging regression and calibration model for
inferring total phosphorus concentration from
diatoms in British Columbia (Canada) lakes.
Freshwater Biology 27:417-434.
Hecky, R. E. and H. J. Kling. 1981. The
phytoplankton and protozooplankton of the
euphotic zone of Lake Tanganyika: Species
composition, biomass, chlorophyll content, and
spatio-temporal distribution. Limnology arid
Oceanography 26:532-547.
Heiskary, S.A. 1989. Integrating ecoregion
concepts into state lake management programs.
Pages 89-100 In Enhancing the States' Lake
Management Programs Proceedings. North
American Lake Management Society, Madison,
WI.
Heiskary, S.A. and C.B. Wilson. 1989. The
regional nature of lake water quality across
Minnesota: An analysis for improving resource
management. Journal of the Minnesota Academy
of Science 55:71-77.
Heiskary, S.A., C.B. Wilson, and D.P. Larsen,
1987. Analysis of regional patterns in lake
water quality: Using ecoregions for lake man-
agement in Minnesota. Lake Reservation
Management 3:337-344.
Helsel, D.R. and R.M. Hirsch. 1992. Statistical
Methods in Water Resources. Elsevier,
Amsterdam.
Hergenrader, G.L. and D.C. Lessig. 1980.
Eutrophication of the salt valley reservoirs,
1968-1973, III. The macroinvertebrate commu-
nity: Its development, composition, and change
in response to eutrophication. Hydrobiologia
75:7-25.
Hickman, G.D. and T.A. McDonough. 1996.
Assessing the reservoir fish assemblage index:
A potential measure of reservoir quality. Ameri-
can. Fisheries Society Symposium 16:85-97.
Hill, K.R. 1986. Classification of Iowa lakes
and their fish standing stock. Lake Reservation
Management 2:105-109.
Hill, M.O. 1979. TW1NSPAN-A FORTRAN
program for arranging multivariate data in an
ordered two-way table by classification of
individuals and attributes. Cornell University,
Ithaca, NY.
Hilsenhoff, W.L. 1987. An improved biotic
index of organic stream pollution. Great Lakes
Entomology 20:31-39.
Hosper, S.H. and M.L. Meijer. 1993.
Biomanipulati on, will it work for your lake? A
simple test for the assessment of chances for
clear water, following drastic fish-stock reduc-
tion in shallow, eutrophic lakes. Ecological
Engineering 2:63-72.
Hosper, S.H., M.L. Meijer, and P.A. Walker, eds.
1992. Handleiding Actief Biologisch Beheer.
RIZA/OVB (Netherlands Institute for Inland
Water Management and Wastewater Treatment/
Organization for Improvement of Inland Fisher-
ies) Lelystad. The Netherlands (in Dutch).
Hughes, R.M., SJi. Heiskaiy, W.J. Matthews, and
C.O. Yoder. 1994. Use of ecoregions in biological
monitoring. Pages 125-151 In S.L. Loeb and A.
Spacie, eds. Biological monitoring of aquatic
systems. Lewis Publishers, Boca Raton, FL.
Hughes, R.M. and D.P. Larsen. 1988.
Ecoregions: An approach to surface water
protection. Journal of Water Pollution Control
Federation 60:486-493.
Hurlbert, S.H. 1971. The nonconcept of species
diversity: A critique and alternative parameters.
Ecology 52:577-586.
Hurlbert, S.H. 1984. Pseudoreplication and the
design of ecological field experiments. Ecological
Monography 54(2).
Hutchinson, G.E. 1957. A treatise on limnology.
Vol 1. Geography, physics and chemistry. John
G-5
-------
Appendix G
Wiley and Sons, New York, NY.
Hutchinson, G.E. 1965. The ecological theater
and the evolutionary phxy. Yale University Press,
New Haven, CT.
Hutchinson, G.E. 1967, A treatise on limnology.
Vol 2, Introduction to lake biology and. the
llmnqplankton. John Wiley and Sons, New York,
NY.
Hynes, H.B.N. 1994. Historical perspective and
future direction of biological monitoring of
aquatic systems. Chapter 2 In S.L. Loeb and A.
Spacie, eds. Biological monitoring of aquatic
systems. CRC Press, Boca Raton, FL.
1TFM. 1992. Ambient water-quality monitoring
in the United States. First year review, evalua-
tion, and recommendations. ITFM. Interagency
Advisoiy Committee on Water Data. Intergov-
ernmental Task Force on Monitoring Water
Qualify, Water Information Coordination Pro-
gram, Washington, DC.
Intergovernmental Task Force on Monitoring
Water Qualify (ITFM). 1994. Water Quality
Monitoring in the United States (Region V): 1993
Report of the Intergovernmental Task Force on
Monitoring Water Quality. Intergovernmental
Task Force on Monitoring Water Qualify, Wash-
ington, DC.
Jackson, D.A. 1993a. Multivariate analysis of
benthic invertebrate communities; The implica-
tions of choosing particular data standardiza-
tions, measures of association, and ordination
methods. Hydrobiologia 268: 9-26
Jackson, D.A. 1993b. Stopping rules in
principal components analysis: A comparison of
heuristical and statistical approaches. Ecology
74: 2204-2214.
