&EPA
United States
Environmental Protection
Agency
Office of Water
(WH - 556F)
EPA 503/8-91-002
August 1991
Monitoring Guidance for the
National Estuary Program
Interim Final
saw*
printed on recycled paper
-------
MONITORING GUIDANCE FOR THE
NATIONAL ESTUARY PROGRAM
Interim Final
Office of Wetlands, Oceans, and Watersheds
U.S. Environmental Protection Agency
Washington, D.C. 20460
August, 1991
-------
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
\
WASHINGTON, D.C. 20460
AUG 2 9 1991
OFFICE OF
WATER
MEMORANDUM
SUBJECT: Monitoring GuiiQanc4tor the National Estuary Program
FROM: C^Marian Miay^Dirbctor
(P "^Oceans a'rid Coas*4l Pr
1 Protection Division (WH-556F)
Enclosed is the interim final guidance document on
environmental monitoring in EPA's National Estuary Program (NEP).
This document describes the development of environmental
monitoring programs to assess the effectiveness of Comprehensive
Conservation and Management Plans (CCMPs) implemented as part of
the NEP.
Section 320 of the Clean Water Act requires the development
of monitoring programs to assess CCMP effectiveness. A
monitoring plan must therefore be submitted to EPA for approval
with each completed CCMP. This guidance document outlines the
required components of NEP monitoring plans, and discusses the
key technical and programmatic issues to be addressed in
monitoring program design and implementation. The document also
describes the methods most frequently used in estuary monitoring
programs.
This document has been reviewed by EPA's Science Advisory
Board (SAB) and other experts within and outside of EPA. The SAB
has endorsed the approach to monitoring program design and
performance evaluation outlined in the document. The guidance is
being printed in interim final form pending receipt of the final
EPA Science Advisory Board report early in fiscal year 1992. All
program requirements contained in the document are, however,
effective immediately. Any additional SAB comments will be
addressed in a final guidance document. We would like to thank
the many reviewers who offered valuable comments on earlier
drafts of the document.
Printed on Recycled Paper
-------
Acknowledgments
Technical support for development of this document was provided by Tetra
Tech, Inc. under EPA Contract No. 68-C1-0008. EPA would like to thank the
many reviewers in who offered valuable comments on earlier drafts of this
document. Key reviewers included Dr. John Armstrong of EPA Region 10
members of EPA's Science Advisory Board, members of Ae National Oceanic
and Atmospheric Administration's Interagency Ecosystems Monitoring
Workgroup, and program directors and staff of National Estuary Program
Management Conferences.
-------
-------
Ill
Table of Contents
1.0 Introduction 1
1.1 Background 2
1.2 Recommended Monitoring Design Procedures 6
1.3 Monitoring Program Management 12
2.0 Develop Monitoring Objectives and Performance Objectives 17
2.1 Monitoring Program Objectives 17
2.2 Performance Criteria 20
2.3 Additional Guidance 20
3.0 Establish Testable Hypotheses and Select Statistical Methods 23
3.1 Establish Testable Hypotheses 23
3.2 Selection of Statistical Method 26
4.0 Select Analytical Methods and Alternative Sampling Designs 29
4.1 Selection of Field and Laboratory Methods 29
4.2 Alternative Sampling Layouts 30
4.3 Use of Existing Monitoring Programs 33
5.0 Evaluate Monitoring Program Performance 39
5.1 Evaluate the Expected Performance of Individual Monitoring Program Components 40
5.2 Evaluate Overall Program Performance 41
5.3 Statistical Power Analysis Methods 41
6.0 Implement Monitoring Study and Data Analysis 47
6.1 Data Management 48
6.2 Data Analysis 50
7.0 Communicate Program Results 51
8.0 References 55
Appendices
A Case Studies
Al.O The Puget Sound Ambient Monitoring Program Case Study A-3
Al.l Purpose and Approach A-3
A1.2 Development of PS AMP: Institutional Arrangements A-4
-------
IV
A1.3 The Puget Sound Ambient Monitoring Program A-9
A1.4 Implementation of PSAMP and Cost A-13
A1.5 Summary and Recommendations A-19
A1.6 References A-23
A2.0 Chesapeake Bay Monitoring Program: Detection of Trends in Estuaries A-25
A2.1 Purpose and Approach A-25
A2.2 Chesapeake Bay Program A-25
A2.3 Evaluation of Monitoring Program Performance A-29
A2.4 Use of Power Analysis Results A-40
A2.5 References A-40
B Methods
B.I Methods - Introduction B-3
B.I Methods Chapter Format B-3
B.2 Quality Assurance/Quality Control(QA/QC) Considerations B-5
B.3 Statistical Design Considerations B-9
B.4 Literature Cited and References B-12
Bl.O Water Column Physical Characteristics B-15
Bl.l Rationale B-15
B1.2 Monitoring Design Considerations B-15
B1.3 Existing Analytical Methods B-16
B1.4 QA/QC Considerations B-21
B1.5 Statistical Design Considerations B-24
B1.6 Use of Data B-25
B1.7 Summary and Recommendations B-25
B1.8 Literature Cited and References B-28
B2.0 Water Column Chemistry B-31
B2.1 Rationale B-31
B2.2 Monitoring Design Considerations B-31
B2.3 Existing Analytical Methods B-33
B2.4 QA/QC Considerations B-37
B2.5 Statistical Design Considerations B-40
B2.6 Use of Data B-40
B2.7 Summary and Recommendations B-41
B2.8 Literature Cited and References B-43
-------
B3.0 Sediment Grain Size B-47
B3.1 Rationale B-47
B3.2 Monitoring Design Considerations B-47
B3.3 Existing Analytical Methods B-50
B3.4 QA/QC Considerations B-52
B3.5 Statistical Design Considerations B-52
B3.6 Use of Data B-52
B3.7 Summary and Recommendations B-53
B3.8 Literature Cited and References B-55
B4.0 Sediment Chemistry B-57
B4.1 Rationale B-57
B4.2 Monitoring Design Considerations B-57
B4.3 Existing Analytical Methods B-61
B4.4 QA/QC Considerations B-63
B4.5 Statistical Design Considerations B-69
B4.6 Use of Data B-69
B4.7 Summary and Recommendations B-70
B4.8 Literature Cited and References B-72
B5.0 Plankton: Biomass, Productivity and Community Structure/Function B-75
B5.1 Rationale B-75
B5.2 Monitoring Design Considerations B-75
B5.3 Existing Analytical Methods B-78
B5.4 QA/QC Considerations B-81
B5.5 Statistical Design Considerations B-81
B5.6 Use of Data B-82
B5.7 Summary and Recommendations B-82
B5.8 Literature Cited and References B-84
B6.0 Aquatic Vegetation B-93
B6.1 Rationale , B-94
B6.2 Monitoring Design Considerations B-94
B6.3 Existing Analytical Methods B-98
B6.4 QA/QC Considerations B-101
B6.5 Statistical Design Considerations B-101
B6.6 Use of Data B-102
B6.7 Summary and Recommendations B-103
B6.8 Literature Cited and References B-105
-------
B7.0 Benthic Infauna Community Structure B-l 11
B7.1 Rationale B-l 11
B7.2 Monitoring Design Considerations B-l 11
B7.3 Existing Analytical Methods B-l 18
B7.4 QA/QC Considerations B-125
B7.5 Statistical Design Considerations B-127
B7.6 Use of Data B-127
B7.7 Summary and Recommendations B-128
B7.8 Literature Cited and References B-130
B8.0 Fish Community Structure B-137
B8.1 Rationale B-137
B8.2 Monitoring Design Considerations B-137
B8.3 Analytical Methods Considerations B-139
B8.4 QA/QC Considerations B-144
B8.5 Statistical Design Considerations B-145
B8.6 Use of Data B-145
B8.7 Summary and Recommendations B-146
B8.8 Literature Cited and References B-147
B9.0 Fish and Shellfish Pathobiology B-151
B9.1 Rationale B-151
B9.2 Monitoring Design Considerations B-153
B9.3 Existing Analytical Methods B-156
B9.4 QA/QC Considerations B-161
B9.5 Statistical Design Considerations B-162
B9.6 Use of Data B-163
B9.7 Summary and Recommendations B-166
B9.8 Literature Cited and References B-170
B10.0 Bioaccumulation B-l81
B10.1 Rationale B-181
B10.2 Monitoring Design Considerations B-l82
B10.3 Existing Analytical Methods B-191
B10.4 QA/QC Considerations B-193
B 10.5 Statistical Design Considerations B-197
B10.6 Use of Data B-197
B10.7 Summary and Recommendations B-197
B10.8 Literature Cited and References B-202
-------
VII
Bll.O Bacterial and Viral Pathogens B-207
Bll.l Rationale B-207
B11.2 Monitoring Design Considerations B-208
B11.3 Analytical Methods Considerations B-209
B11.4 QA/QC Considerations B-216
B11.5 Statistical Design Considerations B-216
B11.6 Data Use B-216
B11.7 Summary and Recommendations B-217
B11.8 Literature Cited and References B-219
-------
VIII
-------
IX
Tables
Table Page
1-1. Elements of Systems Approach 7
2-1. Puget Sound Ambient Monitoring Program Design Specifications for
Sediment Chemistry Sampling 19
3-1. Example Monitoring Program Objectives and Associated Questions 24
3-2. References to Basic Monitoring Design and Statistical Texts 27
Al-1. Alternatives for Improving Puget Sound Monitoring A-6
Al-2. MMC Activities Required by PSWQA in Developing PSAMP A-8
Al-3. PSAMP Design and 1989-90 Activities A-12
Al-4. PSAMP Task Assignments by Agency A-15
A2-1. Seasonal Kendall Tau Result A-37
B-l. Sampling Methods Described B-3
B-2. Format of Methods Section B-4
B-3. Technical Support and Guidance Documents B-6
B-4. Monitoring Topics Included in Puget Sound Estuary Program Protocols B-8
Bl-1. List of Methods and Equipment B-17
Bl-2. Recommended Sample Preservation and Storage Requirements B-22
Bl-3. Recommended Analytical Methods B-23
B2-1. List of Existing Analytical Techniques B-34
B2-2. Sample Preservation and Storage Parameters B-38
B2-3. Definitions for Selected Limits of Detection B-39
B3-1. Sediment Grain Size: Withdrawal Times for Pipet Analysis as a Function of
Particle Size and Water Temperature B-51
B4-1. List of Existing Analytical Techniques B-61
B4-2. Summary of Sample Collection and Preparation QA/QC Requirements for Organic Compounds B-66
B4-3. Sample Preservation and Storage Parameters B-67
B4-4. Summary of Quality Control Sample B-68
B4-5. Summary of Warning and Control Limits for Quality Control Sample B-69
B6-1. List of Terms B-93
B6-2. List of Analytical Methods B-98
B7-1. Biological Indices B-l 19
B8-1. Biological Indices B-141
B9-1. List of Pathobiological Terms B-152
B9-2. Highest Ranking Candidate Fishes for Use as Pathobiology Monitoring Species B-154
B10-1. List of Terms B-181
BIO-2. Highest Ranking Candidate Fishes for Use as Bioaccumulation Monitoring Species B-184
-------
Table Page
B10-3. Recommended Large Macroinvertebrate Species for Bioaccumulation Monitoring B-186
B10-4. List of Existing Analytical Techniques B-193
B10-5. Sample Preservation and Storage Paramters B-194
B10-6. Summary of Sample Collection and Preparation QA/QC Requirements for
Organic Compounds B-195
B10-7. Summary of Quality Control Sample B-198
B10-8. Summary of Warning and Control Limits for Quality Control Sample B-199
Bll-1. Microorganisms Responsible for Causing Adverse Human Health Effects B-210
Bll-2. Laboratory Procedures for Bacterial Indicators B-213
-------
XI
Figures
Figure Page
1-1. Monitoring Program Design 8
1-2. Impacts on the Marine Environment of the Southern California Bight 10
4-1. Description of Various Sampling Methods 31
5-1. Hypothesis Testing: Possible Circumstances and Test Outcomes 42
5-2. Minimum Detectable Difference vs. Number of Replicates for Fixed Set of
Design Parameters 44
5-3. Power vs. Minimum Detectable Difference 45
7-1. Sample Cover of Puget Sound Notes 53
7-2. Sample Cover of Chesapeake Bay Barometer 54
7-3. Sample Covers of Santa Monica Bay Restoration Project Reports 54
A1 -1. Events Leading to the Development and Implementation of the Puget Sound
Ambient Monitoring Program A-4
Al-2. Area Included in Puget Sound Ambient Monitoring Program A-ll
A2-1. Station Locations Included in Data Subset of the Chesapeake B ay
Ambient Monitoring Program A-31
A2-2. Example Box Plot A-33
A2-3. Time Series for N as NH3 for Surface Stations A-34
A2-4. Time Series for Dissolved O2 for Bottom Stations A-34
A2-5. Box Plot for N as NH3 for Surface Stations A-35
A2-6. Box Plot for Dissolved 02 for Bottom Stations A-35
A2-7. Power Analysis for N as NH3 for Surface Stations A-39
A2-8. Power Analysis for Dissolved 02 for Bottom Stations A-39
B4-1. Examples of Acceptable and Unacceptable Samples B-65
B7-1. Generalized SAB Diagram of Changes Along a Gradient of Organic Enrichment B-121
B7-2. Diagram of Changes in Fauna and Sediment Structure Along a Gradient of Organic Enrichment B-123
B7-3. Examples of Acceptable and Unacceptable Samples B-126
-------
XII
-------
1.0 Introduction
The Clean Water Act as amended by the Water Quality Act of 1987 establishes the
National Estuary Program (NEP) to promote long term planning and management in
nationally significant estuaries threatened by pollution, development, or overuse.
Section 320 of the Clean Water Act describes the establishment of a management
conference in each estuary to develop a Comprehensive Conservation and Manage-
ment Plan. It also establishes requirements to monitor the effectiveness of actions
taken pursuant to the plan.
This document provides guidance on the design, implementation and evaluation of the
required monitoring programs. The intended audience is the members of the manage-
ment conference (i.e., the Management Committee, Scientific and Technical Advisory
Committee, and the Citizens Advisory Committee) and the program coordinators and
scientific staff of the individual estuary programs. The purpose of this document is to
identify the steps involved in developing and implementing estuarine monitoring
programs. Given the large number of participants in the management conference and
the diversity of technical backgrounds, it is also intended to provide a technical basis
for discussions on the development of monitoring program objectives, the selection of
monitoring program components, and the allocation of sampling effort. Some of the
issues addressed in this document are:
What is the role of monitoring in the estuary programs?
Why are monitoring programs necessary, and what are examples of goals
and objectives?
What criteria should be used to select components of the monitoring
program, and what should drive the decision making process?
What is the importance of historical data, and how can ongoing monitoring
programs be incorporated into the estuary monitoring program?
What is the relationship between estuary characterization and the monitor-
ing program?
The key technical and programmatic aspects associated with each of these issues are
described. This document also includes several examples. Two case studies are used
to provide examples from existing estuarine monitoring programs. The first case study
from the Puget Sound Estuary Program (Appendix Al .0) provides an example of the
process of developing an effective monitoring strategy. Key issues addressed include
problem and goal definition, the process of monitoring program design and implemen-
-------
tation, and options for funding estuary monitoring programs. The first case study also
demonstrates how ongoing monitoring studies can be coordinated to develop a com-
prehensive basin-wide monitoring program. The second case study (Appendix A2.0)
provides a detailed example of the application of methods for determining the effec-
tiveness and feasibility of monitoring efforts in the Chesapeake Bay Program. Ex-
amples from other environmental monitoring programs are also described.
Appendix B of this document describes existing methods that may be used to monitor
the effectiveness of management actions in the estuary. The National Estuary Program
does not impose national requirements for standard methods. However, emphasis is
placed on the importance of using standardized monitoring protocols within each
estuary and developing performance-based criteria to evaluate the comparability of
analytical methods. One long-term goal of the NEP is to develop a standardized set of
key variables which, if measured in all NEP programs, will provide comparative data
on nationally significant estuaries.
Limited federal funding is available for National Estuary Program monitoring activi-
ties. Therefore, this document discusses the integration of existing monitoring efforts
into the estuary monitoring program. It is also essential that National Estuary Program
monitoring activities be coordinated with existing federal agency status and trends
monitoring programs such as EPA's Ecosystem Monitoring and Assessment Program
(EMAP), the National Oceanic and Atmospheric Administration's Status and Trends
Program, and the U.S. Geological Survey's National Water Quality Assessment
Program. If individual estuary monitoring programs will be measuring the same
parameters as these federal programs, and the protocols in use provide adequate data,
estuary programs should work to ensure compatibility of their data with the existing
federal programs.
Data management, effective data analysis, and the communication of monitoring
program results to a wide range of audiences at several technical levels is also essential
to the success of the estuary program. This document discusses data management
strategies and provides several examples of ongoing efforts to disseminate information
developed by the National Estuary Program.
National Estuary Program 1.1
Background
The purpose of the NEP is to identify estuaries of national significance and to promote
the preparation of comprehensive management plans to ensure their ecological integ-
rity. The process of identifying nationally significant estuaries and nominating estuar-
-------
ies to become part of the NEP is described in A Primer for Establishing and Managing
Estuary Projects (U.S. EPA, 1989a) and Final Guidance on the Contents of a
Governor's Nomination (U.S. EPA, 1990a). Following selection of an estuary to be
included in the National Estuary Program, a management conference is convened by
the EPA Administrator. The management conference has the responsibility to imple-
ment a four-phased program:
Phase I - The Planning Initiative. The planning phase is intended to build
the management framework for identifying and solving problems and
defining the steps in the decision making process.
Phase II - Characterization and Problem Definition. The goal of estuary
characterization is to gather and summarize the existing knowledge con-
cerning the state of the estuary as well as the physical, chemical and
biological factors controlling spatial and temporal changes. The collection
of new data may also be required to develop an understanding of problems
and their causes. The characterization process focuses on identifying
existing and potential problems, gaps in knowledge, and approaches that
could be used to develop the necessary information to fill these data gaps.
Estuary characterization is intended to provide an understanding of the
underlying processes and to describe the cause and effect linkages between
human activities and environmental change. Environmental problems are
described in a manner that provides decision makers with the information
necessary to develop priorities, set management strategies, and devise
restoration measures. The characterization process is described in a
separate EPA guidance document.
Phase III - Development of a Comprehensive Conservation and Manage-
ment Plan (CCMP). The CCMP is a major product of the estuary program.
It is developed by the management conference to summarize findings,
identify and prioritize problems, determine environmental quality goals
and objectives, identify action plans and compliance schedules for pollu-
tion control and resource management, and to ensure that designated uses
of the estuary are protected.
Phase IV - CCMP Implementation. The management conference also has
the responsibility for the coordinated implementation of the CCMP. The
development of a monitoring program to evaluate the effectiveness of
actions specified in the CCMP is a required task of the management
conference.
-------
Environmental sampling is required in Phases II and IV of the estuary programs. The
studies conducted in Phase II are focused on filling identified data gaps, providing
baseline data on point and nonpoint loadings of pollutants, and developing estimates of
the degree of spatial and temporal variability. Sampling is conducted in Phase IV of
the estuary programs as part of a long term environmental monitoring strategy.
Environmental monitoring that is conducted during Phase IV is considered to be
different from the sampling that is conducted during Phase II. Monitoring involves
repeated sampling over time. For example, short term sampling may be conducted in
Phase II to collect specific information on the concentration, distribution and variabil-
ity of chemical contaminants in sediments. The goal of the corresponding sampling
that is conducted during Phase IV is to evaluate trends in monitored variables and to
link the observed patterns to specific management actions. A second distinction
between Phase II sampling and Phase IV monitoring is that comprehensive environ-
mental monitoring programs conducted during Phase IV will require the integration of
information from several concurrent sampling efforts. Since environmental sampling
is costly and resources will be limited, there will be a need to evaluate the efficacy of
different monitoring program components and to allocate sampling efforts accord-
ingly. It will also be necessary to carefully plan and coordinate monitoring efforts
among individual monitoring components and other preexisting monitoring programs
to determine interactions and streamline monitoring efforts. A third distinction
between environmental sampling and monitoring is the need to periodically analyze
the monitoring program results and modify the level of sampling effort to maximize
program effectiveness.
NEP Monitoring Program Goals
The two primary goals of the monitoring programs implemented in Phase IV by the
management conferences are:
to measure the effectiveness of management actions and programs imple-
mented under the CCMP
to provide essential information that can be used to redirect and refocus the
management plan.
These monitoring programs are an essential part of the management program review
and evaluation process for each estuary, and they will be conducted throughout the
implementation of the CCMP.
-------
This document describes the steps involved in designing a monitoring program to meet
these two interrelated goals. However, secondary goals may also be incorporated in
the monitoring program. For example, the goals of the Puget Sound Ambient Monitor-
ing Program (described in Case Study 1) include the characterization and interpretation
of spatial and temporal patterns of conditions in Puget Sound, development of a
permanent record of significant natural and human-caused changes in key environmen-
tal indicators, and the support of research activities through the availability of consis-
tent, scientifically valid data. This document provides guidance on developing an
effective monitoring program to meet these secondary goals as well.