Jennings, M.J.. L.S. Fore, and J.R. Karr. 1995.
Biological monitoring of fish assemblages in
Tennessee Valley reservoirs. Regulated Rivers:
Research & Management 11:263-274.
Johnson, R.A. and D.W. Wichern. 1992.
Applied Multivariate Statistical Analysis. John
Wiley & Sons. New York, NY.
Johnson, R.J. 1989. Water qualify standards
for lakes. Pages 123-128 In G.H. Flock, ed.
Water quality standards for the 21st Century.
U.S. Environmental Protection Agency, Office of
Water, Washington, DC.
Jongman, R.H.G., C.J.F. ter Braak, and O.F.R.
vanTongeren. 1987. Data analysis and commu-
nity and landscape ecology. Pudoc. Wageningen,
the Netherlands.
Karr, J.R. 1981. Assessment of biotic integrity
using fish communities. Fisheries 6:21-27.
. 1991. Biological integrity: Along-ne-
glected aspect of water resource management.
Ecological Application 1:66-84.
Karr, J.R. and D.R. Dudley. 1981. Ecological
perspective on water quality goals. Environmen-
tal Management 5:55-68.
Karr, J.R, K.D. Fausch, P.L. Angermeier, P.R.
Yant, and I.J. Schlosser. 1986. Assessing
biological integrity in running waters: A method
and its rationale. Illinois Natural History Survey,
Special Publication 5.
Kenkel, N.C. and Orloci, L. 1986. Applying
metric and nonmetric multidimensional scaling
to ecological studies: Some new results. Ecol-
ogy 67: 919-928.
Kennedy, R.H., K.W. Thornton, and D. Ford.
1985. Characterization of the reservoir ecosys-
tem. In D. Gunnison, ed. Microbial processes in
reservoirs. Dr. W. Junk Publishers, Boston, MA.
Kennedy, R.H. and W.W. Walker. 1990. Reser-
voir nutrient dynamics. Pages 109-132 In K.W.
Thornton, B.L. Kimmel and F.E. Payne, eds.
Reservoir limnology: Ecological perspective.
John Wiley and Sons, New York, NY.
KDEP. 1993. Methods for assessing biological
integrity of surface waters. Kentucky Depart-
ment of Environmental Protection, Division of
Water, Frankfort, KY.
Kerans, B.L., and J.R, Karr. 1994. A Benthic
Index of Biotic Integrity (B-IBI) for rivers of the
Tennessee Valley. Ecological Application 4:768-
785.
Kingston, J.C. 1986. Diatom analysis. In D.F.
Charles and D.R, Whitehead, eds. Paleoecologi-
cal investigation of recent lake acidification:
Methods and project descriptions. EA-4906, Res.
Proj. 2174-10. Electric Power Research Insti-
tute, Palo Alto. CA.
G-6
-------
Literature Cited
Knowlton, M,F. and J.R. Jones, 1989. Com-
parison of surface and depth-integrated com-
posite samples for estimating algal biomass and
phosphorus values and notes on the vertical
distribution of algae and photosynthetic bacteria
in midwestern lakes, Archiv fur Hydrobiologte.
Supplement 83:175-96.
Kratzer, C.R. and P.L. Brezonik. 1981. A
Carlson-type trophic state index for nitrogen in
Florida lakes. Water Research Bulletin 17:713-
715.
Lahiri, K. and G.H. Moore. 1991. Leading
economic indicators: New approaches and
forecasting records. Cambridge University Press,
Cambridge, MA.
Larsen, D.P. and H. Mercier. 1976. Phospho-
rus retention capacity of lakes. Journal of
Fisheries Research Board of Canada 33:742-
750.
Leach, J.H. and R.C. Herron. 1992. A review of
lake habitat classification. Chapter 3 In W.D.N.
Busch and P.G. Sly, eds. The development of an
aquatic habitat classification system for lakes.
CRC Press, Boca Raton, FL.
Leonard, P.M. and D.J. Orth. 1986. Applica-
tion and testing of an index of biotie integrity in
small, coolwater streams. Transaction of Ameri-
can Fisheries Society 115:401-414.
Lewis, W.M., Jr. 1979. Zooplarikton community
analysts. Springer-Verlag. New York, NY.
Line, J.M., C.J.F. ter Braak, and H.J.B. Birks.
1994. WACALIB version 3.3-a computer pro-
gram to reconstruct environmental variables
from fossil assemblages by weighted averaging
and to derive sample-specific errors of predic-
tion. Journal of Palaeolimnology 10:147- 152.
Ludwig, J.A. and J.F. Reynolds. 1988. Statisti-
cal ecology. John Wiley and Sons, New York,
NY.
Martin, R.G. and R.D. Arneson. 1978. Com-
parative limnology of a deep-discharge reservoir
and a surface-discharge lake on the Madison
River, Montana. Freshwater Biology 8:33-42.
Masters, A.E. 1992. Reservoir vital signs
monitoring, 1991 - Benthic macroinvertebrate
community results. TVA/WR-92/3. Tennessee
Valley Authority, Water Resources Division,
Chattanooga, TN.