Monitoring Design Issues
Marine monitoring is the continued, systematic time-series observation of
predetermined pollutants or pertinent components of the ecosystem over a
period of time sufficient t6 determine (1) the existing level, (2) trends, and
(3) natural variations of measured components (NOAA,1979).
Recent reports by the National Research Council (NRC, 1990a and 1990b) evaluated
current marine monitoring programs and practices, identified needed improvements in
monitoring strategies, and made a series of recommendations to improve the useful-
ness of monitoring information.
The extensive NRC review of monitoring practices found numerous inadequacies in
the design and use of monitoring data. Three broad problem areas were identified:
Monitoring programs are often poorly designed. While some of the
identified problems could be attributed to the inherent difficulty of separat-
ing the effects of human activities from natural variability, the primary
deficiencies in current monitoring practices resulted from the failure to
clearly define monitoring objectives and to apply available design tools.
There is a lack of communication and coordination among the regulatory,
scientific and management entities sponsoring or conducting monitoring
programs. Specific concerns included the inflexibility of regulatory
requirements that limit opportunities to adapt programs to new needs and
regional objectives. The need to adopt standardized sampling and quality
assurance procedures to ensure data comparability was also identified.
-------
The results of monitoring programs are not presented in a form that is
useful in developing broad public policy or evaluating specific control
strategies. It is essential to link data management strategies and data
analysis methods to the objectives of the monitoring effort. It is also
necessary to devise a plan for effectively communicating monitoring
results to the identified audience.
Additionally, the NRC study found that poorly designed monitoring programs and the
lack of communication and coordination among programs has often resulted in:
Limitation of water quality and public health monitoring to areas adjacent
to point-source outfalls and other known/potential problem areas, and to
the sampling and analysis of conventional water quality and public health
parameters
Living resource monitoring programs restricted to assessment of a few
higher-trophic levels of commercially or recreationally important species,
without consideration of key prey species upon which these species depend
Incomplete monitoring of living resources and habitats key to the establish-
ment of potential declines in living resources and water quality (i.e., near
shore habitats, estuarine wetlands, plankton communities)
A wide range of guidance on the design of environmental monitoring programs is
available in the literature. Important contributions have been made that address the
principles and options for designing monitoring networks (Green, 1979; Segar and
Stamman, 1986), the development of monitoring objectives (Beanlands and Duinker,
1983), the appropriateness of monitoring variables (Bilyard, 1987), and the application
of statistical methods in the design process (Ferraro et a/., 1989). The recent NRC
study (NRC, 1990a) also provides comprehensive review of the steps involved in the
design of marine environmental monitoring programs.
Outlined below is a systems approach to the design of the required NEP monitoring
programs that incorporates existing information and that will ensure the collection,
analysis and reporting of adequate information to meet the goals of individual estuary
programs.
A systems design approach places emphasis on the optimum design of the overall
monitoring program. The essential elements (shown in Table 1-1) are the assessment
1.2
Recommended
Monitoring Design
Procedures
-------
TABLE 1-1. ELEMENTS OF SYSTEMS APPROACH
Define the Objective - the overall objective of the design process is
stated in a succinct manner.
Establish Information Needs - information requirements to meet the
objectives are established.
Establish the Objectives of - the objectives of all possible monitoring
Individual Program Components program components and performance
criteria are established.
Evaluation of Trade-Offs - the combination of monitoring components
that best meet the overall objectives are
selected.
Feedback to Initial Design Step - modifications to the system's design are made
to improve the product's performance.
of trade-offs between individual aspects of the monitoring program and the use of
feedback mechanisms to modify individual monitoring program components based on
periodic assessments of overall program performance. This approach is well suited for
design problems that involve complex, highly variable systems, such as estuaries, and
that involve a large number of investigators that must interact as a group to produce
the product.
The plan for implementing this approach to designing the monitoring program is
shown in Figure 1-1. It involves the specification of monitoring objectives, the
completion of a series of steps to translate these objectives into clearly defined moni-
toring activities, and the use of feedback mechanisms to refine the objectives and
adjust the sampling effort.
The overall objective of monitoring undertaken during Phase IV of the estuary pro-
grams will be to measure the effectiveness of management actions implemented as part
of the CCMP. Meeting this broad program objective will require the specification of
several individual, highly-interrelated monitoring objectives. Examples of individual
monitoring objectives (shown in Figure 1-1 as Objectives 1 through n) include:
determine the response of key water quality variables to management actions; deter-
mine trends in sediment contaminant concentrations; and, evaluate the persistence of
PCBs in the tissue of recreational and commercial fish. The identification of these
specific objectives begins during Estuary Characterization. The characterization
-------
Public
Concerns
Develop
Monitoring Objectives,
Performance Criteria
DEVELOP / REFINE MON TOR NG OBJECTIVES
Establish Testable
Hypotheses and Select
Statistical Methods
Select Analytical Methods
and Alternative
Sampling Designs
Evaluate Expected
Monitoring Study
Performance
EVALUATE / ASSESS PROGRAM PERFORMANCE
Monitoring
Study
Performance
Adequate?
COMMUNICATE
MONITORING
RESULTS/REDIRECT
MANAGEMENT
PROGRAM
Implement
Monitoring Study
and Data Analysis
process identifies public concerns and formulates a series of corresponding manage-
ment issues. During characterization, conceptual and predictive models are developed
and research results are evaluated to provide a basic understanding of important
physical, chemical and biological processes in the estuary. This information is used in
the design of the monitoring program to specify a set of variables and ecological
processes that can be used to detect changes in the estuary in response to management
actions.
The recent reviews of marine monitoring efforts by the National Research Council
(NRC, 1990a and b) described several approaches for defining issues and establishing
monitoring objectives. One of these approaches, adapted from a framework developed
by Clark (1986) for identifying cumulative atmospheric impacts, has been effectively
used in the Southern California Bight to identify impacts and develop monitoring
Figure 1-1. Monitoring
program design.
-------
objectives (NRC, 1990b; Bernstein et al., 1991). The key to this approach is the
construction of the matrix, shown in Figure 1-2, which identifies Valued Ecosystem
Components and sources of perturbations. This matrix also summarizes the under-
standing of the relative importance of each source of perturbation and the degree of
scientific certainty associated with knowledge about each impact. Each cell in the
matrix summarizes the effects of each perturbation on a single ecosystem component.
For example, the information summarized in Figure 1-2 indicates that wastewater
outfalls are a controlling factor of soft bottom benthos communities and that there is a
moderate degree of certainty regarding the scientific understanding of the effects of
these outfalls in the Southern California Bight. The effects of each identified source of
perturbation (e.g., storm water runoff) on all identified resources (VECs) are summa-
rized along a single row. Similarly, each column summarizes the existing knowledge
on the impacts on a single resource caused by the complete range of identified sources
of perturbations.
This matrix summarizes existing information on the resources of the estuary and
potential impacts in an easily accessible manner. The process of developing this
matrix also provides an effective tool for building consensus among the wide range of
interested parties in the estuary program on the relative priority of monitoring objec-
tives and the different components of proposed monitoring programs. Bernstein et al.
(1991) describe an Integrated Assessment methodology based on the work by Clark
(1986) that guides the selection of monitoring objectives based on the relative impor-
tance of the identified resources, the understanding of the underlying controlling
processes, and the ability to detect changes in monitoring variables.
The individual steps involved in designing each component of the monitoring program
are shown on the right hand side of Figure 1-1:
Step 1. Develop Monitoring Objectives and Performance Criteria. Clear
objectives and corresponding performance criteria must be developed for
each component of the monitoring program. Performance criteria specify
the level of change or trend that the monitoring program must be able to
detect. For example, in the Chesapeake Bay Program, a target of 40%
reduction in nutrient loads to the bay was established based on modeling
results. One of the predicted benefits of nutrient reduction is the allevia-
tion of anoxic conditions in the bay and an increase in the minimum
dissolved oxygen to approximately 1.5 mg/l. Therefore, the corresponding
objective and performance criterion for the water quality component of the
monitoring program would be to detect a long term change in dissolved
oxygen concentration in the bottom waters of the bay equal to 1.5 mg/1
-------
10
VALUED
ECOSYSTEM
COMPONENTS
\
SOURCES OF
PERTURBATION
Intertidal
Phytoplankton
|
l
8
!<
CO
Q.
O
a
Soft Bottom Benthos
Hard Bottom Benthos
w
"O
0)
CO
a
&
Storms H7 j (J? q ff?
El Nines ,xP 0 0|(|?
California Current ( vJ7 i W
Upwelling
Blooms/Invasions
;/TS
Ecol. Interactions \]7
Power Plants
Wastewater Outfalls
Dredging
Rivers/Storm Runoff
Commercial Fishing
Sport Fishing
Habitat Loss/Mod.
*
e
ฎ
^3
ฎ
All
ฉ
ฉ
n
'&
^
V3
ฉ
1
g
)
1
1
{
ฎ
^
Wetlands & Estuaries
Commercial Shellfish
r:
W i
U.
o
1
0>
Q.
g
V3
g
^3
^3
^7
1?
w
i
Ti
W
e
i
Demersal Fish J
ฉ (b
ฉ
ฉ
F
Q^w
^
ฎ
W
g
i
&
^l-s'
q?
ฉ
g
w
V]
\P
<37
1
ง
1
08
c/>
D)
C3)
LU
CO
iZ
@
m
&
^
1
1
Marine Mammals |
3
V]
^
^3
1
i
W
o
CD
O
1
Human Health |
q
5
37
<3
^7
^
^3
W
w
Net effect of all sources on each component
q
w
^
^y
^
Net effect of each source on all components
i
KEY
Potential Importance
Q Controlling ^
(~7 Major ^
ry Moderate
i Some
Understanding
r~~1 Moderate
EH] Low
Figure 1-2.
Impacts on the marine
environment of the
Southern California Bight
(Bernstein siaL., 1991).
-------
11
Step 2. Establish Testable Hypotheses and Select Statistical Methods. The
study objectives must be translated into statistically testable hypotheses;
for example, no trend exists in measured values of dissolved oxygen
concentration. Establishing testable hypotheses ensures that the results of
the monitoring program will be unambiguous and that the objectives of the
program can be met. The establishment of testable hypotheses also guides
the development of statistical strategies for determining sample locations
and times as well as the selection of statistical tests that will be used to
analyze the resulting data.
Step 3. Select Analytical Methods and Alternative Sampling Designs.
Detailed specifications for each monitoring variable (measurable endpoint)
of the monitoring program must be developed. These include: field
sampling and laboratory procedures, and QA/QC procedures. Addition-
ally, alternative sampling designs that specify the number and location of
stations must be devised for input to the next step in the design process.
Step 4. Evaluate Expected Monitoring Program Performance. It is essen-
tial to evaluate the expected performance of the initial sampling design to
determine the minimum difference that can be detected over time or
between locations. Without this evaluation there is a risk of collecting and
analyzing too few samples to detect statistically significant temporal or
spatial trends or of analyzing an excessive number of samples (with
associated high costs). As indicated by the feedback loop shown in Fig-
ure 1-1, the results of this evaluation are used to identify modifications to
the initial design in order to increase monitoring program effectiveness.
Information from this evaluation will also be used to assess the ability of
monitoring components to provide information used to modify the manage-
ment plan.
Step 5. Implement Monitoring Study and Data Analysis. The develop-
ment of a data management system is an essential task that is often over-
looked in the design of monitoring programs. The data management
system should be operational prior to implementation of the monitoring
program. Data analysis methods and a timetable for analyzing the data,
assessing CCMP implementation progress and monitoring program perfor-
mance, and reporting program results should also be specified. The results
of the performance assessment are used to refine program objectives and
modify individual study elements to satisfy these objectives.
-------
12
These individual steps in the monitoring design process shown in Figure 1-1 are
described in Sections 2 through 6 of this document.
Overall program performance must be assessed at periodic intervals, and the results
should be used to refine monitoring program objectives and methodologies. The
original monitoring design must remain open to modification. The monitoring pro-
gram should take advantage of new information and innovative sampling approaches
as they are developed, and the link between modeling and monitoring efforts should be
fully exploited. The results of the monitoring program should be used to refine and
modify conceptual and mathematical models of the system, and modeling results
should be used to guide changes in individual monitoring program components,
variables monitored, the frequency of sampling, and overall monitoring strategies. It is
essential that new information, from both independent research and the monitoring
program results, is integrated into the monitoring program.
There are a number of management issues related to the design, implementation, and
maintenance of NEP monitoring programs that must be addressed early by the man-
agement conference. These include setting the timetable for the design and implemen-
tation process, assigning responsibilities for coordinating the design effort, and plan-
ning for the long term success of the monitoring program.
Timetable for the Design and Implementation of the Monitoring Program
The design and implementation of the Puget Sound Ambient Monitoring Program
(PS AMP), described in Case Study 1, required four years of uninterrupted effort. The
development of the monitoring program was a consensus building effort among
numerous agencies and organizations, and several iterations were required to finalize
the initial design plans. In addition to the decisions regarding the basic sampling
strategy, agreement was required on numerous interrelated issues, such as sampling
protocols, appropriate quality assurance/quality control methodology, and the selection
of an information management system.
Planning for the monitoring program should be initiated during the first year of the
management conference. Milestones for the monitoring program effort should be
clearly stated in the State-EPA Management Conference Agreement (a five year action
plan for CCMP completion that is negotiated shortly after a management conference is
convened). Development of the monitoring program should be given a high priority
by the management conference after it is convened. It is important to begin early
1.3
Monitoring
Program
Management
-------
13
planning of the monitoring program to ensure that it is in place at the time the CCMP
is implemented. A detailed monitoring program plan must accompany the CCMP that
is submitted to the EPA Administrator for approval. The monitoring plan must contain
the following elements described in this document:
Definition of program objectives and performance criteria (parameter
values needed to guide management decisions)
Identification of testable hypotheses
Detailed specifications for each monitoring variable, including sampling
locations and frequency, field sampling procedures, field and laboratory
analytical procedures, quality assurance and control procedures
Specification of the data management system and statistical test that will be
used to analyze the monitoring data
Description of the expected performance of the initial sampling design
(i.e., the minimum difference that can be detected in measured variables
over time and between locations)
Timetable for analyzing data and assessing program performance
Management Tasks for Developing the Monitoring Program
The first management task in developing the monitoring program should be the
establishment of an organization or committee, such as a monitoring subcommittee of
the Scientific and Technical Advisory Committee (STAC), to develop the monitoring
program. The membership in the monitoring subcommittee should not be limited to
STAC members but should include representatives from federal, state and local
agencies, universities, industry, environmental groups and others currently conducting
monitoring or planning monitoring programs within the estuary or surrounding waters.
The subcommittee should be charged with the following tasks:
Define the goals and objectives of the monitoring program
Propose an initial design which includes recommendations for sampling
and analytical protocols, data management system specifications, quality
assurance guidelines, data reporting requirements and cost estimates
-------
14
Develop the final monitoring design using workshops and other mecha-
nisms to solicit comments and suggestions from the public and scientific
community
Coordinate the activities of the numerous interested and participating
agencies and develop interagency agreements that will promote the moni-
toring effort
Identify funding mechanisms and opportunities to contain costs
As discussed in Section 1.2, the process of developing a comprehensive monitoring
program must begin with a clear statement of the objectives. The explicit statement of
objectives, and options for obtaining these objectives, is necessary as a starting point
for describing the problem areas in the estuary and developing the consensus among
interested agencies and other parties that is essential to the success of the monitoring
effort.
The primary goals of the estuary program will be to measure the success of the CCMP
and to provide information that can be used to redirect and refocus the management.
However, it will also be the responsibility of this subcommittee to develop secondary
goals and program objectives that will be used to focus the sampling effort. These
objectives could include: continued characterization of spatial and temporal patterns
of change in water quality, sediment and biological resources of the estuary; develop-
ment of a permanent record of significant natural and human caused changes in
environmental indicators in the estuary over time, and support for research activities
through the availability of consistent, scientifically valid data. The process of develop-
ing these objectives is described in Section 2.0.
There are several options for initiating the design process. The Puget Sound Ambient
Monitoring Program began with an initial monitoring design developed by consultants
to EPA Region 10. The initial design included goals and objectives, plans for opera-
tion of the program, and methods for sampling, analysis, and reporting the data. The
first draft of the monitoring program design was refined by EPA Region 10 through a
process involving meetings and discussions among managers and scientists working in
the Puget Sound region. The Puget Sound Monitoring Management Committee then
proceeded to review, modify and refine the proposed monitoring design. This resulted
in the development of a revised draft that was released for public and further scientific
review. The Santa Monica Bay Restoration Project began the development of the
monitoring program with an "Assessment of Monitoring and Data Management Needs
Workshop" to develop and evaluate monitoring options. In preparation for the work-
-------
15
shop, background information was compiled and used to develop a "straw man" to
guide workshop discussions. In both the Puget Sound and the Santa Monica programs,
public and scientific workshops were held to discuss the proposed monitoring options
and to build the required consensus among participating entities.
The next task following the adoption of the monitoring program design is to ensure
that it is implemented as planned. The implementation of a regional or basin-wide
monitoring program requires the development of commitments and the coordination of
activities among the participating agencies. This was achieved in the Puget Sound
Ambient Monitoring Program by negotiating memoranda of agreement (MOA) with
each of the participating agencies. The MO As specified commitments that each
agency carry out its responsibilities identified in the monitoring program design, and
that each agency maintain monitoring program funding levels and staff support to the
monitoring program committees. As described in Case Study 1, the negotiating of the
MOAs was essential to the success of PS AMP, but the complexity of the process was
underestimated. It is recommended that the coordination efforts begin as soon as
possible in the design process.
It is also important that the monitoring subcommittee begin to identify, early in the
process of monitoring program development, the costs associated with existing moni-
toring in the basin, the costs of additional monitoring that will be needed, potential
sources of additional funds to support the monitoring program, and appropriate mecha-
nisms to fund the monitoring program. Costs of the Puget Sound Ambient Monitoring
Program were calculated by a technical costing subcommittee of the monitoring
management committee with the assistance of a technical consultant.
The estimates provided by this subcommittee demonstrate that the costs of comprehen-
sive monitoring programs can be substantial. In addition to the $200,000 in staff and
consultant time required to develop the monitoring program design, the calculated
costs of full implementation of PSAMP are approximately $3.2 million per year (in
1987 dollars). The initial sampling program was reduced in scope due to resource
constraints, and costs for the program have been $250,000 to $350,000 over the first
two years. The projected costs of implementing the monitoring plan submitted as part
of the Buzzards Bay CCMP are $750,000 per year. This includes $200,000 funded by
the estuary program to supplement ongoing state and citizen monitoring programs.
The costs of the comprehensive water quality monitoring in the mainstem of the
Chesapeake Bay are on the order of $900,000 annually. This monitoring includes the
sampling of 20 variables at 50 stations, 20 times. In addition to mainstem sampling,
the states surrounding the bay have undertaken their own extensive sampling programs
in the bay's tributaries.
-------
16
The substantial costs of these programs require equally substantial efforts to secure
long term funding. Most of the costs of monitoring should be borne by the state and
local agencies that will complete specific monitoring tasks. Therefore, the best
opportunity for securing the required funding is indirectly by incorporating existing
federal, state and local monitoring programs into the overall design. The critical step
will be to motivate the participating agencies to make modifications to existing moni-
toring efforts to meet the goals of the estuary program. The best way to achieve this
objective is to involve these agencies in the estuary program and the monitoring design
process and to demonstrate the benefits of a regional / basin-wide monitoring program
to the scientific and regulatory community. If state and local funds are not immedi-
ately available for full implementation of the monitoring program, a plan for phased
implementation can be developed through a priority setting process involving the
monitoring management committee and other technical experts.
Section 4.0 of this document discusses the incorporation of individual components of
ongoing local monitoring programs into the National Estuary Program monitoring
efforts. Opportunities for taking advantage of monitoring programs of national scope,
such as EPA's Environmental Monitoring and Assessment Program (EMAP) and
NOAA's Status and Trends Program, as well as locally coordinated citizen monitoring
programs are also discussed.
Financing Marine and Estuarine Programs: A Guide to Resources (U.S. EPA, 1988a)
provides guidance for obtaining direct financing of marine and estuarine monitoring
programs. This document provides a primer on basic financing concepts and explains
the initiatives needed to begin financial planning for long term resource management
activities. Case studies are also included that provide examples of local financial
efforts.
Planning for the Long Term Success of the Monitoring Program
A structure for managing the monitoring program must be established by the monitor-
ing subcommittee or its equivalent. A lead coordinating agency or organization must
be identified and given the responsibility for implementing the monitoring design as
planned, maintaining the monitoring effort, and reviewing the monitoring program
results. The lead agency or organization should provide a full-time manager/coordina-
tor and a staff that is responsible for keeping all the implementing agencies and other
participants informed of progress, resolving disputes, carrying out technical and
administrative duties, and reporting the results of the program.
-------
17
2.0 Develop Monitoring Objectives and
Performance Criteria
The overall objective of the estuary monitoring program is to determine the
effectiveness of the CCMP. However, this overall objective may encompass several
related goals. For example the estuary characterization process may identify the need
for additional information on spatial and temporal variability in natural resources. The
estuary monitoring program may also be viewed as a regionalized monitoring effort
designed to provide both a measure of the health of the estuary and to provide a
permanent record of changes in the state of the estuary.
Regardless of the scope of the proposed monitoring program, it is essential to develop
explicit statements of the monitoring objectives as well as to establish performance
criteria with which to measure monitoring program success.
DEVELOP/REFINE MONITORING OBJECTIVES
DEVELOP
MONITORING OBJECTIVES,
PERFORMANCE CRITERIA
2.1 Monitoring The development of the monitoring objectives is the culmination of the estuary
Program characterization and the preparation of the CCMP. The characterization process
Objectives should identify public concerns and potential water quality, biological and public
health problems in the estuary. Conceptual and predictive models can be used to
-------
18
summarize the physical, chemical, geological and biological status of the estuary and
identify the factors controlling spatial and temporal changes. The products of the
characterization process should also include the identification of the primary
management issues. The CCMP should set environmental quality goals and objectives
for the estuary and specify action plans for achieving these goals.
Case Study 1 (Appendix A 1.0) describes the process that led to the development of the
goals of the Puget Sound Ambient Monitoring Program. The essential steps were:
The development of the Puget Sound Water Quality Management Plan - a
comprehensive management plan for Puget Sound and its related water-
ways. In developing this plan, nine issue papers were prepared. One of
these, Comprehensive Monitoring of Puget Sound (PSWQA, 1986),
reviewed existing monitoring programs and described the process of
developing a comprehensive water quality, sediment, and biological
monitoring program.
During the same period of time, the Office of Puget Sound (EPA Re-
gion 10) developed a proposed Sound-wide monitoring program which
included goals and objectives, sampling design, operation of the program
and methods for sampling, analysis and reporting of the data (Tetra Tech,
1986a). This document provided a basis for discussions between members
of the Puget Sound Estuary Program (PSEP) technical advisory committee,
scientists, agency staff, and other interested parties.
The Puget Sound Water Quality Management Plan appointed a Monitoring
Management Committee to define the goals and objectives of the monitor-
ing program and to modify the monitoring strategy proposed by the Office
of Puget Sound. Key aspects of the monitoring program development
included workshops that were held for public and scientific peer review,
including the Technical Advisory Committee (TAG) of the Puget Sound
Estuary Program.
The final report of the Monitoring Management Committee (PSWQA, 1988) summa-
rized the proposed monitoring program and provided explicit statements of the goals
and objectives and expected performance. For each monitoring component the ratio-
nale, methods, replication and statistical sensitivity, and a listing of appropriate
sampling protocols were identified. A summary of the design specifications for the
sediment monitoring component of the Puget Sound Ambient Monitoring Program is
presented in Table 2-1. These design specifications provide an example of PS AMP
-------
19
TABLE 2-1. PUGET SOUND AMBIENT MONITORING PROGRAM DESIGN
SPECIFICATIONS FOR SEDIMENT CHEMISTRY SAMPLING (PSWQA, 1988)
Rationale
Measuring toxic chemicals of concern will provide data to:
Assess the potential for sediment toxicity to resident biota.
Identify areas of Puget Sound that have been, or are, accumulating substantial amounts of toxic chemicals.
Evaluate temporal changes of toxic chemicals accumulating in sediments.
Interpret biological and sediment toxicity bioassay data.
Methods
Samples for sediment chemistry will be collected from the upper two centimeters (cm) of sediment, using either a
0.06 m2 box corer or a 0.1 m2 van Veen grab. Three grab samples will be taken at each station and composited. The
same composite will be used for sediment toxicity bioassays and conventional sediment variables. A minimum of the
upper five to 10 cm of sediment will be collected for benthic macro-invertebrate abundance determination. Each
sampling device has advantages and disadvantages. Although a box corer takes a deeper and possibly less disturbed
sample than does a van Veen grab, the box corer is more difficult and more expensive to use. An evaluation of benthic
sampling equipment for use in PS AMP is in progress.
Variables to be monitored will include selected EPA priority pollutant metals and selected EPA priority pollutant
organic compounds, as well as additional compounds of concern in Puget Sound.
Miscellaneous organic acids and volatile organic compounds will be measured only where a suspected source is
present. Intensive surveys conducted by individual agencies under other programs may be triggered by results from
this program.
Tributyl-tin has recently been implicated as a human health risk (U.S. EPA, 1985). Studies from other parts of the
country have shown accumulations in sediments and animal tissue around large marinas and harbors. The present
concern warrants a comprehensive survey for tributyl-tin in sediments and bottom fish tissue in Puget Sound, but it is
not included in the monitoring program at this time due to inconclusive sampling results from other parts of the
country. Periodic spot checks for this and other contaminants are recommended. Costs of such analyses have not been
included in cost estimates for the ambient monitoring program.
Replication and statistical sensitivity
Replicate samples will not be collected for sediment chemistry, thereby precluding statistical analyses among indi-
vidual stations within a survey. Replicated sampling at all stations was precluded because of the high cost of laboratory
analysis and has not been recommended by PSEP sampling and analysis protocols. The variability of sediment
chemistry estimates will be reduced however, by the compositing technique recommended. Field and laboratory
replication will be required for sediment chemistry samples as part of the quality assurance program. Stations for field
replicates will be chosen so as to be representative of certain areas or embayments and sediment types.
Statistical analyses may be performed for related groups (clusters) of stations within a survey or for selected stations
over time.
Replicate data from studies on the chemical composition of sediments within the Commencement Bay waterways
indicated that coefficients of variation for several groups of organic chemicals ranged from 17-61 percent (Tetra Tech,
1985). Given a coefficient of variation of 30 percent and three to four replicates (in space or time), the minimum
detectable difference in mean chemical concentration among stations, at the 95 percent confidence level with a power
of 0.8, would be equal to about 100 percent of the overall mean among stations.
Protocols
* Field and Laboratory References: Tetra Tech (1986b, 1986c, 1986d).
* Supporting literature: U.S. EPA (1983), Plumb (1981), U.S. EPA (1982).
-------
20
efforts to clearly state the monitoring objectives and to establish performance criteria.
Four major objectives are stated for the sediment chemistry component of the monitor-
ing effort: assess the potential of sediment toxicity, identify toxic accumulation areas
within Puget Sound, evaluate temporal changes, and to provide supplemental data that
will be used to interpret bioaccumulation data and bioassay test results.
The establishment of program requirements, i.e. performance criteria, in terms of the
level of precision that is necessary to make decisions regarding the success of the
monitoring effort will define the expectation of the monitoring program. Issues that
must be addressed include: 1) What level of detail will be necessary to make decisions
regarding the success of the CCMP?; 2) What level of difference must be detected in
the monitoring program to initiate modifications in the design and implementation of
the monitoring program? This concept of explicitly stating performance criteria is the
cornerstone of the systems design approach that is described below. The explicit
statement of the monitoring program requirements provides a basis for evaluating
expected and actual monitoring program performance (Section 5.0). Evaluation of the
effectiveness of alternative monitoring designs also provides a basis for discussion of
alternative monitoring approaches and sampling layouts.
Performance criteria are addressed in the design specifications for the sediment
chemistry sampling of the Puget Sound Ambient Monitoring Program (Table 2-1).
Under the discussion of replication and statistical sensitivity, basic information is
provided on background variability and the minimum detectable difference between
groups of stations. However, the specification of performance criteria associated with
monitoring program designs should be made more explicit than those presented in
these design specifications. For example, one of the stated objectives is to evaluate
temporal changes in toxic accumulation. Corresponding performance criteria should
be developed by determining the magnitude of trend that can be expected to be de-
tected with the planned level of sampling effort and the period of time that will be
required to detect statistically significant trends. The use of historical data and the
application of statistical methods to evaluate and establish decision criteria are ad-
dressed in Section 5. An example of the use of statistical methods to determine the
magnitude of trends that can be measured in a particular monitoring design is pre-
sented in Case Study 2 (Appendix A2.0).
2.2
Performance
Criteria
The Quality Assurance Management Staff (QAMS) of EPA has developed an ap- 2.3
proach to designing data collection programs based on the development of Data Additional
Quality Objectives (DQOs). This approach has many of the same elements of the G uidance
-------
21
systems approach for designing monitoring programs that are discussed in this docu-
ment. The DQO process places emphasis on defining the objectives of data collection
programs, specification of the decisions that will be made with the data collected, and
the possible consequences of the decision being incorrect (QAMS, 1986). A key
aspect to the development of DQOs is the evaluation of the desired degree of certainty
in conclusions to be derived from the data. This evaluation is similar to the process
described in this document for the development of decision criteria. In both methods,
the decision criteria are developed to ensure that adequate information is obtained for
making decisions. A case study describing the use of DQOs in Superfund Remedial
Investigations is presented by Neptune et al. (1990).
-------
22
-------
23
3.0 Establish Testable Hypotheses and
Select Statistical Methods
Broad monitoring objectives are usually identified in the characterization process.
Table 3-1 provides examples of program objectives that are common to many NEP
monitoring programs. The range of corresponding questions provides some idea of the
breadth of monitoring issues that are encompassed in the monitoring objectives.
In designing the monitoring program, these broad objectives must be focused to
identify specific monitoring variables and corresponding monitoring activities.
ESTABLISH TESTABLE
HYPOTHESES AND SELECT
STATISTICAL METHODS
3.1 Establish The questions identified in Table 3-1 give rise to a number of alternative scientific
Testable hypotheses. For example, questions regarding the water quality component of a
Hypotheses monitoring program and hypoxia can lead to several hypotheses regarding the response
of dissolved oxygen concentrations in the water column. Example hypotheses include:
Nutrient reduction strategies will result in increased dissolved oxygen
concentrations throughout the estuary
Any increases in dissolved oxygen concentrations will be long term,
improvements will not occur for 20 - 30 years
-------
24
It is clear that the monitoring program must be designed to address a wide range of
alternative hypotheses. The recommended procedure for ensuring that sufficient
information and the right type of information is developed in the monitoring program
is to specify, prior to the collection of any samples, the statistical model that will be
used to analyze the resulting monitoring data, and to specify testable (null) hypoth-
eses.
TABLE 3-1.
EXAMPLE MONITORING PROGRAM OBJECTIVES
AND ASSOCIATED QUESTIONS
Objective
Document response of water quality
variables to management actions
Characterize spatial and temporal
patterns in bioaccumulation
Monitor the status of the ecosystem
Questions
Are nutrient reduction strategies effective?
Is the risk of hypoxia reduced?
Is there a decrease in phytoplankton
biomass?
Is light transmittance affected?
Is fish community structure affected?
What is the risk of consuming seafood
products from within the estuary?
Are there trends in fish and shellfish
contaminant concentrations?
Do toxic hot spots exist and what are
the influences on bioaccumulation?
What is the relationship between
sediment concentrations of contaminants
and observed tissue concentrations?
What is the relative contribution of
different sources of pollutants?
What are the trends in selected indicators?
What are the consequences of the
physiological, morphological and
molecular changes on which indicators
are based on organism survival and
population health?
What is the status of species of
commercial and recreational importance?
-------
25
For example, expanding on the water quality issues above, the process of selecting the
statistical model and testable hypotheses for detecting trends in dissolved oxygen
concentration is outlined below.
A regression model is used to partition field observations (Xjj) between several factors
that potentially influence dissolved oxygen concentrations:
Xy = Po+Plti+Wj + fctilj + ejj
where:
X^ = field observation from time i and location j of dissolved oxygen
concentration at a specified depth
Po = mean of all X;: observations
pjtj = temporal component of the measurement
P21; = spatial component of the measurement
p3tiL = location-time interaction component of the measurement
ฃ. = random errors not accounted for by P0, p^, p2lj, ^3^
One possible null hypothesis to be tested in this case is that there is no temporal trend
in measured dissolved oxygen concentrations [H0: p1 = 0]. This is called the null
hypothesis because it states that there is no temporal trend in dissolved oxygen concen-
trations. If it is concluded that H0 is false, then an alternative hypothesis [Ha: $ฑ is
positive (increasing trend) or negative (decreasing trend)] will be assumed to be true.
The specification of this model also provides the basis for addressing other scientific
hypotheses of interest. For example, is there evidence of location-time interaction
[H0: P3 = 0]?
Green (1979) emphasizes the importance of developing testable hypotheses during the
design phase of environmental studies and points out that hypothesis formulation is a
prerequisite to the application of statistical tests. As pointed out below in Section 5,
the development of testable hypotheses and the selection of statistical methods are also
the first steps in evaluating the expected performance of the monitoring program. For
the dissolved oxygen example, this means determining the minimum trend that can be
detected for a specified level of sampling effort. The process of devising testable
-------
26
hypotheses should also be used to initiate discussions between the management
committee and technical experts on the importance of individual monitoring objectives
and the relevance of the associated questions. Platt (1964) provides an excellent
review of the importance of hypothesis development in the design of scientific investi-
gations in general.
The statistical tests that will be used to analyze the resulting data must be specified for 3.2 Selection of
each hypothesis developed. The applicability of univariate, multivariate, parametric Statistical Method
and nonparametric methods must be carefully evaluated. Selected references that
provide background information on the use of various statistical methods are summa-
rized in the annotated list presented in Table 3-2. These references will provide
program coordinators and scientific staff with a basic understanding of the analytical
options available. However, it is essential to involve the statisticians responsible for
the analysis of the data in both the development of testable hypotheses and the selec-
tion of analytical methods.
The statistical software that will be used to analyze the data should also be identified at
this step in the design process. Berk (1987) describes the attributes of effective
microcomputer statistical software. Meads (1990) summarizes the results of a user's
survey of six advanced statistical packages available for PCs. The American Statisti-
cian, a publication of the American Statistical Association, also regularly reviews new
statistical software and updates to existing packages.
-------
27
TABLE 3-2.
REFERENCES TO BASIC MONITORING DESIGN AND STATISTICAL TEXTS
Reference
Monitoring Design
Sampling Design and Statistical Methods for
Environmental Biologists (Green, 1979)
Statistical Methods for Environmental
Pollution Monitoring (Gilbert, 1987)
Sampling Techniques (Cochran, 1977)
Statistical Principles in Experimental
Design (Winer, 1971)
General Statistics
. Biometry (Sokal and Rohlf, 1981)
Biostatistical Analysis (Zar, 1974)
Applied Statistics, Principals and Examples
(Cox and Snell, 1981)
Multivariate Statistics
Applied Multivariate Statistical Analyses
(Johnson and Wichern, 1982)
Applied Regression Analysis
(Draper and Smith, 1981)
Multivariate Statistical Methods: A Primer
(Manly, 1986)
Description
Introduction to principles and options for sampling and
statistical design. Examples of monitoring design and
application of statistical methods.
General reference that describes a wide range of statistical
methods and their application. Intended for nonstatisticians.
Description of several statistical techniques that are not
commonly seen in general references.
Comprehensive review of sampling methods and theory.
Detailed presentation of topics at an introductory level.
Applicability not limited to environmental sampling and
monitoring.
Description of hypothesis testing concepts.
Basic reference to statistical techniques most frequently used
in the biological sciences. Numerous examples of applications.
Basic reference to statistical techniques most frequently used
in the biological sciences. Emphasis on analysis of variance
techniques.
In depth examples of many of the most common
statistical applications
General introduction to multivariate methods.
Detailed introduction to regression techniques. Includes
section on planning large regression studies.
Brief description of the most common multivariate
techniques, with examples
-------
28
-------
29
4.0 Select Analytical Methods and
Alternative Sampling Designs
The goal of this step in the design process is to develop detailed monitoring program
specifications. In addition to the statistical methods, these specifications include the
field collection and laboratory analysis methods for individual monitoring variables,
and the appropriate quality assurance/quality control procedures. Alternative sampling
layouts including numbers and location of sampling points, sample frequency, and the
level of sample replication should also be developed. This information will then be
used in the next step (see Section 5) to evaluate expected monitoring program perfor-
mance and to select the most efficient sampling layout among the alternatives.
SELECT ANALYTICAL
METHODS AND ALTERNATIVE
SAMPLING DESIGNS
4.1 Selection Of Appendix B of this document provides descriptions of numerous sampling methods
Field and that are routinely utilized in estuarine monitoring programs. These descriptions
LaboratO ry include information on how the data can be used to address the goals of the monitoring
Methods programs and to evaluate the success of the CCMP. The essential elements of quality
assurance and quality control programs are also described. The purpose of selecting
-------
30
field and laboratory methods at this stage of the design process is to ensure: the
feasibility of using the selected methods in conjunction with the proposed level of
sampling effort; that any data used to evaluate expected monitoring performance, a
crucial step in the design process, are directly comparable with data that would be
collected in the proposed monitoring effort; and, that standardized methods are used.
Standardized protocols or performance criteria should be developed to ensure that the
data collected by the different groups participating in the estuary monitoring program
are directly comparable. Becker and Armstrong (1988) describe the process of devel-
oping standardized sampling and analysis protocols for measuring selected environ-
mental variables in Puget Sound. The key feature of the development of these proto-
cols was a series of workshops at which regional scientists and managers worked
together to develop standardized methodologies. The protocols which have been
developed are described in the introduction to Appendix B. These protocols were also
used as a primary source of information for the description of methods that are in-
cluded in Appendix B.
In the development of alternative sampling layouts, consideration should be given to 4.2 Alternative
the trade-offs between the benefits of a comprehensive monitoring effort and available Sampling Layouts
funding. In general, federal support for these programs will be limited both in the
amount of funding available annually and the duration of funding. Consideration
should be given to both limiting the scope of monitoring efforts and making the most
efficient use of ongoing monitoring programs in the estuary.
Ideally, the scope of the program should be adequate to meet all identified monitoring
objectives. Where the funding is not available to meet all the objectives, however, the
individual monitoring components should be prioritized on the basis of the relative
importance of related management issues and the availability of existing information.
Generally, emphasis should be placed on focusing monitoring efforts in order to attain
the level of precision necessary to evaluate the effectiveness of individual management
actions, rather than implementing a comprehensive monitoring program that lacks the
ability to detect the level of changes expected over time.
Given the wide variety of habitats in individual estuaries, the large variability gener-
ally associated with environmental samples, and the limit of funding, alternative
sampling strategies should be investigated. Through design optimization, the sampling
effort can be distributed spatially and temporally in such a way as to maximize the
amount of information obtained within the area sampled. The strategy behind most
-------
31
sample design optimizations is either to minimize the detectable difference or trend for
a fixed cost or to minimize the cost for a specified minimum detectable difference or
trend. The strategy adopted will depend upon the specific situation for each monitor-
ing program. In either case, the goal is to obtain the maximum amount of information
per dollar spent.
The choice of a sampling design depends on several factors including the objectives of
the monitoring program, the type of data that are required to test the null hypothesis,
the underlying assumptions of statistical tests, and the spatial and temporal distribution
of the monitoring parameters. These factors can affect both the validity of the test
results and the efficiency of the monitoring program (cost to obtain a given level of
detection). A brief summary of common sampling designs is presented in Figure 4-1.
Figure 4-1.
Description of various
sampling methods.
Sampling Methods
Simple Random: Samples are independently located
at random
Systematic:
Stratified:
Multistage:
Samples are located at regular
intervals
The study area is divided into
nonoverlapping strata and samples
are obtained from each
Large primary units are selected
which are then subsampled
-------
32
The most basic method of collecting monitoring data is simple random sampling. With
this design, samples are selected randomly and with equal probability. While this
method is easy to implement, there are a variety of sampling designs which can be
more efficient. Such designs will produce estimates with smaller standard errors for
the same sampling effort, or require fewer samples to obtain the same standard error as
would be obtained with simple random sampling.
One such design is systematic sampling. In systematic sampling, sample units are
selected at fixed intervals in space or time, usually with a random start. There are
many variations of systematic sampling, however they all share some common advan-
tages and disadvantages. The even coverage obtained with systematic sampling tends
to ensure that each sample, on average, is more representative of the population as a
whole than a simple random sample. Therefore, such samples tend to have a smaller
standard error associated with them. Problems can arise that lead to bias or increased
variance if there is an underlying pattern or periodicity in the population over space or
time, which is common with environmental data. In addition, it can be difficult to
obtain a valid estimate of the standard error if the data cannot be assumed to be
distributed randomly.