Maughan, J.T. 1993. Ecological assessment of
hazardous waste sites. Van Nostrand Reinhold,
New York. NY.
Maxted, J.R., E.L. Dickey, and G.M. Mitchell.
1994. Habitat quality of a Delaware nontidal
stream. Delaware Department of Natural
Resources and Environmental Control, Dover,
DE.
Mayewski, P.A., M. Wumkes, J. Klinck, M.S.
Twickler, J.S. Putscher, K.C. Taylor, A.J. Gow,
E.D. Waddington, R.B. Alley, J.E. Gibb, P.M.
Grootes, D.A. Meese, M. Ram, M. Whalen, and
A.T. Wilson. 1994. Record drilling depth struck
in Greenland. Eos 75:113-124.
Mazumder, A. 1994. Patterns of algal biomass
in dominant odd- vs. even-link lake ecosystems.
Ecology 75(4): 1141-1149.
Mazumder, A., W.D. Taylor, D.J. McQueen, and
D.R.S. Lean. 1990. Effect of fish and plankton
on lake temperature and mixing depth. Science
247(1):312-315.
Meador, M.R., T.F. CufTney and M.E. Gurtz.
1993. Methods for sampling Jlsh communities as
part of the National Water Quality Assessment
Program. USGS/OFR 93-104, U.S. Geological
Survey, Washington, DC.
Meinert, D.L., S.R, Butkus, and T.A.
McDonough. 1992. Chickamauga Reservoir
Embayment Study - 1990. TVA/WR-92/28.
Tennessee Valley Authority, Water Resources
Division, Chattanooga, TN.
Miller, D.L., P.M. Leonard, R.M. Hughes, J.R.
Karr, P.B. Moyle, L.H. Schradcr, B.A. Thomp-
son, R.A. Daniels, K.D. Fausch, G.A. Fitzhugh,
J.R. Gammon, D.B. Halliwell, P.L. Angermeier,
and D.J. Orth. 1988. Regional applications of
an index of biotic integrity for use in water
resource management. Fisheries 13:12-20.
Mills. E.L., D.M. Green, and A. Schiavone, Jr.
1987. Use of zooplankton size to assess the
community structure of fish populations in
freshwater lakes. North American Journal of
Fisheries Management 7:369-378.
Mills, E.L. and A. Schiavone, Jr. 1982. Evalua-
G-7
-------
Appendix G
tion of fish communities through assessment of
zooplankton populations and measures of lake
productivity. North American Journal of Fisher-
ies Management 2:14-27.
Moore, J.W. 1981. Factors influencing the
species composition, distribution and abun-
dance of bcnthic invertebrates in the profundal
zone of a eutrophic northern lake.
Hydroblohogia 85:505-510.
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. Freshwater Biology 41-52.
MALMS. 1992. Developing eutrophicatlon stan-
dards for lakes and reservoirs. North American
Lake Management Society, Alachua, FL.
Naumann, E. 1932. Grundzuge der regionalen
limnologie. Binnengewasser 11.
Norris, R.H. 1995. Biological monitoring: The
dilemma of data analysis. Journal of North
American Benthological Society 14:440-450.
NRC. 1992. Restoration of aquatic ecosystems.
National Research Council. National Academy
Press, Washington, DC.
Numberg, G.K. 1996. Trophic state of clear
and colored, soft- and hardwater lakes with
special consideration of nutrients, anoxia,
phytoplankton and fish. Lake and Reservation
Management 12(4):432-447.
O'Brien, W.J. 1979. The predator-prey interac-
tion of planktivorous fish and zooplankton.
American Science 67:572-581.
. 1990. Perspectives on fish in reservoir
ecosystems. Pages 209-226 In K.W. Thornton,
B.L. Kimmel, and F.E. Payneln, eds. Reservoir
limnology: Ecological perspectives. John Wiley
and Sons, New York, NY.
O'Gorman, R., E.L. Mills, and J.S. DeGisi.
1991. Use of zooplankton to assess the move-
ment and distribution of alewife (Atosa
pseudoharengus) in south-central Lake Ontario
in spring. Canadian Journal of Fisheries Aquatic
Sciences 48:2250-2257.
Oglesby, R.T. 1977. Relationships of fish yield
to lake phytoplankton standing crop, produc-
tion, and morphoedaphie factors. Journal of
Fisheries Research Board of Canada
34(12):2271-2279.
Ohio EPA. 1987. Biological criteria for the
protection of aquatic life. Vols. 1-3. Ohio Envi-
ronmental Protection Agency, Monitoring and
Assessment Program, Surface Water Section,
Division of Water Quality, Columbus, OH.
. 1990. TheuseofbiocriteriaintheOhioEPA
surface water monitoring and assessment
program. Ohio Environmental Protection
Agency, Ecological Assessment Section, Division
of Water Quality Planning and Assessments,
Columbus, OH.
OCC. 1993. Development of rapid
bioassessment protocols for Oklahoma utilizing
characteristics of the diatom community. Okla-
homa Conservation Commission, Oklahoma
City, OK.
Omemik, J.M. 1987. Ecoregions of the conter-
minous United States. Annals of the Association
of American Geography 77(1): 118- 125.
Oreskes, N., K. Schrader-Frechette, and K.