Another design often used is stratified sampling. By dividing the study area into
nonoverlapping homogeneous strata it is possible to optimize the sampling effort in
several ways. First, the samples can be allocated to the different strata in proportion to
the size of the strata and the variability within the strata, and in inverse proportion to
the cost of sampling in those strata. This method will insure that the minimum vari-
ance will be obtained for a given cost. Other criteria such as the ecologic importance
of strata and the parameters being measured can also be taken into account when
allocating sampling effort. Stratified sampling also allows the use of the best sampling
designs within strata to further increase the sampling efficiency. Stratified sampling
works well in a tiered approach because it allows monitoring performance assessment
and design modification to be made on a stratum by stratum basis. Stratified sampling
also yields estimates for each stratum, providing information which better represents
the area being sampled, and is therefore more ecologically meaningful. As an ex-
ample, two strata may each have a significant trend for a given parameter, but the
trends may be in opposite directions. If the data were combined (such as in systematic
sampling), the trends might cancel each other out and result in a conclusion of no
significant trend. Because stratified sampling ensures that some samples will be taken
from each stratum, over the entire study area, it helps insure that the overall estimate
will be more representative on average than one obtained from a simple random
sample. This advantage of stratified sampling will be realized even if optimal alloca-
tion is not used and the strata are defined arbitrarily with respect to the parameters of
-------
33
interest. In general, if the variability within individual strata is less than the overall
variability in all combined strata, the standard errors obtained with stratified sampling
will be less than those obtained from systematic sampling, which will be less than
those obtained from simple random sampling. Applications of stratified sampling to
environmental studies are found in U.S. EPA (1982a), Jensen (1973) and Reckhow and
Chapra (1983).
Multistage sampling is another cost effective method of allocating sampling effort over
large areas. In this design, large primary sampling units are composed of smaller
secondary units. For large-scale studies, third stage units may also be used. Within
each of the selected first stage units, one or more second stage units are selected. In
addition to increased sampling efficiency, this method allows intensive sampling to be
done only at the second stage, while parameters which are inexpensive to measure can
be obtained for the first stage units. This provides some level of monitoring over a
wide area and if problem areas are detected, the distribution of second stage units can
be reallocated. The information collected from the first stage units can also be used to
implement variable probability sampling at the second stage to further increase the
sampling efficiency.
Further information on these sampling designs and methods for calculating required
sample sizes and optimal distribution of samples can be found in Gilbert (1987) and
Cochran(1977).
4.3 Use Of A monitoring strategy which incorporates ongoing monitoring programs or elements
Existing from these programs can significantly reduce the cost of the monitoring effort. Exist-
Monitoring ing compliance and resource monitoring programs may produce data which can
Prog rams completely satisfy, or augment, the spatial and temporal coverage required by an NEP
monitoring program. Additionally, by adopting sampling and analytical methods of
ongoing monitoring programs as standard protocols for NEP monitoring program
components, data from these existing monitoring programs may be used in evaluating
the effectiveness of the CCMP.
An inventory and evaluation of existing federal, state, local, and volunteer monitoring
programs within the estuary drainage basin should be conducted in order to assess their
usefulness and applicability in evaluating the effectiveness of the CCMP. Key tasks of
this assessment process include:
Identify existing and planned programs as well as special projects which
may contribute data useful in evaluating the effectiveness of the CCMP
-------
34
Determine whether NEP monitoring program objectives could be cost-
effectively met by incorporating sampling and analytical methodologies
from these existing monitoring programs
Specific monitoring variables, statistical and analytical methods, and quality assur-
ance/quality control (QA/QC) protocols selected will depend upon the stated objec-
tives and action plans outlined in the CCMP, and specific hypotheses to be tested in
the monitoring program. Therefore, techniques from existing monitoring programs
may not always be adopted. However, whenever possible, it is recommended that
programs work to ensure comparability of their methods with existing programs.
Several specific opportunities for taking advantage of ongoing monitoring efforts are
described below.
NOAA's National Status and Trends Program (NS&T)
NOAA's National Status and Trends (NS&T) Program is designed to assess the
current conditions of environmental quality in the nation's coastal zone and to deter-
mine whether these conditions are improving or deteriorating (NO A A, 1989). In order
to achieve this objective, NOAA currently supports the Benthic Surveillance Project
and the Mussel Watch Program.
The Benthic Surveillance Project regularly measures concentrations of contaminants in
sediments and tissues of bottom-dwelling fish. The occurrence of external and internal
symptoms of disease (e.g., fin erosion and liver tumors) are also documented. The
Benthic Surveillance Project currently collects and analyzes sediment and bottomfish
samples at 75 estuary sites.
The Mussel Watch Program collects mussels and/or oysters once a year from approxi-
mately 200 sites nationwide (NOAA, 1989). Analyses of sediment and bivalve tissue
concentrations of trace metals, DDE, PCBs, aromatic hydrocarbons, and radionuclides
are conducted.
NS&T has developed standardized methods for analysis, quality assurance (QA), and
quality control (QC) for fish and shellfish tissue, and sediment contaminants. The
methods employed in this program and sampling locations should be reviewed to
determine the feasibility of their incorporation into the planned monitoring efforts in
each estuary. For detailed information concerning analytical and QA/QC methods,
contact:
-------
35
NOAA National Status and Trends Program
NOAAN/OMA32
6001 Executive Blvd
Rockville.MD 20852
EPA's Environmental Monitoring and Assessment Program (EMAP)
The goal of EMAP's Near Coastal Program (EMAP-NCP) is to monitor the condition
of near coastal ecosystems, evaluate the relationship between ecological condition and
anthropogenic disturbance, and assess the effectiveness of pollution control actions
and environmental policies on a regional and national scale. EMAP will provide
information on biological and chemical indicators, sampling design and methods,
analytical methods, and QA/QC protocols.
EMAP-NCP sampling design consists of three schemes (U.S. EPA, 1990b). EMAP's
Regionalization scheme divides the nation's estuaries and coastal resources into
biogeographical provinces (e.g., Virginian, Carolinian, Louisianian, Acadian,
Columbian, and Californian Provinces). EMAP's Classification scheme classifies
estuaries into three resource classes that have similar physical features:
Large, continuously distributed estuaries (e.g., Chesapeake Bay, Long
Island Sound)
Large, continuously distributed tidal rivers (e.g., Potomac, Delaware,
Hudson Rivers)
Small, discretely distributed estuaries, bays, inlets, and tidal creeks and
rivers (e.g., Bamegat Bay, NJ, Indian River Bay, DE, Lynnhaven Bay, VA)
EMAP's sampling scheme consists of elements of systematic, random, and fixed
location sampling. Large, continuously distributed estuaries are sampled using a
randomly placed systematic grid, with grid points about 18 km apart. Large tidal
rivers are sampled along systematically spaced lateral transects. Transects are located
about 25 km apart. Only two sampling points are located on each transect, one ran-
domly selected, and one selected using scientific judgement to identify sampling
locations that may be indicative of degraded conditions in the system. Small estuaries
are sampled by partitioning them in groups of four, selecting one estuary randomly
from each group of four, and sampling at two stations in each small estuary selected.
EMAP will operate on a four year sampling cycle, with one fourth of the sites in a
-------
36
region sampled each year. Sampling will be undertaken only during the months of
July and August. Given this sampling design and schedule, it is clear that a very small
amount of EMAP data, if any, is likely to be available to most NEP estuary programs
in a particular year. The stratified random sampling approach developed by the EMAP
program may be of use to some NEP estuary programs if applied on a much smaller
scale. The EMAP program is currently evaluating the advantages of sampling at
alternative spatial scales.
EMAP-NCP indicators of environmental quality are described in EMAP's Near
Coastal Program Plan for 1990 and Ecological indicators documents (U.S. EPA, 1990b
and 1990c). Ecological indicators include:
Response Indicators
- Benthic species composition and biomass
- Gross pathology of fish
- Fish community composition
- Relative abundance of large burrowing shellfish
- Histopathology of fish
- Apparent Redox Potential Discontinuity
Exposure Indicators
- Sediment contaminant concentration
- Sediment toxicity
- Contaminants in fish flesh
- Contaminants in large bivalves
- Water column toxicity
- Continuous and point measurements of dissolved oxygen concen-
tration
Habitat Indicators
- Salinity
- Sediment characteristics
- water depth
Stressor Indicators
- Fresh water discharge
- Climate fluctuations
- Pollutant loadings by major category
- Land use patterns of watershed by major categories
-------
37
- Human population density/demographics
- Fishery landings statistics
Methods for collecting many of these indicators are addressed in the methods section
(Appendix B).
For detailed information concerning EMAP analytical, statistical, and QA/QC meth-
ods, consult the Director of EMAP's Near Coastal Waters Monitoring Program at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
27 Tarzwell Drive
Narragansett, RI02882
Local Monitoring Programs
In addition to ongoing compliance and resource (e.g., fisheries) monitoring that is
conducted in each estuary, there are also other opportunities to augment the estuary
monitoring programs. For example, states are required under 305(b) of the Clean
Water Act (CWA) to conduct water quality assessments, and they are encouraged to
include assessments of trends in their reports (U.S. EPA, 1989b) which are submitted
biennially to EPA. Other monitoring programs are conducted by state agencies,
universities and pollutant dischargers. These programs should be evaluated to deter-
mine their use in monitoring CCMP implementation.
State sponsored and private volunteer monitoring programs have been shown to be
very valuable in collecting data on estuarine water quality, beach litter, marine mam-
mal strandings, and the status of other estuarine resources (Armitage et al., 1989).
Volunteers in the Chesapeake Bay watershed have collected data used to help verify
water quality models, and to identify correlations between measured variables, such as
low dissolved oxygen, and the frequency of observed events such as fish kills and
algae blooms (Ellett, 1988).
Volunteer monitoring programs have also been useful in developing an educated and
involved constituency committed to protecting water resources (U.S. EPA, 1990d).
Partnerships between the public and government agencies responsible for management
must be developed if National Estuary Program CCMPs are to be effectively imple-
mented. Establishment of volunteer monitoring programs has proven to be an effec-
tive way to build public commitment toward achieving environmental quality goals
-------
38
and objectives in the CCMP. Through participation in monitoring programs citizens
learn how they contribute to pollution problems, and develop a sense of stewardship
toward the waters they are monitoring. Volunteer monitoring programs also help the
public to understand the difficulties faced by the scientific community in linking water
quality changes to impacts on living resources.
Each National Estuary Program Management Conference should establish a volunteer
monitoring component as an integral part of its monitoring program. The experience
of citizen monitoring programs throughout the country proves that volunteers can be
trained to carry out a wide variety of environmental monitoring tasks, provided they
are given the appropriate equipment and instruction. In estuaries, volunteers can be
especially helpful in upstream areas not normally covered by a state's monitoring
network. Basic water quality measurements such as pH, transparency, salinity,
dissolved oxygen, and temperature can provide useful information to the comprehen-
sive monitoring program. Trained volunteers can also be used to assess aquatic
vegetation in the estuary, and can provide information on acute problems such as
spills, fish kills, and algae blooms. EPA has published a guidance document describ-
ing how to plan and manage effective volunteer environmental monitoring programs
(U.S. EPA, 1990d). The document provides an overview of the use of citizen volun-
teers in environmental monitoring. It discusses how to plan and organize volunteer
monitoring projects, how to involve the media, and how to prepare quality assurance
plans for volunteer programs.
In general, estuary monitoring programs must select cost-effective methods which
provide data essential to assessing the effectiveness of the CCMP. Whenever possible,
programs should work to ensure comparability of its methods with applicable ongoing
programs. Key aspects of the incorporation of existing state agency and volunteer
monitoring programs into the overall design for the Puget Sound Ambient Monitoring
Program are described in Case Study 1 (Appendix A 1.0).
-------
39
5.0 Evaluate Monitoring Program
Performance
Although often overlooked and neglected, the evaluation of monitoring program
performance is potentially the most important step in the design and review process.
The performance evaluation motivates the development of explicit statements of
program objectives as well as the specification of quantitative performance criteria
during the design phase. During the course of the monitoring effort, performance
evaluations provide a systematic procedure for measuring success in terms of the
ability to meet program goals. The periodic evaluation process also identifies the need
to modify sampling design and methods.
Evaluation procedures are essential because the information developed in the monitor-
ing programs must be sufficiently precise and scientifically defensible. The monitor-
ing programs will provide the primary source of information that will be used to
evaluate the success of the CCMP. This information will be used as a basis for
determining the efficacy of selected management strategies and the accuracy of model
predictions upon which many management decisions have been based. The monitor-
ing programs will also provide quantitative information that will guide decisions
regarding needed modifications to the management plan.
Additionally, the cost of the estuary monitoring programs will be substantial. In order
to protect this investment, it is essential to assess expected performance prior to
collecting the first samples. This performance information will provide the basis for
determining the feasibility of proposed sampling strategies, selecting the most effec-
tive monitoring components and variables, and optimizing the overall monitoring
effort.
The two types of performance evaluations are shown in Figure 1-1 and highlighted in
the schematic on the next page. The first is the evaluation of the expected perfor-
mance of individual components of the monitoring program (e.g., the evaluation of
trends in toxic chemical accumulation in sediments). The first evaluation takes place
during the design phase. The second type of performance evaluation is the assessment
of overall program performance. This assessment takes place after the monitoring
program has been implemented (e.g., after the first full year of data have been col-
lected). The objective is to determine if the overall goals of the program are being met
by the individual monitoring components, and if the program should be modified by
adding, deleting or expanding the scope of individual monitoring components. The
essential feature of both types of evaluation is the existence of a feedback loop that
provides the pathway for modifying the system's design based on monitoring program
performance.
-------
40
EVALUATE EXPECTED
MONITORING STUDY
PERFORMANCE
EVALUATE/ASSESS PROGRAM PERFORMANCE
IS
MONITORING
STUDY
PERFORMANCE
ADEQUATE?
The establishment of performance criteria (e.g., the ability to detect a change in
chlorophyll concentrations of 5 [ig/l over a period of five years) is a fundamental part
of developing monitoring objectives. As indicated in Section 2.0, these performance
criteria represent the level of change that must be detected in order to make manage-
ment decisions regarding the effectiveness of the CCMP. It is these performance
criteria that will be used to evaluate the applicability of individual components of the
monitoring program. The specification of sampling methods for proposed monitoring
program components described in Step 3 (Section 4.0) includes the development of
alternative sampling strategies, including the monitoring variablesAndicators and the
level of sampling effort (numbers of sampling stations and sample replicates). The
goal of the performance assessment is to evaluate the effectiveness of these alternative
sampling designs in terms of the established performance criteria. The results will
5.1 Evaluate the
Expected
Performance of
Individual
Monitoring
Program
Components
-------
41
provide the basis for determining the relative benefits of individual monitoring compo-
nents and selecting the final monitoring design.
The questions that will be addressed in these analyses are:
Can the proposed sampling effort meet the needs of the monitoring
program as defined in the performance criteria associated with the stated
objectives?
How can the proposed program be modified to ensure that these objectives
are met?
As indicated by the feedback loop shown in Figure 1-1, this is an iterative process.
Proposed sampling designs are evaluated and modified, if necessary, to meet the
overall objectives. The tools for conducting these analyses are described in Sec-
tion 5.3.
5.2 Evaluate
Overall Program
Performance
The overall performance of the monitoring program should be evaluated at periodic
intervals. Initially, this evaluation should take place at the conclusion of the first year
of sampling. This evaluation should compare the results with the expected monitoring
performance, and a list of required modifications should be prepared. Opportunities
for streamlining the program should be identified, and the performance criteria should
be reviewed and revised, if necessary, for subsequent evaluations.
5.3 Statistical
Power Analysis
Methods
The primary tool for conducting these analyses is statistical power analysis. Statistical
power analysis provides an evaluation of the ability to detect statistically significant
differences in a measured monitoring variable. The importance of these analyses can
be seen in the examination of the possible outcomes associated with testing the null
hypothesis (e.g., H0: sampling location has no effect on observed sediment contami-
nant concentrations) shown in Figure 5-1:
The null hypothesis is true, and it is rejected. This is referred to as a Type I
error (a) and is commonly called the significance level of the test. By
convention this value is routinely set at 0.05, i.e., the investigator accepts a
small probability of incorrectly concluding that there are differences in
sediment contaminant concentrations between sampling locations.
-------
42
g ACCEPT
C/5
O
Q REJECT
HYPOTHESIS
ACTUALLY TRUE ACTUALLY FALSE
1-cc
ซ Illl ;;;
iS;VXXf:::.yx.X^;-;.^:&XX.;V?'3X-
sซ;y?';;;:ffi>::.;.:sH::,:ซ;:Hvi: ; mifm^t :'
03
1-P
Figure 5-1. Hypothesis
testing: possible circum-
stances and test outcomes.
The hypothesis is true, and it is accepted. This is the complement of the
Type I error (1-a).
The hypothesis is false, and it is accepted. This is referred to as a Type II
error ((3). While the significance level of the test is consistently reported
with the results of statistical tests, the probability of accepting the null
hypothesis when it is not true (Type II error) is almost never reported with
statistical test results. Moreover, the consequences of the Type II error are
probably not fully comprehended by many investigators.
The hypothesis is false, and it is rejected. This is the complement of the
Type II error (1- P) and is referred to as the power of the test. Therefore,
statistical power is the probability of correctly detecting an effect.
Consideration of these possible outcomes of a statistical test leads to three important
conclusions. The first is that failure to reject the null hypothesis does not justify its
acceptance. For example, the failure to reject the null hypothesis (e.g., that location
has no effect on sediment contaminant concentrations) may occur either because there
really is no effect (probability = 1 - a) or because the power of the test is so low
(probability = (3). In the later case, a Type II error occurs because the statistical test is
weak. This may be due to the highly variable nature of the sampling environment or
the low level of sampling effort. In either case, it is possible to evaluate the probabil-
ity of the Type II error for any statistical comparison of environmental data. There-
fore, the second conclusion is that the probability of the Type II error should be
reported with all statistical test results. The third conclusion is most relevant to the
-------
43
design of the NEP monitoring programs - the expected power of the statistical test
should be evaluated prior to implementing the sampling program.
Although statistical power analyses are not routinely conducted with statistical tests of
significance, there is ample guidance available, and the tests can be easily performed.
Power analyses can be conducted using standard equations and tables or nomographs.
The basic tools for conducting these analyses are provided by Cohen (1977), Dixon
and Massey (1969), and Winer (1971). More comprehensive tables and nomographs
are provided by Pearson and Hartley (1951), Tang (1938), Lehmer (1944), and Scheffe
(1959), but the use of these references requires more advanced statistical training.
There are also a wide range of computer programs available for conducting statistical
power analyses. Goldstein (1989) reviewed several MS/PC-DOS power analysis
programs and compared the methods they cover, their ease of use, graphics capabilities
and their computational accuracy. Additionally, a statistical power analysis tool for
analysis of variance is available on the EPA's Ocean Data Evaluation System (U.S.
EPA, 1987a), and the U.S. EPA guidance for conducting fish histopathology studies
(U.S. EPA, 1987b) provides nomographs for the power analyses associated with
contingency table analysis.
Most basic statistical texts address hypothesis testing and the concepts of statistical
power analysis. The statistical texts by Dixon and Massey (1969) and Winer (1971)
provide introductory level descriptions as well as the tables for conducting these
analyses. The statistical text by Scheffe (1959) provides a theoretical description of
power tests for the analysis of variance. The Technical Support Document for the
ODES Statistical Power Analysis Tool (U.S. EPA, 1987a) provides a detailed intro-
ductory level description of power tests for the analysis of variance.
The power of all statistical tests is dependent upon the following design parameters:
significance level of the test (a)
level of sampling effort (i.e., number of sampling stations and sample
replicates)
minimum detectable difference in the effect that can be detected
natural variability within the sampling environment
This relationship between the power of a statistical test and the design parameters
makes several types of power analyses possible. The power of the test can be deter-
mined as a function of any of these design parameters. Alternatively, the value of any
individual design parameter required to obtain a specified power of a statistical test can
be determined as a function of the other parameters.
-------
44
The results of statistical power analyses are generally reported in two formats. In the
first, shown in Figure 5-2, the minimum difference in the effect that can be detected is
shown as a function of level of sampling effort (number of replicate samples). The
results of this type of analysis provide a quantitative comparison of alternative sam-
pling layouts, and they are especially useful in the evaluation of proposed NEP moni-
toring programs. Using these results, the level of sampling effort required to obtain a
selected level of precision in the monitoring program can be determined. This ex-
ample was taken from the evaluation of bioaccumulation monitoring strategies (U.S.
EPA, 1987c). Historical data for liver concentrations of PCBs in winter flounder were
<
LLJ
o
o"-
LLJ
O
cc
LLJ
CO
o
LLJ
(
LJJ
Q
350-1
300-
250-
Lu 200-
150-
100-
50 -
Fixed Design Parameters
Statistical Significance (a) = 0.05
Power (1-p) =0.80
Stations = 4
Estimated Variance (a2) = 2.06
T
2
T
4
T
6
T
8
10
T-
12
"T"
14
16
NUMBER OF REPLICATES
Figure 5-2. Minimum
detectable difference vs.
number of replicates for
fixed set of design
parameters.
-------
45
used to evaluate the expected performance of alternative sampling designs. The results
shown in Figure 5-2 were taken from the first phase of the analyses. They indicate that
the minimum difference in tissue concentrations of PCBs that could be detected
between sampling locations with the collection and analysis of tissue from five fish at
each location was approximately 120 percent of the overall mean at all stations or
4.9 mg/kg. Additionally, these results also show that the sampling program would
require the collection and analysis of tissue from seven fish at each location to detect a
difference equal to the overall mean among sampling locations. Based on the results
of these early analyses during the design phase of the monitoring program, alternative
sampling strategies were evaluated. Additional power analyses were conducted to
evaluate the compositing of fish tissue prior to laboratory analyses. The results
indicated that the collection of replicate composite samples would permit the detection
of substantially smaller differences in tissue concentrations of PCBs at a much lower
cost.
Statistical power analyses may also be displayed as shown in Figure 5-3. The power
of the test (the probability of detection) is shown as a function of the minimum detect-
able difference that can be detected between locations or over time. These results,
taken from Case Study 2 (Appendix A2.0), were obtained from the analysis of a subset
Figure 5-3.
Power vs. minimum
detectable difference.
I
5
V
Q
*
o
>>
ฃป
1
.Q
s
(L,
OC
LU
I
a
o.o
I I I I I
0.08 0.12
0.00 0.04 0.08 0.12 0.16 0.20
Minimum Detectable Difference (SLOPE, mg/l-yr)
-------
46
of water quality data that were collected in Chesapeake Bay. As part of the 1987
Chesapeake Bay Agreement a commitment was made to reduce the total inputs of
nitrogen and phosphorus entering the mainstream of the Chesapeake Bay by at least
40% by the year 2000. The target of 40% nutrient reduction was developed by model-
ing the environmental conditions in the Chesapeake Bay with a two-dimensional,
steady-state model. The modeling results indicated that, after full implementation of
the 40% reduction goal, the lowest average dissolved oxygen would be 1.6 mg/l and
no waters would be anoxic. Therefore, the minimum performance criterion for the
monitoring program should be the ability to detect a difference in dissolved oxygen
equal to l.6m/l.
To test the ability of the existing monitoring program to meet the performance crite-
rion, historical data were used to estimate measurement variability. Statistical power
analyses were then conducted, using estimates of the maximum and minimum vari-
ance. The results shown in Figure 5-3 indicate that the minimum trend in dissolved
oxygen concentration that can be detected with a probability of 0.80 and ten years of
data is on the order of 0.06 to 0.13 mg/^-yr. These results indicate that the existing
monitoring program would meet the specified performance criterion.
The relationship between power and minimum detectable difference, shown in Fig-
ure 5-3, provides the information required to evaluate the probability of a Type II
error, and the probability of detecting specific levels of effects in a proposed sampling
program. Recent examples of the application of this type of power analysis include:
Parkhurst (1985); Toft and Shea (1983); and, Rotenberry and Wiens (1985). The
necessity for this type of analysis in the evaluation of statistical test results is provided
byPeterman(1989).
-------
47
6.0 Implement Monitoring Study and Data
Analysis
Data management and data analysis, key issues that are often overlooked in the design
of monitoring programs, are as important to the success of the monitoring effort as the
collection and laboratory analysis of field data. Moreover, the cost of effective data
management/data analysis can be substantial. On the order of 20 percent of the budget
allocated for the monitoring program should be reserved for data management and data
analysis activities. Failure to plan for these costs can result in the loss of information
due to inadequate data preservation and limited analysis of the monitoring data that are
collected.
IMPLEMENT
MONITORING STUDY
AND DATA ANALYSIS
Recent characterization efforts conducted by individual estuary programs, at consider-
able expense, found that historical data are not readily available and that essential
quality assurance information necessary to evaluate the comparability of data sets is
often not preserved. Generally, it was found that while large expenditures are often
made on data collection, the amount of funding allocated to data management and data
analysis is relatively small and inadequate.
-------
48
The need to assimilate and integrate historical information, as well as planned monitor- 6.1
ing efforts, should drive the development of a data management strategy. Failure to Data Management
plan for data management can result in the loss of the information due to inadequate
data preservation.
The development of a data management strategy must consider the following ques-
tions:
Where will the data go?
How will these data be stored?
Who will maintain the data base?
How will data be checked and loaded into the data base?
How accessible will the data be?
Will statistical, graphical, and report generating tools be available?
How much will it cost?
A computer system will be essential for the management of the data collected by the
estuary monitoring programs. It should be operational prior to implementation of the
monitoring program and should have the following attributes:
Centralized Storage of Raw Data
Easy Access and Use
System Documentation
Quality Assurance Procedures
Linkage to Graphical, Statistical and Report Generation Routines
Long Term Availability and Flexibility
A centralized data source provides the ability to transfer data between investigators
and to conduct analyses that utilize data from different monitoring program compo-
-------
49
nents. An additional feature of a centralized data management approach is the ability
to designate a system administrator who has responsibility for system documentation
and data quality assurance.
The quality assurance information that must be reported with each data set must be
defined prior to implementation of the monitoring effort. The objective is to identify
key field, laboratory and quality assurance information that would allow future users
of the data to make informed decisions regarding the comparability of historical data
sets. This set of basic reporting requirements should be developed for all data types
collected. The Ocean Data Evaluation System (ODES), which was developed by the
Office of Marine and Estuarine Protection, and the data system developed for the
Puget Sound Ambient Monitoring Program provide good models for the implementa-
tion of data quality assurance procedures. Each data set that is submitted to ODES, for
example, undergoes an extensive quality assurance review, and the results of this
review are summarized in a report that is accessible from within the system. Extensive
documentation for ODES is available (U.S. EPA, 1987d and 1988b).
The data management system that is adopted by the estuary programs must also
provide basic graphical, statistical and report generation capabilities and/or the ability
to download data easily to data analysis packages. It must also be flexible enough to
address new data types and analytical needs. Finally, a long term financial commit-
ment must be made to the system in order to establish user confidence in the ability to
store and access data over a long time period.
The NEP has developed a data management policy to ensure that all potential users,
both inside and outside the NEP, have access to environmental data generated under
the program. This policy states:
Management Conferences convened under the NEP are required to submit
all environmental data generated with NEP funds to EPA in ODES format
and machine readable form for storage on the NCC mainframe (National
Computer Center in Research Triangle Park, North Carolina). This policy
applies to data generated with funds awarded in FY90 and beyond.
Responsibility for identifying and selecting data management support
remains with the NEP Management Conferences. The selection of a data
management system by each program should be based on an evaluation of
characterization and monitoring requirements for the estuary.
-------
50
As indicated in Section 5.0, an essential element of the monitoring plan will be the 6.2
specification of a timetable for analyzing the data and assessing monitoring program Data Analysis
performance. The assessment of monitoring program performance should be used to
refine monitoring program objectives and modify individual monitoring program
elements to satisfy these objectives. Initially, monitoring program evaluations should
be conducted after the first year of data collection. Subsequent interim evaluations
should be conducted at two or three year intervals.
The primary purpose of other data analysis activities will be to test the hypotheses
developed in Step 2 of the design process (see Section 3.0). Additional goals are to
summarize the data, generate new hypotheses, and evaluate the uncertainty associated
with the measurements and conclusions. Additional analyses should be designed to
produce information for use by groups with diverse technical backgrounds.
A wide range of statistical and graphical tools are readily available for use in meeting
these goals. Recently, there has been an increased interest in the development of
Geographical Information System applications for graphical analysis and information
display. The National Estuary Program is currently funding projects to demonstrate
the use of this tool to synthesize a broad range of data collected by the estuary pro-
grams, and to effectively communicate this information to interested groups.
-------
51
7.0 Communicate Program Results
One of the primary goals of the monitoring program is to provide information that can
be used to redirect and refocus the CCMP. To achieve this goal, emphasis should be
placed on distributing the data that are collected in the estuary monitoring programs.
The data collected by individual investigators should be made readily available to the
scientific community for comparative studies that relate information from different
components of the program. Section 6.0 discusses the need for the development and
implementation of an effective data management strategy. Emphasis should also be
placed on the analysis of these data. The dissemination of the recorded monitoring
data is not a sufficient mechanism for communicating the monitoring results. Statisti-
COMMUNICATE
MONITORING
illttilii
-------
52
cal analysis of the monitoring data is essential, and graphical and written summaries
should be produced for agency managers charged with implementing the CCMP. The
results must be effectively communicated to an audience with a wide range of techni-
cal backgrounds and interests.
Figure 1-1 and the schematic above show two feedback loops associated with the
evaluation of monitoring data. One provides direct feedback of analytical results that
are used to modify and refine the monitoring program to increase efficiency (see
Section 5.0). The other provides feedback to three basic factors that influence the
design, development and refinement of monitoring program objectives: public con-
cerns, modeling, and research. Data analyses must provide information that addresses
the needs of program managers, scientists and the public.
Graphical and written summaries should be produced that demonstrate the results of
individual components of the monitoring program as well as the relationships between
monitoring activities. These summaries should serve as tools to effectively communi-
cate information on the effectiveness of the actions taken under the management plan,
and to build public awareness of actions taken by the estuary program. Demonstration
materials that summarize program results should be produced for use in newsletters,
workshops, poster sessions, and public forums. The results of the monitoring effort
and the data analyses should also be made available to the scientific community, and
use of the monitoring data should be encouraged. Data analyses should be conducted
to test for trends, test and generate new hypotheses, evaluate the uncertainties associ-
ated with the data, and to identify the source of these uncertainties. These analyses
should serve as a basis for: extending existing knowledge of the estuary, making
refinements to conceptual and numerical models of the system developed in the
characterization phase of the program, and identifying new research. Collectively the
analytical results should provide the necessary information for redirecting and
refocusing the CCMP.
There are several examples of publications by estuary programs to generate public
interest and support, and to disseminate information on monitoring results. In addition
to preparing annual technical reports on monitoring efforts, the Puget Sound Estuary
Program produces a quarterly newsletter - Puget Sound Notes (Figure 7-1). This
technical newsletter has a distribution of 2,750. It is intended to report on recent
program results, and to inform interested individuals about events that affect the
estuary. The Chesapeake Bay Barometer is a one page monthly publication of the
Chesapeake Bay Program. The Barometer provides a summary of dissolved oxygen
concentration and water clarity in the bay during the previous month. It also includes
short, non-technical summaries of topics of general interest (Figure 7-2). Preyious
-------
53
topics have included: salinity in the bay, boat pollution, bald eagle populations, and
the striped bass fishery. The Santa Monica Bay Restoration Project is producing a
series of paired reports on pathogens (Figure 7-3). These reports summarize the
results of the long term assessment of inputs of fecal indicators and enteric viruses
from storm drains. The results of each phase of the study are summarized in both a
Technical Report and a Public Summary.
Figure 7-1. Sample cover
ofPuget Sound Notes.
Sound
Number 25-June, 1991
editor Timothy W. Ransom, PSWQA
Contributors: Timothy W. Ransom, PSWQA
Ronald M. Thorn, Btttelle Marine Sciences Laboratory
LoAnn Hallum, Fisheries Research Institute
Charlee A. Slmenetad, Fisheries Research Institute
Curtis 0. Ttnner, Port of Seattle
Fred Weinmann and Michael RyHto, U.S. EPA-fleglon 10
Roberta Feint, PSWQA
The Estuarine Habitat Assessment Protocol
ovChmrtetA. Slmtnatad, Wetland Ecosystem Team, Fisheries Research Institute, Univarsif/of Washington, CunttD. Tanner, PortotSeartle, andFnd
Wflnmann and Michael Rylko, U.S. Environmental Protection Agency-Region 10, Wetland Program and Officd of Coastal Waters.
Introduction
Presently, our ability to manage and preserve wetland and
other habitats of the waters of the United States is inhibited to
a large degree by the lack of consistent assessment and
monitoring. This is particularly true in the Pacific Northwest,
which is relatively young in terms of the scientific information
base. Net loss of wetlands continues in this region, even within
the regulatory domain of the CWA Section 404 permit process
(Kentula et al., in revision; Rylko and Storm, in prep.). This loss
is largely due to a lack of consistency in the assessment and
monitoringrequ/redto determine which wetlandfunctions will be
lost to a development project and thus the compensation
actions to be taken. Furthermore, this lack of consistency
impairs our ability to determine quantitatively if lost functions
have been fully replaced by the mitigation action. We argue that
one reason behind our inabilrty to compel functional
replacement is the lack of a uniformly acceptable assessment
procedure that developers and their consultants and permitting
and resource agencies can mutually embrace. For instance, of
73 Section 404 permNs issued between 1980 and 1990, the fish
and wildlife habitat function of the site to be filled was evaluated
for only 11 %, and monitoring of the mitigation site was required
f oronly 53% (M. Rylko, pars. comm.). Those proposing wetland
irrigation are hampered by the lack of any guidelines for
assessing the potential wetland impacts of their project, and the
regulating agencies are equally hindered by the inconsistency
and inappropriateness of the proposed assessment and
monitoring plans.
There are a number of wetland habitat assessment
procedures available, but the broad consensus among
technical experts is that they are too general to address the
unique wetland functions in specific geographic regions and are
typically too subjective to provide consistent results. To
increase the effectiveness of our management and
conservation of estuarine habitats, we need assessment and
monitoring procedures that: (1) are based explicitly on habitat
function; (2) are specific to the region of application; (3) use
methods mat are standardized, consistent and comparable; (4)
generate quantitative data rather than qualitative indices; (5)
are designed to be thoroughly objective arrong different users
and sites; (6) will be adaptive in terms of building on prior results;
and (7) enable the user to measure wetland function from both
a single species and multi-species (community) levels.
The Estuarine Habitat Assessment Protocol ("lhป
Protocol"; Slmenstad et al., in press) represents such an
approach to assessing the function of estuarine habitats for fish
and wildlife Fish and wildlife support functions of estuanne
habitats were the chosen focus of fhis Protocol because they
have historically been the driving criteria behind resource
agency requirements tor compensatory mitigation. However.
other potentially more important habitat functions, such as
maintenance of water quamy or flood desynchronlzatlon, also
need to be assessed with similar rigor. The Protocol is intended
to address the need for a systematic procedure that can be
applied uniformly across a variety of wetland and associated
nearshore habitats, using objective, scientific methods.
Although it is directly applicable to the evaluation of
compensatory mitigation projects in estuarine habitats, in the
long term the Protocol may facilitate the development of
successful design criteria for estuarine habitat restoration as
well.
Concept, Development, Content and Limitations of
the Protocol
Concept
There are two fundamental problems with previous methods
of assessing estuanne habitats: (1) documentation of simple
absence or presence of fish and wiidlife, or even more
quantitative information on population sizes, in a specific habitat
or at a specific site has not necessarily indicated utilization per
se; and (2) the causal association between characteristics of the
habitat and the function and extent of fish and wildlife utilization
has neither been identified nor quantified. For example, while
unique nesting of wetland birds is obviously illustrative of a
functional dependence upon the habitat, the occurrence of
highly mobile animals such as migrating juvenile salmon does
not directly infer functional dependence.
The Protocol differs from most prior approaches to estuarine
habitat assessment and monitoring fqrfisn and wildlife because
it focuses on the attributes ot the habitats that promote fish and
wildlfle utilization. The strategy for developing the Protocol was
one of identifying the biological and physical attributes of
estuarine wetlands which determine the extent of fish and
wildlife utilization. In addition, our approach to gathering the
information for the Protocol was to identify those attributes
important to the greatest number of fish and wildlrte species.
Thus, an assessment of these attributes is presumed to
address a broad scope of the habitat's biotic communrty, rather
than just a single target species. It then follows that the
incorporation ot such critical attributes in a wetland mrtigation
and restoration design would increase the function of the habitat
lor the representative community.
Development
The concept of the Protocol emerged from the deliberations
of an ad hoc group called the Urbanized Estuary Mitigation
Working Group (UEMWG), a technical gathering of government
agency, tribal, university and industry representatives which
has met since 1986 to discuss mitigation in urbanized estuaries.
After the basic concept of the Protocol and a process by which
i! could be developed were generated by the UEMWG, funding
was assumed by the U.S. Environmental Protection Agency,
-------
54
CHESAPEAKE BAY BAROMETER
ENVIRONMENTAL CHARACTERISTICS OF THE BAY
MARCH 1990
DISSOLVED OXYGEN
Diuolved oiygen I DO) u the amount of oiyget comsined in
aiti Surface wateป usually hu I greater amount of DO than
ปitci near ihe bottom due 10 iu interaction with the atmosphere
jrxj oiygcn production by plant photosynthesis. DO teveb in
March were somewhat be tow !*a year i leveia m bout the tot-
lom and surface waien of the Bay'} matn jicm. The to* levels
mav be due lo ihe extremely warm air emperaiurei around ihe
Bay bum during March. Thu ปairnih heati the waiet which, in
luro. leueni ihe anxxini of DO thai UK wale/ can hold.
WATER CLARITY
Water cla/iiy 11 a mcajure erf the depth ID which light can pone-
trait wjiei. The greater the value, ihe clearer the wiiet. Sus-
pended material, including Tine sediment and microscope or-
ganism* (plankton), reduce water clarity The Uippled area n M
average cone for March. AJ uiuai. waher claniy u dropptng in
ihe Bay. especially in the upper reaches. Spnnf runoff and pre-
cipiiaion *ปjh sodimcni into the water and wtih ihe warming
temperatures, plankton proliferate. The wannih of March may
tuvซ allowed plankton 10 get an eaily sian ihtj year.
Along the
Susquehanna
Mixed
(including urban)
APRIL HIGHLIGHT: THE SUSQUEHANNA RIVER
Near Coopcnv>*rt, N.Y., a place belter known for the baaetall Hall of
Fame, Ouego Lake tups a small stream of wiler thai with the mita beconea
the sweeping Siuquehanna. Eilending 444 milej 10 ihe head of ihx Chesapeake
Bay at Havre de Grace, the Susquehanm River dnina 13% of New York') land
and alfflon half of Pentsytvnia. Yet K> Maryland and Vinjinii eyes, focuaed
wty of life in the ferule ftoodpUin valleys. The ptcluresque farm) mter^ersed
wiiJi forests dntinguidied Inn valley from the urban Jprซwl uirounding oUier
YeL in nme ways, the charm of ihe valley] hid the problem* [hat began to
burden the river and ultimately the Bay. Particularly in ine lower bum.
Figure 7-2. Sample cover
of Chesapeake Bay
Barometer.
Figure 7-3. Sample
covers of Santa Monica
Bay Restoration Project
Reports.
-------
55
8.0 References
Armitage, T., E. Baptiste and K. Ellet. 1989. Citizen Volunteer Monitoring, a Tool
for Estuarine Management. Coastal Zone '89, Proceedings of the Sixth Symposium on
Coastal and Ocean Management. American Society of Civil Engineers. Vol. l,pp
887-898.
Beanlands, G.E., and P.N. Duinker. 1983. An ecological framework for environmental
impact assessment in Canada. Halifax, Nova Scotia: Dalhousie University Institute for
Resource and Environmental Studies and Federal Environmental Assessment Review
Office.
Becker, D.S. and J.W. Armstrong. 1988. Development of regionally standardized
protocols for marine environmental studies. Marine Pollution Bulletin. 19(7)
310-313.
Berk, K.N. 1987. Effective microcomputer statistical software. The American Statisti-
cian 41 (l):64-76.
Bernstein, B.B., B.E.Thompson and R.W. Smith. 1991. A combined science and
management framework for developing regional monitoring objectives. Presented at
the National Estuary Program Science Symposium, Sarasota, FL. 25-27 February
1991.
Bilyard, G.R. 1987. The value of benthic infauna in marine pollution monitoring
studies. Marine Pollution Bulletin. 18(11)581-585.
Clark, W.C. 1986 The cumulative impacts of human activities on the atmosphere.
Pages 113-123. In: Cumulative Environmental Effects: A Binational Perspective.
The Canadian Environmental Assessment Research Council, Ottawa, Canada, and the
U.S. National Research Council, Washington, DC.
Cochran, W.G. 1977. Sampling Techniques. Third ed. New York: John Wiley & Sons.
Cohen, J. 1977. Statistical Power Analysis for the Behavioral Sciences. New York:
Academic Press.
Cox, D.R., and E.J. Snell. 1981. Applied Statistics, Principals and Examples. New
York: Chapman and Hall.
Dixon, W.J., and FJ. Massey Jr. 1969. Introduction to Statistical Analysis. Third ed.
New York, NY: McGraw-Hill.
-------
56
Draper, N.R., and H. Smith. 1981. Applied Regression Analysis. Second ed. New
York: John Wiley & Sons.
Ellett, K. 1988. An Introduction to Water Quality Monitoring Using Volunteers: A
Handbook for Coordinators. Alliance for the Chesapeake Bay, Inc. Baltimore, MD.
Ferraro, S.P., F.A. Cole, W.A. DeBen, and R.C. Swartz. 1989. Power-cost efficiency
of eight macrobenthic sampling schemes in Puget Sound, Washington, USA. Can. J.