Belitz. 1994. Verification, validation, and
confirmation of numerical models in the earth
sciences. Science 263:641-646.
Orth, R.J. and J.W. Moore. 1983. Chesapeake
Bay: An unprecedented decline in submerged
aquatic vegetation. Science 222:51-53.
Palmer, M.W. 1993. Putting things in even
better order: The advantages of canonical
correspondence analysis. Ecology 74(8):2215-
2230.
Patalas, K. 1971. Crustacean plankton and the
eutrophication of St. Lawrence Great Lakes.
Journal of Fisheries Research Board of Canada
29:1451-1462.
Patrick, R. 1949. A proposed biological mea-
sure of stream conditions, based on a survey of
the Conestoga Basin, Lancaster County, Penn-
sylvania. Proceedings of the Philadelphia Acad-
emy of Natural Sciences 101:277-341.
Periy, J.A. and N.H. Troelstrup. 1988. Whole
ecosystem manipulation: A productive avenue
for test system research? Environmental Toxicol-
ogy and Chemistry 7:941 -951.
G-8
-------
Literature Cited
Peterman, R.M. 1990, The importance of
reporting statistical power: The forest decline
andacidic deposition example. Ecology
71(5):2024-2027.
Peters, J.A. 1988. Quality control infusion into
stationary source sampling. Chapter 22 In L.H.
Keith, ed. In Principles of environmental sampling.
American Chemical Society, Washington, DC.
Peters, RH. 1991. A critique for ecology.
Cambridge University Press, Cambridge, MA.
Pielou, E.C. 1977. Mathematical ecology. 2nd
edition. John Wiley and Sons, Inc., New York.
NY.
Porter, K.G. 1977. The plant-animal interface
in freshwater ecosystems. American Science
65:159-170.
PSEP. 1995. Recommended guidelines for
conducting laboratory bioassays on Puget Sound
sediments. Prepared for U.S. Environmental
Protection Agency Region 10, Office of Puget
Sound. Seattle, WA and Puget Sound Water
Quality Authority, Olympia. WA.
Ranasinghe, J.A., S.B. Weisberg, J. Gerritsen,
and D.M. Dauer. 1994. Assessment of Chesa-
peake Bay benthic macroinvertebrate resource
condition in relation to water quality and water-
shed stressors. Prepared by Versar, Inc. for
Maryland Department of Natural Resources,
Chesapeake Bay Research and Monitoring
Division, Annapolis, MD.
Reynolds, C.S. 1984. The ecology of freshwater
phytoplankton. Cambridge University Press,
Cambridge, MA.
Reynoldson, T.B., R.C. Bailey, K.E. Day, and
R.H. Norris. 1995. Biological guidelines for
freshwater sediment based on BEnthic Assess-
ment of SedimenT (the BEAST) using a multi-
variate approach for predicting biological state.
Australian Journal of Ecology 20:198-219.
Rohm, C.M., J.M. Omeraik, C.W. Kiilsgaard.
1995. Regional patterns of total phosphorus in
lakes of the Northeastern United States. Lake
and Reservoir Management 11(1): 1-14.
Roth, N.E., M.T. Southerland, J.C. Chaillou,
J.H. Volstad, S.B. Weisberg, H.T. Wilson, D.G.
Heimbuch, J.C. Seibel. 1997. Maryland biologi-
cal stream survey: Ecological status of non-tidal
streams in six basins sampled in 1995. CBWP-
MANTA-EA-97-2. Prepared for Maryland
Department of Natural Resources, Chesapeake
Bay and Watershed Programs, Annapolis, MD.
Ryder, R.A., S.R. Kerr, K.H. Loftus, and H.A.
Regier. 1974. The morphoedaphic index, a fish
yield estimator: Review and evaluation. Journal
of Fisheries Research Board of Canada 31:663-
688.
Scheffer, M., M.R. de Redelijkheid, and F.
Noppert. 1992. Distribution and dynamics of
submerged vegetation In a chain of shallow
eutrophic lakes. Aquatic Botany 42:199-216.
Schindler, D.W. 1988. Effects of acid rain on
freshwater ecosystems. Science 239:149-158.
Schindler, D.W., S.E.M. Kasian, and R.H.
Hesslein. 1989. Biological impoverishment in
lakes of the midwestem and northeastern
United States from acid rain. Environmental
Science and Technology 23:573-580.
Schueler, T.R. 1987. Controlling urban runoff:
A practical manual for planning and designing
urban BMPs. Document No. 87703. Metropoli-
tan Washington Council of Governments,
Washington, DC.
Schuster, E.G. and H.R. Zuuring. 1986. Quan-
tifying the unquantifiable: Or have you stopped
abusing measurement scale? Journal of For-
estry 1986:25-30.
Scott, E.M., G.D. Hickman, and A.M. Brown.
1992. Reservoir vital signs monitoring, 1991-
Fish community results. TVA/WR-92/5. Ten-
nessee Valley Authority, Water Resources
Division, Chattanooga, TN.
Scott, L.C., J .A. Ranasinghe, A.T. Shaughnessy,
J. Gerritsen, T.A. Tornatore, R. Newport. 1991.
Long-term benthic monitoring and assessment
program for the Maryland portion of Chesapeake
Bay: Level 1, Comprehensive report. Vol I.
Versar, Inc., Columbia, MD.
Simon, T.P. and J. Lyons. 1995. Application of
the index of biotic integrity to evaluate water
resource integrity in freshwater ecosystems.
Pages 245-262 In W.S. Davis and T.P. Simon,
eds. Biological assessment and criteria. Tools for
water resource planning and decision making.
G-9
-------
Appendix G
CRC Press, Boca Raton, FL.
Smith, V.H. 1983. Low nitrogen to phosphorus
ratios favor dominance by blue-green algae in
lake phytoplankton. Science 221:669-671.
Smol, J.P. 1983. Paleophyeology of a high
Arctic lake near Cape Herschel, Ellesmere
Island. Canadian Journal of Botany 61:2195-
2204.
Smol, J.P. 1986. Chrysophyte analysis, Section
7. In D.F. Charles and D.R. Whitehead, eds.
Paleoecological investigation of recent lake
acidification: Methods and project description.
EA-4906, Research Project 2174-10. Electric
Power Research Institute, Palo Alto, CA.
Smol, J.P., D.F. Charles, and D.R. Whitehead.
1984. Mallomonadacean microfossils provide
evidence of recent lake acidification. Nature
307(5952) :628-630.
Smol, J.P., B.F. Cumming, and J.C. Kingston.
1990. QA in paleolimnology. In D.R. Hart, ed.
QA guidelines for biology in aquatic environmen-
tal protection. Prepared for Environment
Canada, National Water Research Institute,
Burlington, Ont.
Sokal, R.R. and F.J. Rohlf. 1969. Biometry.
W.H. Freeman & Co., San Francisco, CA.
Southerland, M.T., and J.B. Stribling. 1995.
Status of biological criteria development and
implementation. Pages 81-96 In W.S. Davis and
T.P. Simon, eds. Biological assessment and
criteria; Tools for water resource planning and
decision making. CRC Press, Boca Raton, FL.
Southwood, T.R.E. 1977. Habitat, the templet
for ecological strategies? Journal of Animal
Ecology 46:337-65.
Snedecor, G.W. and W.G. Cochran. 1980.
Statistical Methods. 7s1 edition. The Iowa State
University Press, Ames, IA.
Steedman, R.J. 1988. Modification and assess-
ment of an index of biotic integrity to quantify
stream quality in southern Ontario. Canadian
Journal of Fisheries and Aquatic Sciences
45:492-501.
Steinhart, C.E., L.4. Schierow, and W.C.
Sonzogni. 1982. An environmental quality
index for the Great Lakes. Water Resources
Bulletin 18:1025-1031.
Stemberger, R.S. and J.M. Lazorchak. 1994.
Zooplankton assemblage responses to distur-
bance gradients. Canadian Journal of Fisheries
and Aquatic Sciences 51:2535-2547.
Stribling, J.B., J. Lathrop-Davis, M.T. Barbour,
J.S. White, and E.W. Leppo. 1995. Evaluation
of environmental indicators for the wetlands of
Montana: The multimetric approach using benthic
macroinvertebrates. Prepared for the State of
Montana, Department of Health and Environ-
mental Sciences, Helena, Montana.
Suter, G.W., II. 1993. Ecological risk assess-
ment Lewis Publishers, Boca Raton, FL.
SYSTAT. Inc. 1992. SYSTAT: Statistics,
Version 5.2 edition. SYSTAT, Inc., Evanston, IL.
ter Braak. C.J.F, 1986. Canonical correspon-
dence analysis: A new eigenvector technique for
multivariate direct gradient analysis. Ecology
67:1167-1179.
. 1988. CANOCO - A FORTRAN program for
canonical community ordination. Agricultural
Mathematics Group, Wageningen, The Nether-
lands.
ter Braak, C.J.F. and S. Juggins. 1993.
Weighted averaging partial least squares
regression(WA-PLS): An improved method for
reconstructing environmental variables from
species assemblages. Hydrobiologia 269/
270:485-502.
Thompson, S.K. 1992, Sampling. John Wiley
and Sons, New York, NY.
Thornton, K.W. 1990a. Perspectives on reser-
voir limnology. Pages 1-14 In K.W. Thornton,
B.L. Kimmel and F.E. Payne, eds. Reservoir
limnology: Ecological perspectives. John Wiley
and Sons, New York, NY.
. 1990b. Sedimentary process. Pages 43-
69 In K.W. Thornton, B.L. Kimmel, and F.E.
Payne, eds. Reservoir Limnology: Ecological
Perspectives. John Wiley and Sons Inc., New
York, NY.
Tilman, D. 1982. Resource competition and
community structure. Princeton University Press,
Princeton, NJ.
G-10
-------
Literature Cited
Trebitz, A.S., S.A. Nichols, S.R. Carpenter, and
R.C. Lathrop. 1993. Patterns of vegetation
change in Lake Wingra following a Myriophyllum
spicatum decline. Aquatic Botany 46:325-340.
Tukey, J. W. 1977. Exploratory data analysis.
Addison-Wesley, Reading, MA.