Fish. Aquatic. Sci. 46 (10):2157-2165.
Gilbert, R.O. 1987. Statistical Methods for Environmental Pollution Monitoring. New
York: Van Nostrand Reinhold Company.
Goldstein, R. 1989. Power and sample size via MS/PC-DOS computers.
43(4):253-260.
Green, R.H. 1979. Sampling Design and Statistical Methods for Environmental
Biologists. New York: John Wiley and Sons.
Jensen, A.L. 1973. Statistical analysis of biological data from preoperational-
postoperation industrial water quality monitoring, Water Research 7:1331-1347.
Johnson, R.A., and D.W. Wichern. 1982. Applied Multivariate Statistical Analysis.
Englewood Cliffs, NJ: Prentice-Hall, Inc.
Lehmer, E. 1944. Inverse tables of probabilities of error of the second kind. Ann.
Math. Stat. 15:388-398.
Manly, B.F.J. 1986. Multivariate Statistical Methods, A Primer. New York:
Chapman and Hall.
Mead.L. 1990. The PC Week Poll of Corporate Satisfaction. PC Week. 7(48): 151-
152.
Neptune, D., E.P. Brantly, MJ. Messner, and D.I. Michael. 1990. Quantitative
decision making in Superfund: a Data Quality Objectives case study. May/June: 18-27.
NO A A. 1979. Report of the subcommittee on ocean pollution monitoring. In Reports
of the Subcommittee on: National Needs and ProblemsData Collection, Storage,
-------
57
and Distribution; Monitoring; Research and Development. 87-165. Rockville, MD:
National Oceanic and Atmospheric Association, U.S. Department of Commerce.
NOAA. 1989. National Status and Trends Program for Marine Environmental
Quality. A summary of data on tissue contamination from the first three years (1986-
1989) of the Mussel Watch Project. NOAA Technical Memorandum NOS OMA 49.
NOAA Office of Oceanography and Marine Assessment, Ocean Assessments Divi-
sion, Rockville, MD. 22 pp. and appendices.
NRC. 1990a. Managing troubled waters: the role of marine environmental monitor-
ing. Washington, D.C.: National Academy Press.
NRC. 1990b. Monitoring southern California's coastal waters. Washington, D.C.:
National Academy Press.
Parkhurst, D.F. 1985. Interpreting failure to reject a null hypothesis. Bull. Ecol. Soc.
66:301-302.
Pearson, E.S., and H.O. Harley. 1951. Charts of the power function for analysis of
variance tests, derived from the non-central F-distribution. 38:112-130.
Peterman, R.M. 1989. Application of statistical power analysis to the Oregon Coho
Salmon (Oncorhynchus kisutch) problem. Can. J. Fish. Aquatic Sci. 46:1183-1187.
Platt,J.R. 1964. Strong inference. Sci 146 (3642):347-352.
Plumb, R.H. 1981. Procedure for handling and chemical analysis of sediment and
water samples. Technical Report EPA/CE-81-1. U.S. EPA and Corps of Engineers,
U.S. Army Engineers Waterways Experimental Station, Vicksburg, MS.
PSWQA. 1986. Issue paper: comprehensive monitoring of Puget Sound. Puget
Sound Water Quality Authority, Seattle WA.
PSWQA. 1988. Puget Sound Ambient Monitoring Program. Monitoring Manage-
ment Committee Final Report. Puget Sound Water Quality Authority, Seattle, WA.
QAMS. 1986. Development of Data Quality Objectives:description of stages I and II.
U.S Environmental Protection Agency, Quality Assurance Management Staff, Wash-
ington, D.C.
-------
58
Reckhow, K.H. and S.C. Chapra. 1983. Engineering approaches for lake management
Volume I. Data analysis and Empirical Modeling. Butterworth, Boston.
Rosenberg, D.M., V.H. Resh, S.S. Balling, M.A. Barnby, IN. Collins, D.V. Durbin,
T.S. Flynn, D.D. Hart, G.A. Lamberti, E.P. McElravy, J.R. Wood, T.E. Blank, D.M.
Schultz, D.L. Marrin, and D.G. Price. 1981. Recent trends in environmental impact
assessment. Can. J. Fish, and Aquatic Sci. 38:591-624.
Rotenberry, J.T., and J.A. Wiens. 1985. Statistical power analysis and community-
wide patterns. Am. Nat. 125:164-168.
Scheffe, H. 1959. The Analysis of Variance. New York: John Wiley & Sons, Inc.
Segar, D.A. and E. Stamman. 1986. Fundamentals of marine pollution monitoring
programme design. Marine Pollution Bulletin. 17(5)194-200.
Sokal, R. R., and F. J. Rohlf. 1981. Biometry. Second ed. San Francisco, CA: W.H
Freeman and Company.
Tang, P.C. 1938. The power function of the analysis of variance tests with tables and
illustrations of their use. Sta. Res. Mem. 2:126-149.
Tetra Tech, Inc. 1985. Commencement Bay nearshore/tideflats remedial investigation.
Final Report. 2vols. Prepared for Washington State Department of Ecology and U.S.
Environmental Protection Agency. Tetra Tech, Inc., Bellevue, WA.
Tetra Tech, Inc. 1986a. Puget Sound monitoring program: a proposed plan. Draft
report prepared for U.S. Environmental Protection Agency, Seattle WA.
Tetra Tech, Inc. 1986b. Quality assurance and quality control (QA/QC) for 301(h)
monitoring programs guidance on field and laboratory methods. Prepared for Marine
Operations Division, Office of Marine and Estuarine Protection, U.S. Environmental
Protection Agency, Washington, DC. Tetra Tech, Inc., Bellevue, WA. 267 pp. plus
appendices.
Tetra Tech, Inc. 1986c. Recommended protocols for measuring metals in Puget Sound
sediment and tissue samples. Draft Report. Prepared for U.S. Environmental Protec-
tion Agency, Region 10, Office of Puget Sound. Tetra Tech, Inc., Bellevue, WA.
-------
59
Tetra Tech, Inc. 1986d. Recommended protocols for measuring organic compounds
in Puget Sound sediment and tissue samples. Draft Report. Prepared for U.S. Envi-
ronmental Protection Agency, Region 10, Office of Puget Sound. Tetra Tech, Inc.,
Bellevue, WA. 55pp.
Toft, C.A., and P.J. Shea. 1983. Detecting community-wide patterns: Estimating
power strengthens statistical inference. Am. Nat. 122:618-625.
U.S. EPA. 1982a. Handbook for Sampling and Sample Preservation of Water and
Wastewater, EPA-600/4-82-029. Environmental Monitoring and Support Laboratory.
U.S. EPA, Cinncinati, OH.
U.S. EPA. 1982b. Methods for organic chemical analysis of municipal and industrial
wastewater. EPA 600/4-79-057. U.S. EPA, Environmental Montioring and Support
Laboratory, Cincinnati, OH.
U.S. EPA. 1983. Methods for chemical analysis of water and wastes. EPA 600/4-79-
020. U.S. EPA, Environmental Monitoring Support Laboratroy, Cincinnati, OH.
U.S. EPA. 1985. Tributyl tin support document. Office of Pesticide Programs, U.S.
Environmental Protection Agency, Washington DC. -
ป ^
U.S. EPA. 1987a. Technical support document for ODES statistical power analysis.
U.S. Environmental Protection Agency, Office of Marine and Estuarine Protection,
Washington, D.C. EPA 430/09-87-004.
U.S. EPA. 1987b. Guidance for conducting fish liver histopathology studies during
301 (h) monitoring. U.S. Environmental Protection Agency, Office of Marine and
Estuarine Protection, Washington, D.C. EPA 430/09-87-004.
U.S. EPA. 1987c. Bioaccumulation monitoring guidance: Strategies for sample
replication and compositing. U.S. Environmental Protection Agency, Office of Marine
and Estuarine Protection, Washington, D.C. EPA 430/09-87-003.
U.S. EPA. 1987d. Ocean Data Evaluation System (ODES): user guide. U.S. Envi-
ronmental Protection Agency, Office of Marine and Estuarine Protection, Washington,
D.C. EPA 503/8-90-007.
-------
60
U.S. EPA. 1988a. Financing marine and estuarine programs: a guide to resources.
U.S. Environmental Protection Agency, Office of Water, Washington, D.C.
503/8-88/001.
U.S. EPA. 1988b. Ocean Data Evaluation System (ODES) data submissions manual.
U.S. Environmental Protection Agency, Office of Marine and Estuarine Protection,
Washington, D.C. EPA 501/9-90-006.
U.S. EPA. 1989a. Saving Bays and Estuaries: A primer for establishing and manag-
ing estuary projects. U.S. Environmental Protection Agency, Office of Water.
EPA/503/8-89-001.
U.S. EPA. 1989b. Guidelines for the preparation of the 1990 state water quality
assessment (305(b)) report. U.S. Environmental Protection Agency, Office of Water.
U.S. EPA. 1990a. The National Estuary Program: Final guidance on the contents of a
governor's nomination. U.S. Environmental Protection Agency, Office of Water,
Washington, D.C.
U.S. EPA. 1990b. Environmental Monitoring and Assessment Program: Near Coastal
Program Plan for 1990: Estuaries. EPA 600/4-90/033. Office of Research and Devel-
opment.
U.S. EPA. 1990c. Environmental Monitoring and Assessment Program: Ecological
Indicators. EPA 600/3-90/060. Office of Research and Development.
U.S. EPA. 1990d. Volunteer water monitoring: A guide for state managers. U.S.
Environmental Protection Agency, Office of Water, Washington, D.C.
EPA 440/4-90-010.
Winer, B.J. 1971. Statistical Principles in Experimental Design. New York, NY:
McGraw-Hill.
Zar, J.H. 1974. Biostatistical Analysis. Englewood Cliffs, NJ: Prentice-Hall, Inc.
-------
Appendix A - Case Studies
-------
A-3
A1.0 The Puget Sound Ambient Monitoring
Program Case Study*
A1.1 The Puget Sound Ambient Monitoring Program (PSAMP) is a comprehensive moni-
Purpose and toring program which examines environmental variables throughout Puget Sound and
Approach the surrounding watersheds. With the designation of Puget Sound as an estuary of
national significance in 1988, PSAMP was adopted and submitted as part of the
designation package of the National Estuary Program. The development, implementa-
tion, management, and coordination of PSAMP can serve as an excellent model for
estuarine monitoring programs nationwide.
The sequence of events leading to the implementation of the Puget Sound Ambient
Monitoring Program is shown in Figure Al-1. Important features of this time line are
that the development of this comprehensive monitoring plan was well underway prior
to the selection of Puget Sound for the National Estuary Program, and that the devel-
opment and implementation of such a comprehensive monitoring program takes
substantial time. The PSAMP design was developed between 1986 and 1988, and
sampling began in 1989. However, the Puget Sound Water Quality Authority
(PSWQA), which played a key role in its development, was formed in 1983 and issued
a report in 1984 which recognized the need for estuary-wide monitoring.
Throughout the process of developing PSAMP, National Estuary Program (NEP)
objectives were taken into account. PSAMP fulfills the monitoring requirements of
the NEP, including:
identify trends in water quality, natural resources, and uses of the estuary,
develop the relationship between loads and potential uses,
identify causes of environmental problems, and
monitor the effectiveness of actions taken under the Puget Sound Water
Quality Management Plan (CCMP).
This case study describes the steps involved in the development of PSAMP, the goals
of the program, the monitoring design, and the relationship of PSAMP to the Compre-
hensive Conservation and Management Plan (CCMP) for Puget Sound. Emphasis is
placed on describing the institutional aspects of the monitoring program development
process, including the roles of the participating federal, state, local, and tribal govern-
ments. The importance of scientific review and input from the public is also empha-
sized.
This case study was adapted from a paper prepared by Andrea E. Copping of the Puget Sound Water
Quality Authority and John W. Armstrong of the U.S. Environmental Protection Agency, Region 10, for
the January 1991 Puget Sound Research Conference.
-------
A-4
Figure Al-1. Events
leading to the development
and implementation of the
Puget Sound Ambient
Monitoring Program.
- Puget Sound Water Quality Authority (PSWQA) established
- PSWQA Annual Report recommended creating a Sound-wide Water Quality
monitoring program
- Legislation revised; PSWQA required to prepare and adopt a Puget Sound Water
Quality Management Plan and to include recommendations for a comprehensive
monitoring plan
- PSWQA released Issue Paper on comprehensive monitoring
- EPA Region X released a monitoring strategy for a Sound-wide monitoring Program
- Puget Sound Water Quality Management Plan calls for creating an Ambient
Monitoring Program and creates the Monitoring Management Committee
- Public and government workshops reviewed proposed PSAMP
- Puget Sound designated in the National Estuary Program
- Monitoring Management Committee Final Report submitted to PSWQA
- PSWQA incorporated PSAMP in 1989 Management Plan
- Management framework for PSAMP adopted; PSAMP Steering Committee
established
- PSAMP data collection began
- First Annual Report of PSAMP
- PSAMP review and feedback; continued monitoring
Early in the development of PSAMP, it became clear that the cost and logistics of such A1.2 Development
a comprehensive effort would require the participation of all key resource agencies and Of PSAM P:
affected parties. Participating local, state and federal agencies would have to make institutional
modifications to existing monitoring efforts and agree to share data. It was also Arrangements
-------
A-5
recognized that securing long term funding sources would require strong public
support. The key agencies and the milestones in the development and implementation
of PSAMP are described below.
The Puget Sound Water Quality Authority
In 1983 the Governor of the state of Washington appointed the 21 member Puget
Sound Water Quality Authority (PSWQA) and charged them with identifying pollu-
tion-related threats to Puget Sound marine life, evaluating pollution threats to human
health, and investigating the need for coordination among agencies responsible for
protecting Puget Sound water quality. In their 1984 report, the PSWQA called for "a
long-range coordinated plan ...to protect and improve water quality throughout the
Sound" (Booth and Powell, 1984). They further recognized the need for an estuary-
wide management plan, including increased coordination and enhancement of existing
monitoring programs.
In 1985, the Washington state legislature created the PSWQA as a state agency with a
seven member board appointed by the Governor. The enabling legislation charged the
PSWQA with developing a comprehensive management plan for Puget Sound and its
related waterways. The PSWQA was required to prepare and adopt a Puget Sound
Water Quality Management Plan by January 1987, to update the plan in 1989 and
1991, and to oversee the implementation of the plan. The legislation required that the
plan include consideration of many issues, including "...recommendations for a
comprehensive water quality and sediment monitoring program..." PSWQA was also
charged with creating a biennial "State of the Sound" report to describe water quality
and related resource conditions in Puget Sound. Throughout its history the PSWQA
has relied heavily on input from all parties affected by water quality, regulatory
frameworks, and the need to manage the resources of Puget Sound.
PSWQA Monitoring Issue Paper
As part of the development of the 1987 Puget Sound Water Quality Management Plan,
the PSWQA wrote an issue paper on Comprehensive Monitoring of Puget Sound
(PSWQA, 1986). The issue paper reviewed existing Puget Sound monitoring pro-
grams, described the process for developing a comprehensive water quality and
sediment monitoring program which includes biological resources, and presented
alternatives for developing the monitoring program (Table Al-1). The issue paper
identified four major components of a comprehensive monitoring program:
-------
A-6
TABLE AM. ALTERNATIVES
FOR IMPROVING PUGET SOUND MONITORING
Development and adoption of standardized protocols
A centralized or coordinated data management system
Improved data interpretation/report preparation arid information dissemination
Adequate laboratory capacity
Review of existing water column program and changes to collect more pertinent
information. Follow up actions could be tied to monitoring results.
Development of a sediment monitoring program, including ambient sediment moni-
toring, intensive surveys, discharge monitoring, and dredging/dredged material
disposal monitoring
Improved point and nonpoint source monitoring for municipal treatment plants and
industrial facilities, searches for illegal discharges
Greater number of intensive surveys to identify sources of pollution
Development of a biological monitoring program
Improved monitoring of commercial and recreational shellfish beds
Monitoring by citizen volunteers
Reference: PSWQA 1986.
ambient monitoring
discharge monitoring/compliance monitoring
intensive surveys
resource monitoring and other programs (e.g., monitoring paralytic shellfish
poisoning, changes in shoreline and habitat, weather, demographics, and
socioeconomic information.)
The PSWQA included findings of the 1986 issue paper, and comments received on the
issue paper, in the 1987 Puget Sound Water Quality Management Plan (PSWQA, 1987).
-------
A-7
EPA Region 10 Monitoring Guidance Report
In 1986, the Office of Puget Sound, EPA Region 10, directed their consultant to
develop a monitoring strategy for a Sound-wide monitoring program. Through a
process involving consultation with agencies concerned with Puget Sound manage-
ment, the consultant developed a monitoring design which included goals and objec-
tives, sampling design, operation of the program, and methods for sampling, analysis,
and reporting of data. The monitoring design was released in draft form in 1986 (Tetra
Tech, 1986). At that point EPA Region 10 and the PSWQA recognized the need for a
process which would further involve the many agencies, organizations, and individuals
with an interest in Puget Sound monitoring.
Monitoring Management Committee
The 1987 plan required the PSWQA to appoint a Monitoring Management Committee
(MMC) (PSWQA, 1987). The MMC is chaired by the PSWQA and consists of
representatives of state, federal, local, and tribal governments, business, industry,
shellfish growers, university scientists, Canadian agencies, environmental groups, and
the public. The plan directed the MMC to develop a comprehensive ambient monitor-
ing program, and to address 16 specific issues related to the program (Table Al-2).
Staff support was provided to the MMC to carry out coordination and data manage-
ment tasks.
The MMC was appointed in September 1986. Like all PSWQA meetings, MMC
meetings are open to the public. The committee reaches all decisions by consensus.
The committee began its work in late 1986 by defining the goals and objectives of the
monitoring program and then proceeded to review and modify the consultant monitor-
ing strategy (Tetra Tech, 1986). The MMC established technical subcommittees to
study the monitoring design and to make recommendations on program design.
From late 1986 to early 1988 the MMC refined the monitoring program design which
included sampling and analysis recommendations, a data management system design,
quality assurance guidelines, reporting requirements, cost estimates, and funding
alternatives. The committee released a draft program design in May 1987 for public
and scientific peer review. Workshops were held for the public, local government, and
tribal staff. Comments received from the PSEP technical advisory committee (TAG),
scientists, agency staff, and other affected parties were considered by the committee
and a final report entitled "Puget Sound Ambient Monitoring Program - Detailed
Design Considerations" was adopted by the PSWQA and published in April 1988
-------
A-8
TABLE Al-2. MMC ACTIVITIES REQUIRED BY PSWQA
IN DEVELOPING PSAMP
Define the goals and objectives of the monitoring program.
Define how the monitoring data will be used to assist in management of Puget Sound
resources by responsible agencies.
Review and provide for the specific monitoring needs of programs in the Puget Sound
Water Quality Management Plan.
Propose a monitoring design based to the extent applicable on the November 1986
consultant report to EPA on Puget Sound monitoring and review a similar report on
freshwater (watershed) monitoring.
Analyze the data needed to meet the objectives of the monitoring program.
Analyze and evaluate existing programs and their relation to the objectives of a
monitoring program.
Identify comprehensive monitoring needs not being met by existing programs.
Identify mechanisms to obtain needed information either through redirection of
existing programs or initiating of new programs.
Estimate annual costs associated with existing monitoring in Puget Sound and costs of
additional needed monitoring.
Recommend a data management system that meets agency needs.
Define appropriate mechanisms to fund the monitoring program including allocation
of existing funds and new funding.
Define the appropriate roles for agencies in implementing the monitoring program.
Define how the Puget Sound Atlas (Evans-Hamilton and D.R. Systems 1987) would
be updated, managed, and integrated with ongoing activities.
Recommend a structure for managing the monitoring program, including mechariisms
for making needed refinements and adjustments to ensure program objectives are met
in a cost-effective manner.
Identify opportunities for citizens' monitoring of Puget Sound.
Form ad hoc technical working groups to assist in developing the monitoring program.
Reference: PSWQA, 1987.
-------
A-9
(PSWQA, 1988a). The report included PSAMP goals and objectives, and sampling
and analysis strategies for monitoring sediment quality, marine and fresh water quality,
fish, shellfish, birds, marine mammals, river mouths, nearshore habitat, and additional
data. The PSAMP design was incorporated by reference into the 1989 Puget Sound
Water Quality Management Plan (PSWQA, 1988b).
A1.3 The Puget
Sound Ambient
Monitoring
Program
The MMC developed five goals for PSAMP:
1) Characterize and interpret spatial and temporal patterns of conditions of
Puget Sound in relation to its natural resources and for humans, and
recognize contamination.
2) Take measurements to support specific program elements identified in the
Puget Sound Water Quality Management Plan.
3) Measure the success of programs implemented under the Puget Sound
Water Quality Management Plan (as they relate to the overall ambient
monitoring goal and the program goals of the plan).