TVA. 1994. Tennessee Valley reservoir and
stream quality - 1993: Summary of vital signs
and use suitability monitoring. Vol. I. Water
Management Tennessee Valley Authority, Water
Resources Division, Chattanooga, TN.
. 1995. Aquatic ecological health determi-
nations for TVA reservoirs - 1994: An informal
summary of 1994 vital signs monitoring results
and ecological health determination methods.
Water Management Tennessee Valley Authority,
Water Resources Division, Chattanooga, TN.
Uhlmann, D. 1971. Influence of dilution,
sinking and grazing rate on phytoplankton
populations of hyperfertilized ponds and micro-
ecosystems. Mitt. Intemat Verein. Umonl.
19:100-124.
Underwood, A.J. 1994. On beyond BACI:
Sampling designs that might reliably detect
environmental disturbances. Ecological Applica-
tions 4(1):3-15.
USEPA. 1980a. Clean lakes program guidance
manual EPA 440-5-81-003. U.S. Environmen-
tal Protection Agency, Office of Water, Washing-
ton, DC.
. 1980b. National accomplishments in
pollution control 1970-1980: Some case histories.
U.S. Environmental Protection Agency, Office of
Planning and Management, Program Evaluation
Division, Washington, DC.
. 1983. Water Quality Standards Handbook.
U.S. Environmental Protection Agency, Office of
Water Regulations and Standards, Washington, DC.
, 1984. Technical support manual:
Waterbody surveys and assessments for conduct-
ing use attainability analyses. Vol 111: Lake
Systems. U.S. Environmental Protection
Agency, Office of Water Regulations and Stan-
dards, Washington, DC.
. 1985a. Clean lakes program: A review of
the first decade. EPA 440/5-85-033. U.S.
Environmental Protection Agency, Office of Water
Regulations and Standards, Washington, DC.
. 1985b. Methods for measuring the acute
toxicity of effluents to freshwater and marine
organisms. EPA 600/4-85/013. U.S. Environ-
mental Protection Agency, Cincinnati, OH.
. 1985c. Technical support document for
water quality-based toxics control EPA 440/4-
85/03. U.S. Environmental Protection Agency,
Washington, DC.
. 1987a. Handbook of methods for acid
deposition studies. EPA 600/4-87/026. U.S.
Environmental Protection Agency, Office of
Research and Development, Washington, DC.
. 1987b. Report of the national workshop on
in-stream biological monitoring and criteria. In-
stream Biol. Criteria Comm. U.S. Environmen-
tal Protection Agency, Office of Water Regula-
tions and Standards, Environmental Research
Laboratory, Region 10, Corvallis, OR,
. 1987c. Surface water monitoring: A
framework for change. U.S. Environmental
Protection Agency, Office of Water, Office of
Policy, Planning and Evaluation, Washington,
DC.
. 1989a. Overview of selected EPA regula-
tions and guidance affecting POTW management.
EPA 440/69-89/008. U.S. Environmental
Protection Agency, Office of Water, Washington,
DC.
. 1989b. Rapid bioassessment protocols for
use in streams and rivers: Benthic
macroinvertebrates and fish. EPA/440/4-89-
001. U.S. Environmental Protection Agency,
Office of Water, Washington, DC.
. 1990a. Biological criteria: National
program guidance for surface waters. EPA 440/
5-90-004, U.S. Environmental Protection
Agency, Washington, DC.
. 1990b. Lake and reservoir restoration
guidance manual. 2nd edition. EPA 440/4-90-
006. Prepared by North America Lake Manage-
ment Society for U.S. Environmental Protection
Agency, Washington, DC.
. 1990c. Macroinvertebrate field and
laboratory methods for evaluating the
G-11
-------
Appendix G
biologicalintegrity of surface waters. EPA 600/4-
90/030. U.S. Environmental Protection Agency,
Washington, DC.
. 1990d. Near coastal program plan for
1990: Estuaries. EPA 600/4-90/033. U.S.
Environmental Protection Agency, Office of
Research and Development, Environmental
Research Lab, Narragansett, RI.
. 1991a. Biological criteria; Guide to
technical literature. EPA 440/5-91-004. U.S.
Environmental Protection Agency, Office of
Water, Washington, DC.
. 1991b. Biological criteria: Research and
regulation. Proceedings of a symposium. EPA
440/5-91-005. U.S. Environmental Protection
Agency, Office of Water, Health and Ecological
Criteria Division, Washington, DC.
. 1991c. Biological criteria: State develop-
ment and implementation efforts. EPA 440/5-91-
003. U.S. Environmental Protection Agency,
Office of Water, Washington, DC.
. 199Id. Design Report for EMAP, the
Environmental Monitoring and Assessment
Program. EPA 600-3-91-053. U.S. Environmen-
tal Protection Agency, Office of Research and
Development, Washington, DC.
. 199 le. The Environmental Monitoring and
Assessment Program (EMAP) - Surface Waters
monitoring research plan, fiscal year 1991. EPA
600/3-91/022. U.S. Environmental Protection
Agency, Corvallis, OR.
. 199 If. Technical support document for
water quality-based toxics control EPA 505/2-
90-001. U.S. Environmental Protection Agency,
Office of Water, Washington, DC.
. 1992a. Draft ecoregions/subecoregions of
Florida. U.S. Environmental Protection Agency,
Office of Research and Development, Environ-
mental Research Laboratory, Corvallis, OR
. 1992b. Fish field and laboratory methods
for evaluating the biological integrity of surface
waters. EPA 600/R-92/ 111. U.S. Environmen-
tal Protection Agency, Cincinnati, OH.
. 1992c. Framework for ecological risk
assessment EPA 630-R-92-001. U.S. Environ-
mental Protection Agency, Washington, DC.
. 1992d. Monitoring Guidance of the Na-
tional Estuary Program. EPA 842-B-92-004.
U.S. Environmental Protection Agency, Office of
Water, Washington, DC.
. 1992e. Procedures for initiating narrative
biological criteria. EPA 822-B-92-002.
U.S.Environmental Protection Agency, Office of
Science and Technology, Washington, DC.
. 1992f. Sediment classification methods
compendium. EPA 823-4-92-006. U.S. Environ-
mental Protection Agency, Office of Water.
Washington, DC.
. 1993a. EMAP-Surface waters 1991 pilot
report Review draft EPA 620/R-93/003. U.S.
Environmental Protection Agency, Environmen-
tal Research Laboratory, Corvallis. OR.
. 1993b (draft final). EPA requirements for
quality assurance project plans for environmental
data operations. EPA QA/R-5. U.S. Environ-
mental Protection Agency, Quality Assurance
Management Staff, Washington, DC.
. 1993c. Guidelines for preparation of the
1994 state water quality assessments (305(b)
reports). EPA 841-B-93-004. U.S. Environmen-
tal Protection Agency, Office of Water, Washing-
ton, DC.
. 1993d. Statistical methods for the analysis
of lake water quality trends. EPA 841-R-93-003.
U.S. Environmental Protection Agency, Office of
Water, Washington, D.C.
. 1993e. Virginian Province demonstration
report. EMAP-Estuaries: 1990. EPA 620/R-93/
006. U.S. Environmental Protection Agency,
Office of Research and Development, Washing-
ton, DC.
. 1994a. EMAP Surface waters field opera-
tions manual for lakes. U.S. Environmental
Protection Agency, Environmental Monitoring
Systems Laboratory, Las Vegas, NV.
. 1994b. Environmental Monitoring and
Assessment Program: Integrated quality assur-
ance project plan for the Surface Waters Resource
Group, 1994 activities. Rev. 2.00. EPA 600/X-
91/080. U.S. Environmental Protection Agency,
Las Vegas, NV.
. 1994c (draft). EPA program guidance on
G-12
-------
Literature Cited
the TMDL concept U.S. Environmental Protec-
tion Agency, Washington, DC.
. 1994d. Guidance for the Data Quality
Objectives Process. EPA 600/R-96/055. U.S.
Environmental Protection Agency, Office of
Research and Development, Washington, DC.
. 1994e (draft). Guidance on implementation
of biological criteria. U.S. Environmental Protec-
tion Agency, Office of Science and Technology,
Washington, DC.
. 1994f. National water quality inventory:
1992 report to Congress. PA 841 -R-94-001.
U.S. Environmental Protection Agency, Office of
Water, Washington, DC.
. 1994g. Periphytic diatom assemblages of
high elevation Rocky Mountain lakes: Character-
ization of reference conditions. U.S. Environ-
mental Protection Agency, Washington, DC.
. 1994h. Statistical summary: EMAP-
Estuaries Virginian Province - 1991. EPA 620/R-
94/005. U.S. Environmental Protection Agency,
Office of Research and Development, Environ-
mental Research Laboratory, Narragansett, RI.
. 1994i. Statistical summary: EMAP -
Estuaries Virginian Province - 1992. EPA 620/R-
94/019. U.S. Environmental Protection Agency,
Office of Research and Development, Environ-
mental Research Laboratory, Narragansett, RI.
. 1994j. USEPA procedures for assessing
the toxicity and bioaccumulation of sediment-
associated contaminants with freshwater inverte-
brates. EPA 600/R-94/024. U.S. Environmen-
tal Protection Agency, Environmental Research
Laboratory, Duluth, MN.
. 1994k (draft). The watershed protection
approach 1993/94 supplement to 1992 annual
report. U.S. Environmental Protection Agency,
Office of Water, Washington, DC.
. 1995. Generic quality assurance project
plan guidance for programs using community
level biological assessment in wadable streams
and rivers. EPA 841-B-95-004. U.S. Environ-
mental Protection Agency, Office of Water,
Washington, DC.
. 1996a. Biological criteria: Technical
guidance for streams and small rivers. EPA 822-
B-96-001. U.S. Environmental Protection
Agency, Office of Water, Washington, DC.
. 1996b. Biological criteria: Technical
guidance for survey design and statistical
evaluation of biosurvey data. Prepared for U.S.
Environmental Protection Agency, Office of
Science and Technology, Washington, DC.
. 1996c. Guidance for data quality assess-
ment, practical methods for data analysis, EPA
QA/G-9, QA96 Version. EPA 600/R-96/084.
U.S. Environmental Protection Agency, Office of
Research and Development, Washington, DC.