4) Provide a permanent record of significant natural and human-caused
changes in key environmental indicators in Puget Sound over time.
5) Support research activities through the availability of consistent, scientifi-
cally valid data.
In order to achieve these goals, the monitoring program characterizes and interprets
spatial and temporal patterns for the following:
factors that endanger human health
biological populations and communities
ซ factors affecting biological populations
presence of pollutants in the Sound
entry of pollutants into the Sound from rivers and streams
estuarine and wetland habitats
results of water quality degradation, such as shellfish bed closures
improvements in water quality such as the reopening of shellfish beds
factors that affect aesthetic conditions.
-------
A-10
The monitoring program identifies areas where:
resources are damaged or contaminated
resources or conditions are changing over time
intermittent or periodic degradations are occurring.
The Scope ofPSAMP
The design of PS AMP calls for monitoring of sediments, water quality in marine and
fresh waters, fish, shellfish, birds, marine mammals, and nearshore habitat. The focus
of the program is Puget Sound-wide, which includes Hood Canal, portions of the
Straits of Juan du Fuca and Georgia to the Canadian border, and the surrounding
watersheds. The sediment quality sampling stations, for example, are shown in
Figure Al-2. Sampling occurs year-round. Station locations, the number of stations
or surveys, and the timing of sampling varies with each type of measurement. Three
types of stations are sampled in many tasks of the program:
fixed stations that are sampled each year,
rotating stations that are sampled on a 3 year cycle; these stations can
provide extended geographic coverage, and
a small number of floating stations that are located at the discretion of the
implementing agency; these stations can provide additional geographic
coverage, or allow investigation of a suspected water or sediment quality
problem.
Table A1-3 summarizes the design for full PS AMP implementation, as well as the
monitoring data that are being collected during 1989-90.
Citizens' monitoring is an integral part of PS AMP. The advantages of using citizens
to collect data are twofold: samples and observations can be collected in a cost-
effective manner, particularly at remote locations where travel by agency staff is
costly; and as a tool for increased public involvement, education, and stewardship.
Using state funds, PS AMP contracts with citizens' groups to provide volunteers to
collect information and samples under the guidance of implementing agency staff
(funds are used for administrative purposes and to reimburse expenses). Parameters
for collection by citizen monitors are carefully chosen to minimize quality assurance
problems; for example, citizens are used extensively to dig shellfish for tissue analysis,
to catch marine fish and salmon for tissue analysis, and to groundtruth remote sensing
data.
-------
A-11
Figure Al-2. Area
included in Puget Sound
Ambient Monitoring
Program.
d/mpn
-------
A-12
Task
Sediment quality
Marine water
column
Fish
Shellfish
Birds
Marine mammals
Nearshore habitat
Freshwater
Reference: PSWQA
TABLE Al-3.
Task
Subcomponents
Sediment chemistry
Bioassays
Benthic invertebrates
Long-term trends
Known water quality
problems
Algal growth
Tissue chemistry
(bottomfish)
Liver histopathology
(bottomfish)
Tissue chemistry
(cod, rockfish, salmon)
PSAMP DESIGN
Proposed
No. of
Stations/Surveys
75 throughout
Puget Sound
10-12 throughout
Puget Sound
5-10 in selected bays
5-10 in selecteci bays
21 stations
21 stations
5- 10 stations
Abundance 35 beaches
Bacterial contamination 35 beaches
Tissue chemistry 35 beaches V
Paralytic shellfish 35 beaches
poisoning
1990a.
Throughout
Puget Sound
Throughout
Puget Sound
One-third of
Puget Sound
75 throughout
watershed
AND 1989-90
Actual
1988-89
Stations/Surveys
50 throughout
PugetSound
24 throughout
Puget Sound
10 stations
10 stations
4 stations
10 beaches
lO beaches
4 beaches
16 beaches
No activity
No activity
No activity
ACTIVITIES
Proposed
Sampling
Frequency
Annually-
spring
Monthly
Annually-
early summer
Annually-
early summer
Annually,
depending
on species
Annually
Quarterly
Annually
Monthly
Monthly, annually
Monthly, annually
Annually
75 throughout Monthly
watersheds
(limited parameters)
Actual
Sampling
Frequency
March- April
(1989,1990)
Monthly
Seasonally
Summer/winter
solstices
May 1989
May 1989
September/
February/
April 1989-1990
May 1990
Quarterly
April 1990
Monthly
Monthly
-------
A-13
The NEP and PSAMP
With the formal designation of Puget Sound as an estuary of National Significance in
March 1988, the 1991 Puget Sound Water Quality Management Plan was adopted as
the Comprehensive Conservation and Management Plan, and PS AMP formally
became a part of Puget Sound Estuary Program.
PSAMP examines data from selected other monitoring programs and studies underway
in the Puget Sound basin in order to better interpret PSAMP data, and to meet the
needs of the NEP which states that a CCMP "shall survey and utilize existing reports,
data, and studies relating to the estuary that have been developed by or made available
to federal, interstate, state, and local agencies." Monitoring programs and studies that
evaluate environmental conditions and potential sources in the Puget Sound region that
are of interest to PSAMP include:
nonpoint source control watershed monitoring
Puget Sound Dredge Disposal Analysis (PSDDA)
PSEP Urban Bay Studies (UBATs)
NPDES compliance monitoring
Department of Ecology intensive surveys (tracing sources of contamina-
tion)
NOAA National Status and Trends Program
research studies, historical information, and other studies.
Timber/Fish/Wildlife
climate/weather data
demographic and socioeconomic conditions
decision record-keeping.
A1.4 Assuring PSAMP Implementation
Implementation of
PSAMP and Cost In order to ensure that PSAMP would be implemented as planned, the PSWQA
negotiated memoranda of agreement (MOA) with each of the implementing agencies.
These agreements detailed the specific responsibilities of each implementing agency
and the PSWQA in relation to PSAMP. The MOAs specify that each implementing
agency must carry out its responsibilities under the PSAMP design, must maintain
PSAMP funding levels within the agency unless the Governor or the state legislature
requires across-the-board agency funding cuts, and must refer changes in PSAMP to
the PSAMP Steering Committee for approval. Through the MOAs, the PSWQA
-------
A-14
agrees to provide staff support to the PSAMP committees and to work with the imple-
menting agencies to maintain long-term funding for PSAMP.
Under the 1985 legislation creating the PSWQA, the agency was due to terminate in
1991. In order to ensure the future of PSAMP, during fall 1989 the PSAMP Steering
Committee developed a recommendation for future management structure for PSAMP
which was not markedly different from the existing one. The PSAMP Steering
Committee preferred that the coordination and dispute resolution functions remain
with the PSWQA. The committee's recommendation was incorporated into the
PSWQA enabling legislation in 1990, which reauthorized the PSWQA. The new
section in the legislation which deals with PSAMP requires the implementing agencies
to participate in the program, requires the PSWQA to ensure implementation of
PSAMP, and allows the PSWQA to establish an interagency committee to coordinate
PSAMP.
Initial PSAMP Assignments
Agencies represented on the MMC were asked to express their interest in playing a
lead role in specific tasks of PSAMP. "Implementing agencies" were chosen for each
PSAMP task based on these expressions of interest. It came as no surprise that the
agencies that expressed interest in becoming implementing agencies were those with
regulatory and/or management responsibilities as well as expertise in the specific
areas; initially all implementing agencies are state agencies (Table Al-4), although this
could change in the future. Each implementing agency was responsible for preparing a
detailed implementation plan, managing their specific PSAMP task, quality assurance
and quality control, data management and reporting, and transfer of PSAMP data to
the central database, housed at the PSWQA. The MMC felt strongly that the PSWQA
should continue to play the lead role in coordinating PSAMP and providing data
management functions, as the Authority has both the expertise and staff available to
carry out these functions.
Cost
A technical subcommittee to the MMC (Costing Subcommittee), PSWQA staff and a
consultant retained by PSWQA worked to develop a cost estimate for PSAMP. In
1987 dollars, full implementation of PSAMP was calculated to cost approximately
$3.2 million a year, with startup costs of $250,000 to $350,000 over the first two years.
Early in 1988 it became apparent that state funds would not be available for full
implementation of the program. Through a priority-setting process involving the
-------
A-15
TABLE Al-4. PSAMP TASK ASSIGNMENTS BY AGENCY
Monitoring Task
Implementing Agency
Sediment Quality
Marine Water Column
Fish
Shellfish
Birds
Marine Mammals
Nearshore Habitat
Freshwater
Reference: PSWQA 1990b
Washington Department of Ecology
Washington Department of Ecology
Washington Department of Fisheries
Washington Department of Health
Washington Department of Wildlife
Washington Department of Wildlife
Washington Department of Natural Resources
Washington Department of Ecology
MMC and other scientists, a plan for phased implementation of the program was
devised. This process established the order in which PSAMP tasks would be funded
when implementation funding became available. The MMC established sediment
quality, fish tissue, and shellfish tissue monitoring as top-ranked priorities for funding.
The MMC also recognized that agency staff could not manage to fully implement
PSAMP in a single year, even if funds were to become available.
The development phase of PSAMP, including establishment of the data management
system, has been funded largely by PSEP funds administered by EPA Region 10 and
made available to the PSWQA by cooperative agreement. State, federal, local, and
tribal agencies, as well as private sector funds have contributed individuals' time
serving on monitoring committees and subcommittees. The development costs of
PSAMP were approximately $200,000 from 1987-1989. Additional development
costs for data management are still being incurred.
-------
A-16
Funding
Long-term stable funds for monitoring programs are notoriously difficult to secure.
Funding for four types of activities were needed to assure the success of PS AMP:
contractor funds for the development of the draft monitoring strategy and several
workshops; staff time for program development; startup costs; and ongoing implemen-
tation. As previously noted, EPA Region 10 provided PSEP funds for PSWQA staff to
coordinate the development of PSAMP and the data management system. PSEP
resources continue to fund PSWQA staff for initial implementation and further data-
base development through FY91. Start up costs for implementing PSAMP have been
itemized together with long-term implementation costs to clarify what is needed to
properly initiate such a program.
To date, approximately 30% of PSAMP has been funded, including central coordina-
tion and data management tasks. Other than PSEP funds available through EPA, all
PSAMP funding has come from the state of Washington. Like all Puget Sound Water
Quality Management Plan funds, PSAMP funds are given directly to the implementing
agency; the PSWQA submits a funding request for the entire plan, with assistance
from the other agencies, and acts as an advocate for the plan to the Governor and the
state legislature.
During 1989-90, PSAMP received approximately $900,000 of state general fund
money each year (PSWQA 1990a). The Washington state Department of Ecology
receives $500,000 a year for sediment sampling and analysis. $200,000 goes to the
Washington state Department of Fisheries for measuring toxics in bottomfish and
recreational fish tissue and for bottomfish health assessments. The Washington state
Department of Health uses $ 160,000 for sampling and analysis of shellfish tissue for
bacterial and/or chemical contamination. The PSWQA receives $40,000 a year for
coordination and data management. $50,000 in citizens' monitoring activities are
funded by a special fund for public involvement and education (PIE Fund) from the
state Centennial Clean Water Fund (tax on tobacco products). In addition to securing
new funds, the Washington state Department of Ecology has modified existing marine
and fresh water monitoring programs to meet PSAMP goals, saving PSAMP $245,000
a year.
Maintaining current PSAMP funding and securing additional funds is a major focus of
the PSAMP Steering Committee and the PSWQA. The MOAs are intended to protect
PSAMP funds within the implementing agencies, all of which have numerous conflict-
ing priorities for funds. The monitoring committees and the PSWQA are relying
-------
A-17
heavily on legislative advocacy, attention from the media, and garnering of public
support through the dissemination of PS AMP results to aid in securing PSAMP funds.
Management of PSAMP
The management structure for PSAMP consists of the PSAMP Steering Committee,
the MMC and subcommittees, and the PSWQA board. PSWQA staff provide techni-
cal and administrative support to the committees, and chair the MMC and PSAMP
Steering Committee.
As PSAMP moved from the development to the implementation stage, it became
apparent that the MMC was too large and unwieldy a group to coordinate the program
and make day to day decisions. In addition, some state agencies were uncomfortable
with having representatives from the private sector make decisions involving the use
of public funds. To solve these problems, a second monitoring committee was formed,
the PSAMP Steering Committee.
The PSAMP Steering Committee is chaired by the PSWQA and is made up of one
representative from the five implementing agencies, the PSWQA, EPA Region 10,
local government, and tribes. Additional members can be added in future if necessary.
The committee is responsible for the overall coordination and management of PSAMP.
The PSAMP Steering Committee meets once a month and reaches decisions by
consensus.
Should the PSAMP Steering Committee be unable to reach consensus on technical or
other program matters, or should any member be dissatisfied with decisions made by
the committee, the matter can be referred to the PSWQA board for resolution. To
date, there have been no such instances.
The MMC maintains the same representation that it had during the development phase,
and, with its subcommittees, acts as the advisor to the PSAMP Steering Committee.
The MMC meets twice a year or as needed. Major changes in PSAMP design,
changes in assignment of an implementing agency, setting funding priorities, and other
technical issues must be referred to the MMC by the PSAMP Steering Committee.
Although the PSWQA has played the central role in facilitating the development and
initial implementation of PSAMP, the PSWQA is committed to maintaining PSAMP
as an interagency effort which benefits the public, as well as many agencies and
organizations besides the PSEP co-managers and implementing agencies.
-------
A-18
Review ofPSAMP
The PSAMP Steering Committee and the MMC recognize that the long-term success
of PS AMP rests on making the program effective in answering needs of the resource
management agencies, and informing the public of the state of the Sound, within the
available budget. An annual review of PSAMP is built into the process by the partici-
pants in order to keep the program focussed on top priority monitoring needs and to
ensure the efficient use of funds. The first annual review will begin in late 1990. In
addition, PSAMP will undergo periodic independent review; the first review is set for
1994, with subsequent reviews every 3 years.
Data Management, Quality Assurance and Reporting
Data management and quality assurance/quality control (QA/QC) requirements for
PSAMP were designed as integral parts of the program. The data management system
consists of a central database housed at the PSWQA, and agency databases at each of
the implementing agencies. The databases are microcomputer (desktop computer)
based, and exchange data electronically using a set of standardized data transfer
formats developed by PSWQA staff in cooperation with technical subcommittees to
the MMC. The central database and several of the agency databases are on dBase IV
and have been customized by PSWQA staff.
PSAMP is also responsible for developing a geographic information system (GIS) on
Puget Sound. Resource maps from the Puget Sound Environmental Atlas (Evans-
Hamilton and D.R. Systems, 1987) have recently been converted from graphics files to
Arc-Info (GIS program) by the USGS, under an agreement with the PSWQA.
PSWQA staff are presently working with the Washington State Department of Natural
Resources to update the Atlas on GIS and will complete the initial Puget Sound GIS by
1993.
Each implementing agency is responsible for QA/QC of sample collection, analysis,
and data management for their tasks of PSAMP. Limited QA/QC of data are per-
formed at the PSWQA before data are accepted into the central database. QA/QC
guidelines were established by the MMC and are detailed in each agency's implemen-
tation plan.
Each implementing agency is required to prepare an annual technical report detailing
their PSAMP monitoring over the previous year. Using the agency technical reports
and other program information, each year PSWQA staff prepare an integrated PSAMP
-------
A-19
report for public release. The first of these reports was released in May 1990
(PSWQA, 1990a). PSAMP reports and quality assured data are available through the
PSWQA.
A1.5 This case study describes the process by which agencies, the private sector, scientists,
Summary and and the public interest groups interacted to design and implement PSAMP. It also
Recommendations identifies some of the ingredients that have made it a successful process. Addition-
ally, there are several opportunities for other estuary programs to improve upon this
process.
Keys to Success
The success of PSAMP design process is attributable to the agencies and individuals
who saw the need for a comprehensive monitoring program and who made the com-
mitment to make the process work. The PSWQA played an essential role in coordi-
nating and managing the process. One key to the PSWQA's past and ongoing suc-
cesses has been the agency's comprehensive approach to complex issues by simulta-
neously addressing the scientific, resource management, political, and public con-
cerns. PSWQA staff have spent considerable effort trying to balance these often
conflicting pressures. PSWQA has been able to maintain an objective role in manag-
ing and coordinating PSAMP without competing with the individual agencies;
PSWQA does not carry out any sampling operations.
The EPA's Office of Puget Sound was very supportive of the efforts to develop
PSAMP. Not only did EPA provide funds for the development of the program, but
key staff were very much involved in the process and fully supportive of the strong
role taken by the PSWQA in coordinating the program. EPA Region 10 has sup-
ported the development of a number of other products and studies which have laid the
groundwork for PSAMP, including the Puget Sound Protocols and Guidelines; the
Puget Sound Environmental Atlas; reconnaissance studies for sediment and tissue
contamination; seafood risk assessment study; and the Puget Sound chemicals of
concern matrix.
The interest among agencies, legislators, and the public in Puget Sound water quality
during the early 1980s helped set the stage for creating the PSWQA and emphasizing
water quality issues and programs, such as PSAMP. Prior to the establishment of the
NEP, EPA Region 10 and the Washington state Department of Ecology established
the Puget Sound Action Program. These agencies were then joined by the PSWQA
and formed the PSEP.
-------
A-20
The design and initial implementation of PS AMP, including the data management
system, is successful both as a technical product and for the consensus process from
which it came. The involvement of representatives from both the public and private
sectors, a targeted outreach effort, and the process of peer review have all contributed
to the success with which the program has been received by the implementing agen-
cies. The data management system and data transfer formats has been very well
received by state and local agencies. PSWQA staff have had numerous requests for
information, and interest in PSAMP have been communicated from several of the
established and emerging estuary programs around the country as well as from other
state programs, other regions, and from Canada. Staff are frequently asked to present
the program design, institutional structure, data management criteria, and uses of
PSAMP data at seminars and conferences in the Puget Sound region and across the
country. In addition, PSAMP was able to release its first annual report one year after
initial sample collection.
The GIS recommendation was initially prepared to serve many purposes. Funding
constraints now and in the future have necessitated that the effort be reduced to fit the
immediate needs of updating of the Puget Sound Atlas and providing some geographic
data management for PSAMP. The flexibility of the recommendation and the broad-
based consensus through which it was built have allowed for this redirection of effort
without a loss of time or momentum.
The redirection of existing state agency funds and the procurement of new state money
for PSAMP were also aided by the long consensus process that was followed. Al-
though sufficient state funds for full implementation of PSAMP are not currently
available, almost $1 million a year is currently committed to PSAMP, and additional
resources can be expected to follow in the future.
Opportunities for increased productivity and cost containment have been seized on
several occasions during this project. For example, a combination of PSEP and state
funds were used to develop the fish task implementation plan and to sample bottomfish
during 1989. Similarly, a mix of funds were used to design and develop the PSP
database. These opportunities have allowed those agencies to proceed with PSAMP
implementation without delay when state funds became available in July 1989. Simi-
larly, the planning carried out by Fisheries and the responsiveness of Fisheries and
Authority staff allowed the cost-effective "piggybacking" of bottomfish sampling with
other state sampling programs in May 1989.
The transition to a new management structure for PSAMP during the implementation
phase, including the establishment of the PSAMP Steering Committee and the reallo-
-------
A-21
cation of management responsibilities between the PSAMP Steering Committee and
the MMC was accomplished smoothly. The agencies involved displayed a sense of
stewardship for the program and have continued to actively participate. The success
of the management structure is most clearly illustrated in the incorporation of the
PSAMP Steering Committee's recommendation for managing PSAMP into law during
the 1990 state legislative session.
All of the implementing agencies are large complex agencies with, in some cases,
several divisions involved in the implementation. Despite the complexity of the issues
involved in long-term implementation and coordination of PSAMP, three of the five
implementing agencies have signed memoranda of agreement with the Authority.
The establishment of a successful citizens' monitoring program benefits both PSAMP
through cost-effective collection of data, and educational and public involvement
needs. Encouraging an educated and involved public is key to providing support and
commitment to the long-term funding of PSAMP.
Opportunities to Improve Future Projects
Throughout the many successes of PSAMP design and initial implementation, there
were setbacks, minor difficulties, and areas for future improvement. The majority of
these problems could be alleviated in other programs with a combination of lessons
learned from this program and the addition of more staff time.
The time to negotiate the memoranda of agreement between each implementing
agency and the Authority was seriously underestimated. The size and complexity of
several of the agencies involved in the program was not taken into account.
Questions have arisen during the design and initial implementation phases of PSAMP
as to whether it would have been preferable to design a more limited and affordable
ambient monitoring program, rather than the necessary and sufficient program de-
signed to meet the goals set out in the MMC final report (PSWQA, 1988a). A smaller
program would have been designed somewhat differently, and would not have raised
the expectations of all those eagerly awaiting PSAMP data. The design of PSAMP as
a thoroughly adequate program for characterizing Puget Sound and measuring the
success of source control and remedial action programs will act as the blueprint with
which to direct funds as they become available. The danger in designing a less com-
prehensive program is a level of complacency among funding sources and managers
who feel that we are collecting adequate information on Puget Sound. With benefit of
hindsight, we would probably design much the same program again.