. 1996d. Proposed guidelines for ecological
risk assessment. U.S. Environmental Protection
Agency. Federal Register 61(175):47552-47631.
. 1996e. Biological assessment methods,
biocriteria, and biological indicators: Bibliography
of selected technical, policy, and regulatory
literature. EPA 230-B-96-001. U.S. Environ-
mental Protection Agency, Office of Policy,
Planning, and Evaluation, Washington, DC.
. 1996f. Summary of state biological assess-
ment programs for streams and rivers. EPA 230-
R-96-007. U.S. Environmental Protection
Agency, Office of Policy, Planning, and Evalua-
tion, Washington, DC.
. 1997a (review draft). Estuarine/near
coastal marine waters bioassessment and
biocriteria technical guidance. Prepared for U.S.
Environmental Protection Agency, Office of
Water, Washington DC.
. 1997b. Field and laboratory methods for
macroinvertebrate and habitat assessment of low
gradient nontidal streams. Mid-Atlantic Coastal
Streams Workgroup. U.S. Environmental
Protection Agency, Environmental Services
Division, Region 3, Wheeling, WV.
. 1997c (draft). Lake regions of Florida.
U.S. Environmental Protection Agency, Office of
Research and Development, Western Ecology
Division, Corvallis, OR.
. 1997d (draft). Revision to rapid
bioassessment protocols for use in streams and
rivers: Periphyton, benthic macroinvertebrates,
and fish EPA 841-D-97-002. U.S. Environ-
mental Protection Agency, Office of Water,
G-13
-------
Appendix G
Washington, DC.
Vlghi, M. and G. Chiaudani. 1985. A simple
method to estimate lake phosphorus concentra-
tions resulting from natural, background,
loadings. Water Research 19:987-991.
Vinson, M.R., and C.P. Hawkins. 1996. Effects
of sampling area and subsampling procedure on
comparisons of taxa richness among streams.
Journal of the North American Benthological
Society 15:392-299.
Vollenweider, R.A. 1968. Scientific fundamen-
tals of eutrophicatton of lakes and flowing waters
with special reference to phosphorus and nitro-
gen. OECD/DAS/CSI/68.27. OECD, Paris.
. 1975. Input-output models with special
reference to the phosphorus loading concept in
limnology. Schwetzerische Zeltschrift fur
Hydrologie 37:53-83.
Wallace, J.B.. J.W. Grubaugh, and M.R. Whiles.
1996. Biotic indices arid stream ecosystem
processes: Results from an experimental study.
Ecological Applications 6(1):140-151.
Warwick, R.M., and K.R. Clarke. 1991. A
comparison of some methods for analyzing
changes in benthic community structure. Jour-
nal of Marine Biological Association 71:225-244.
Watson, N.H.F. 1974. Zooplankton of the St.
Lawrence Great Lakes: Species composition,
distribution, and abundance. Journal of the
Fisheries Research Board of Ckmada 31:783-794.
Weaver, M.J., J.J. Magnuson, and M.K. Clayton.
1993. Analyses for differentiating littoral fish
assemblages with catch data from multiple
sampling gears. Transactions of the American
Fisheries Society 122:1111-1119.
Weber, S.P., B. Shaw, and S.A. Nichols. 1995.
The aquatic macrophyte communities of eight
northern Wisconsin flowages. Technical Report,
College of Natural Resources, University of
Wisconsin, Stevens Point, WI.
Wetzel, R. G. 1975. Limnology. W.B. Saunders
Company, Philadelphia, PA.
Wetzel, R.G. 1990. Reservoir Ecosystems:
Conclusions and speculations. Pages 227-238
In K.W. Thornton, B.L. Kimmel, and F.E. Payne,
eds. Reservoir limnology; Ecological perspec-
tives. John Wiley and Sons, New York, NY.
Wiederholm, T. 1980. Use of benthos in lake
monitoring. Journal of the Water Pollution
Control Federation 52:537-547.
. 1984. Responses of aquatic insects to
environmental pollution. Pages 508-557 In V.H.
Resh and D.M. Rosenberg, eds. The ecology of
aquatic insects. Praeger, New York, NY,
Wonnacott, T.H. and R.J. Wonnacott. 1969.
Introductory Statistics. 2nd Edition. John Wiley
and Sons, Inc., New York, NY.
Wright, J.C. 1967. Effect of impoundments on
productivity, water chemistry, and heat budgets
of rivers. Pages 188-199 In Reservoir Fishery
Resources. American Fisheries Society. Wash-
ington, DC.
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.
Freshwater Biology 14:221-256.
Yoccoz, N.G. 1991. Use, overuse and misuse of
significance tests in evolutionary biology and
ecology. Bulletin of the Ecological Society of
America 72:106-111.
Yoder, C.O. and E.T. Rankin. 1995. Biological
criteria program development and implementation
in Ohio. Pages 109-144 In W.S. Davis and T.P.
Simon, eds. Biological Assessment and Criteria:
Tools for water resource planning and decision
making. Lewis Publishers, Boca Raton, FL.
Zaret, T.M. 1980. Predation and freshwater
communities. Yale University Press, New Haven,
CT.
G-14
------- |