-------
A-22
Similarly, the amount of effort that went into the design and implementation of the
central database has been questioned as little use has been made of the analytical
capabilities of the system to date. The modular design of the central database has
allowed portions of it to be customized for use by the individual implementing agen-
cies. Without the PSWQA staff effort in developing the central database, it is unlikely
that the implementing agencies would have functional databases in time to manage
their early years of PS AMP data. Additionally, as more data are collected under
PSAMP, the central database will become the site of integrated analyses among the
PSAMP tasks.
Trade offs between efficiency and the need to build broad-based consensus on all parts
of PSAMP were seen throughout the design and initial implementation process. For
example, additional portions of the central database could have been completed by the
end of the development phase if PSWQA staff had begun programming and file design
work earlier. However, an earlier start on database development would not have
allowed as complete review and participation by the implementing agencies, other
participants in PSAMP, and the interested public.
It has proved to be very important to staff the design and implementation of PSAMP
with a full-time manager/coordinator and a systems analyst. The task of keeping all
the implementing agencies and other participants informed of progress, resolving
disputes and misunderstandings, carrying out technical and administrative duties,
coordinating with other closely related monitoring and research programs, and publi-
cizing the existence and results of the program was overwhelming for PSWQA moni-
toring staff. In order to really cover all these bases, additional junior level staff was
needed. In future projects, creative use should be made of interns and entry level
professional staff, as well as additional funding sought for central coordination and
data management.
At times technical questions brought to the MMC and PSAMP Steering Committee
could not be adequately resolved due to the lack of technical expertise on the commit-
tees. Technical subcommittees were formed and proved useful for many questions;
however, inadequate PSWQA staff time did not allow for sufficient meetings and
contacts to resolve every need. In addition, there was frequently insufficient time and
attention given to the program by committee members. In some instances additional
PSWQA staff time could have improved the situation but in general there needs to be a
strong devotion to the program by some of the participants, especially the implement-
ing agencies. It is also important to recognize that there is only a finite pool of Puget
Sound scientists and managers available to participate on a regular or frequent basis in
such programs; a similar-sized pool of experts may prove limiting in many areas.
-------
A-23
The major activity which was given short shrift during the initial implementation of
PSAMP, and which future projects should strive to improve upon was the publicizing
of monitoring data, results, expectations, and interpretations. Even with a very limited
amount of new monitoring data there are ample historical Puget Sound data residing in
the state and federal agency files which could be resurrected, compared to recently
collected information, and displayed to managers and the public. Publicly releasing
results and interpretations of scientific findings helps to raise the level of awareness
and support for a program like PSAMP.
A1.6 Booth, P.N., and S.L. Powell. 1984. Puget Sound Water Quality Annual Report.
References Prepared for PSWQA, Olympia, WA. 39 pp.
Evans-Hamilton and D.R. Systems. 1987. Puget Sound Environmental Atlas. Pre-
pared for U.S. Environmental Protection Agency, Puget Sound Water Quality Author-
ity, and U.S. Army Corps of Engineers, Seattle, WA.
PSWQA. 1986. Issue Paper: Comprehensive Monitoring of Puget Sound. PSWQA,
Seattle, WA. 34 pp. + appendices.
PSWQA. 1987. 1987 Puget Sound Water Quality Management Plan. PSWQA,
Seattle, WA.
PSWQA. 1988a. Puget Sound Ambient Monitoring Program - Detailed Design
Considerations. Monitoring Management Committee Final Report. PSWQA, Seattle,
WA. 145 pp.
PSWQA. 1988b. 1989 Puget Sound Water Quality Management Plan. PSWQA,
Seattle, WA. 276 pp.
PSWQA. 1990a. Puget Sound Update: First Annual Report of the Puget Sound
Ambient Monitoring Program. PSWQA, Seattle, WA. 89 pp.
PSWQA. 1990b. Draft 1991 Puget Sound Water Quality Management Plan. PSWQA,
Seattle, WA. 320pp.
Tetra Tech, Inc. 1986. Puget Sound Monitoring Program: A Proposed Plan. Draft
Report Prepared for U.S. Environmental Protection Agency, Seattle, WA. 103 pp.
-------
A-24
-------
A-25
A2.0 Chesapeake Bay Monitoring Program:
Detection of Trends in Estuaries
A2.1 Purpose This case study demonstrates the use of statistical methods to evaluate the ability of
and Approach ongoing monitoring efforts to detect long term trends in the estuary. As described in
Section 5.0, this type of evaluation can be used to determine if the overall goals of the
monitoring program are being met and if the existing program should be modified.
Water quality data collected as part of the Chesapeake Bay Monitoring Program were
obtained to demonstrate the analytical approach. However, this case study is not
intended as a review of the ongoing monitoring efforts in the Chesapeake Bay. The
data that were used represent a small subset of the data that have been collected as part
of a comprehensive sampling effort. The ongoing monitoring program also has many
objectives that are not addressed in this example analysis.
Background information on the Chesapeake Bay Program is included, and the devel-
opment and implementation of the nutrient reduction strategy that is a cornerstone of
the bay-wide management plan is described. A description of the water quality data
that were obtained from the Chesapeake Bay Program is also included. The focus of
the case study, however, is on describing the use of statistical analyses to evaluate
monitoring effectiveness and the application of the analytical results. An understand-
ing of statistical power analysis is necessary to fully understand the procedures de-
scribed. Familiarity with the information provided in section 5.0 is therefore recom-
mended.
A2.2 Chesapeake The Chesapeake Bay Program, first authorized by Congress in 1977, has developed a
Bay Program "long-term process of collecting critical environmental data throughout a 66,500
square mile ecosystem" (Chesapeake Bay Program Monitoring Subcommittee, 1989).
Due to the size of the bay, a cooperative effort between numerous state and federal
agencies was required to build the current program. Efforts during 1976 through 1983
were spent developing a program and communication structure.
Early characterization efforts indicated that:
The bay was over-enriched with nutrients
Toxic "hot spots" exist
Submerged aquatic vegetation habitats had been lost
Lowered dissolved oxygen levels were causing serious water quality
problems and the loss of valuable habitats
-------
A-26
In particular, increases in the extent of oxygen deficient waters and nuisance algae and
declines in water clarity, submerged aquatic vegetation, finfish, and oysters (Flemer
et al., 1983) were observed. Eutrophication was suspected to be the cause of many of
the problems. Though toxics were a major concern, there was much more uncertainty
about the cause-effect relationships (Magnien and Haire, 1989).
In 1983, the first Chesapeake Bay Agreement was signed by Maryland, Virginia,
Pennsylvania, the District of Columbia, and EPA. The agreement signified a commit-
ment to establish a structure to oversee cooperative and comprehensive measures
needed to restore the bay.
In 1984, the first basin-wide monitoring network was established along with a wide
range of new initiatives and legislation passed in all three States and the District of
Columbia. Previous monitoring information was insufficient to define the changes in
water quality between the 1950's and 1970's. Therefore a long term monitoring
program was designed to improve the understanding of the bay dynamics associated
with pollution reduction, to characterize the current status of water quality problems,
and to determine the long-term effects of management actions (e.g., phosphate deter-
gent ban, improved sewage treatment, nutrient reduction strategy, and best manage-
ment practices).
1987 Chesapeake Bay Agreement
The Chesapeake Bay Program has evolved over a number of years. The 1987 Chesa-
peake Bay Agreement was especially important because it set forth a water quality
goal to "reduce and control point and nonpoint sources of pollution to attain the water
quality condition necessary to support the living resources of the Bay." As a result of
this commitment by the Governors of Virginia, Maryland, and Pennsylvania, the
mayor of the District of Columbia, the Chesapeake Bay Commission, and the EPA
Administrator, bay-wide strategies for the control and reduction of nutrients, toxics,
and conventional pollutants were to be developed.
Nutrient Reduction Strategy
The 1987 Chesapeake Bay Agreement established a commitment to develop, adopt,
and begin implementation of a strategy to equitably reduce the total inputs of nitrogen
and phosphorus entering the mainstream of the Chesapeake Bay by at least 40% by the
year 2000. The target of 40% nutrient reduction was developed by modeling the
environmental conditions in the Chesapeake Bay with a two-dimensional, steady-state
model. According to modeling efforts, there are 1.35 billion cubic meters of anoxic
-------
A-27
water in the bay sometime during the summer and an average of less than 0.5 mg/t of
lowest summer dissolved oxygen in the deepest portions of the bay (based on 1985
nutrient loadings). Model results also suggest that a significant improvement in
dissolved oxygen in bay bottom waters can be achieved if nitrogen and phosphorus
loads are reduced. For example, the model predicts that after full implementation of
the 40% reduction goal that the lowest average dissolved oxygen will be 1.67 mg/t and
no waters will be anoxic.
Other benefits such as decreased algal biomass are expected as well. Chlorophyll
concentrations are expected to decrease (based on model predictions) from 13.6 ug/ฃ
to 8.8 \ig/l. The reestablishment of rooted aquatic plants and improved fish popula-
tions are expected as well but cannot be quantified (Chesapeake Bay Program, 1988).
Model predictions are based on the information available in 1985 and the available
model did not include a mechanism to account for nutrient loading from sediment.
Since each state had a different mix of point and nonpoint source inputs as well as
different programs and policies, each state has a unique plan for nutrient reductions
(Chesapeake Bay Program, 1988). This approach allows each state to implement
individual strategies that are best suited for that state within the limits of its program
constraints, while still reaching the overall goal of 40% reduction at the bay-wide
level.
In addition, the reduction target is to be reevaluated by December, 1991 based on
modeling, research, and monitoring results.
Program Objectives
The first step in developing a bay-wide monitoring program was to review historical
data. Although there were some data shortcomings, the Chesapeake Bay Program was
nevertheless able to demonstrate degraded environmental conditions.
Three general program objectives were developed (Magnien and Haire, 1989):
Quantify the extent and nature of water quality problems
Determine the response of key water quality variables to management
actions
Develop and test hypotheses on how the Bay ecosystem functions, espe-
cially as it relates to anthropogenic stresses and management solutions
-------
A-28
Based on eutrophication concerns and the complexity of the Chesapeake Bay, it was
also clear that the monitoring network would need to:
Incorporate physical, chemical, and biological variables
Collect mainstream and major tributary data
Collect data consistently over an extended period of time
Determine the optimal sampling strategy for each variable
As a result of data characterization and development of program objectives, six
monitoring program components were formed and are summarized below (Magnien
and Haire, 1989 and Chesapeake Bay Program, 1989a):
Chemical/Physical Variables: salinity, temperature, Secchi depth, dissolved
oxygen, suspended solids, nutrient species (nitrogen, phosphorus, carbon,
and silicon), phytoplankton pigments, heavy metal and organic compounds
in surficial sediments
Phytoplankton: species counts, phytoplankton pigments by horizontal and
vertical in vivo fluorescence, primary productivity, light penetration
Zooplankton: micro (44 ^im - 202 |xm) and meso (>202 um) species counts,
biomass
Benthic Organisms: species counts, production, sediment characteristics,
salinity, dissolved oxygen
Ecosystem Processes: sediment-water column exchange rates of dissolved
inorganic nutrients (nitrogen, phosphorus, and silicon) and oxygen,
surficial sediment characteristics, deposition rates of particulate matter
(total seston, nitrogen, phosphorus, carbon, phytoplankton pigments)
River Inputs: flow, suspended solids, nutrient species (nitrogen, phosphorus,
carbon, and silicon), phytoplankton pigments
Variable selection and spatial/temporal intensity were evaluated in order to provide the
level of information necessary to support confident management decision making. In
general, historical station locations were used as initial station locations and were
-------
A-29
"fill&A-'m" with additional stations to provide adequate spatial coverage. Selected
sampling frequency was either monthly or bimonthly.
1991 Reevaluation Study
A review of the progress made toward the goals established in the Nutrient Reduction
Strategy will be completed by December 1991. Included in this review is the develop-
ment of a three dimensional dynamic model that will overcome previous limitations
and permit more precise estimates.
A2.3 Evaluation
of Monitoring
Program
Performance
The use of statistical power analysis to evaluate expected monitoring program perfor-
mance is demonstrated below using a subset of data from the Chesapeake Bay Moni-
toring program. The purpose of evaluating monitoring program performance is not to
determine whether the goals of the nutrient reduction program have been met but
rather to determine what is the minimum change that the monitoring program can be
expected to detect. This information can be used to address a series of questions
regarding the adequacy of the monitoring effort. Examples include:
Are sufficient data being collected to detect a change in ambient nutrient
concentrations equal to the expected decrease?
How long (e.g., number of years) will it take to detect the level of change
predicted by the numerical models?
This example includes a description of the data obtained from the Chesapeake Bay
Program. Several analytical techniques used to explore the statistical nature of the
data are also described. The purpose of these analyses is to identify characteristics of
the data that might affect the applicability of statistical tests. Next, statistical tests for
trends are described, and results of the analysis of the example data set are presented.
Statistical trend analyses were conducted in order to determine whether a trend could
be detected. Finally, a series of power analyses are then presented and discussed. It
should be noted that the results from the small subset of data used in this demonstra-
tion is not necessarily indicative of the expected performance of the Chesapeake Bay
monitoring program. An analysis of the nutrient data that has been collected from
throughout the bay would be required.
However, this example describes the use of general procedures can be applied to these
and other types of monitoring data.
-------
A-30
Example Data Set and Sampling Layout
For this example, data were obtained for dissolved oxygen and ammonium as nitrogen
from 15 stations. These stations were chosen because they provided adequate spatial
coverage as well as having similar sampling methodologies. The station locations
(four stations from segment 3, ten stations from segment 4, and one station from
segment 5) are shown in Figure A2-1. The principal characteristics of the segments
are (Flemer et al, 1983):
Segment 3
Upstream limit of deep water anoxia
Moderate salinity (7-13 ppt. mean)
Two-layer, estuarine circulation driven primarily by freshwater inflow
Segment 4
Water deeper than 30 feet, oxygen depletion in summer, potentially toxic
to fish, crabs, shellfish and benthic animals.
Mean salinity of 9 to 14 ppt.
Rich in nutrients
Segment 5
Influenced by inflow from the Potomac and Patuxent and rich in nutrients.
Mean salinity of 10 to 17 ppt.
Subject to summer anoxia and contains most of the deeper Bay waters.
Observations of the water quality random variables are collected bimonthly (monthly
during winter months), for a total of 20 samples per year at four station depths. These
depths are denoted as surface, above the pycnocline, below the pycnocline, and
bottom. When the estuary was not stratified at selected station locations, the
pycnocline data were collected at equally spaced distances between the surface and
-------
A-31
Figure A2-1. Station
locations included in data
subset of the Chesapeake
Bay Ambient Monitoring
Program.
Washington, DC
-------
A-32
bottom. Approximately 66 months (June, 1984 through November, 1989) of data were
available for each water quality random variable.
The subgroup of stations from the Chesapeake Bay Program was selected based on the
physical grouping provided by Flemer et al. (1983). In this example analysis depth is
used to group station locations. Further studies could be developed to separately
analyze each segment as defined by Flemer et al. (1983).
Exploratory Analyses
Before applying a statistical test for trend, several aspects of the time series need to be
explored. These include: distributional characteristics (i.e., deviations from normal-
ity), the existence of trends, seasonality, and serial correlation.
The choice of the appropriate statistical test will depend on which of these characteris-
tics the data exhibit. The distributions of the data were shown to be non-normal. Time
series plots were examined visually for trend and seasonality. A series of monthly box
plots for each monitoring variable was made to examine the distribution of monthly
data and the seasonal pattern. Each box plot represents that month's data over six
years for a group of stations and shows the mean, median, and several percentiles (see
Figure A2-2).
Figures A2-3 and A2-4 show the time series plot for ammonia as nitrogen for the a
surface station and dissolved oxygen at a bottom station. The recurring sinusoidal
pattern indicates seasonality , a common occurrence in water quality data (Harcum
etal, 1990).
Figures A2-5 and A2-6 show the box plots for nitrogen as ammonia and dissolved
oxygen. If there was no seasonality, the means for all the months would be approxi-
mately the same. However, the monthly mean values vary depending on the month
which indicates that there is seasonal pattern.
The data were also tested for serial independence, (i.e., that the value at one timepoint
is dependent on the value of the previous timepoint). Most stations were found to
exhibit independence.
Statistical Tests for Trend
The choice of statistical test should be selected based on the data characteristics
described above. (Overall, the data were not significantly correlated; however, the
-------
A-33
data were seasonal and nonnormally distributed.) The Seasonal Kendall tau test was
selected since the test accounts for seasonality and is applicable to data that are not
distributed normally. Numerous researchers (e.g., Hirsch et at., 1982; Hirsch and
Slack, 1984; Berryman et al., 1988; Ward et al, 1988; Camoncho and Haywood,
1988; Taylor and Loftis, 1989; and Harcum et al., 1990) have also suggested that the
Seasonal Kendall tau test is an appropriate test for detecting gradual trends with the
type of data collected in the Chesapeake Bay Program and is typically more robust
over a wider range of conditions than alternative tests.
The Seasonal Kendall tau is a nonparametric test based on ranks. Each data point is
assigned rank based on its magnitude (smallest values get the lowest ranks, etc.). The
ranks are compared within a season (in this case, a season is a month) across the
years (May, 1984, May 1985, etc.) to determine if there is a trend. The results from
each season are then tallied and compared to a criterion value obtained from the
normal table.
The uneven sampling frequency of the example data set presents some potential
problems for the analysis described. Each sampling period (season) cannot be treated
equally since more samples were collected in the summer months. Additionally, the
Figure A2-2.
Example box plot.
Box Plots for Column X^
300 -
280 -
o 260 -
S. 240 -
E
O 220 -
0
c 200 -
.0
2 180 -
8 160 -
o
0 140 -
120 -
# > median +
3 x interquartile range
0
o
ฐ median +
1 .5 x interquartile range
+ -mean interquartile
range
25lh peicentile * '
j ^ median -
0 "* 1 .5 x interquartile range
8
^ ^ < median -
"* 3 x interquartile range
pซ
-
100 -J - i
Column
-------
A-34
STATION CBS.1
0.12
0,10
01 0.08
< 0.06
5
o
5 0.04
5
0.02 -
.X
O>
E
Z
UJ
CJ
o
o
UJ
>
_l
o
CO
in
a
0.00
16
14 -
12 -
10
8 -
6 -
4 -
2 -
0 -
Apr-84
o SURFACE
Apr-84 Jun-85 Jul-86 Aug-87
SAMPLE DATE
Sep-88
Oct-89
BOTTOM STATIONS
STATION ID
+ CB3.3E
ซ CB5.1
Jun-85
Jul-86 Aug-87
SAMPLE DATE
Sep-88
Oct-89
Figure A2-3. Time series
for N as NH3 for surface
stations.
Figure A2-4.
Time series for Dissolved
O2 for bottom stations.
-------
A-35
* -^
Figure A2 -5. Box plot for
N as NH3 for surface
stations.
V*riabie=NH.(Ma/t): Surface yซters
i i
I
S6P
I I
I I
Figure A2-6.
Box plot for dissolved 02
for bottom stations.
I I
I
JUM
-------
A-36
likelihood of violating the independence assumption of statistical tests for trends is
increased since summer samples were collected more frequently.
Two typical solutions for handling the unequal sampling are 1) averaging or 2) sub-
sampling (e.g., average monthly values to one monthly value or discard one value per
month during summer). In this case, past research (Harcum et al., 1990) has shown
that averaging data is the preferred method for this type of a monitoring program. For
the analyses presented below, average values were computed for 12 seasons (i.e., one
value was computed for each month). Equal weight was given for each month for long
term trend detection.
For more information about the Seasonal Kendall tau test see Hirsch, et al. (1982),
Hirsch and Slack (1984), or Taylor and Loftis (1989).
The results of the Seasonal Kendall tau tests are summarized in Table A2-1. At the
0.10 significance level, several stations showed statistically significant negative trends
in ammonium concentrations for both the surface and bottom samples. Positive trends
were also detected in dissolved oxygen concentrations in the bottoms waters at two of
the fifteen stations analyzed.
Several of the stations shown in Table A2-1 did not demonstrate a statistically signifi-
cant trend in ammonium or dissolved oxygen concentrations. However, the inability
to reject the null hypothesis (i.e., Ho: no trend in measured values) does not justify its
acceptance. The level of sampling effort or period of observation may not have been
sufficient to detect real trends in the data.
The reader is cautioned that the data preprocessing procedure used here is not appro-
priate for all applications. Consider a monitoring program composed of monthly
observations for the first five years and quarterly observations for the next five years.
It is not appropriate to average the monthly observations to quarterly values for
comparison with the last half of the data record. Rather, to apply the Seasonal Kendall
tau test, one must subsample from the earlier record (discarding two observations per
quarter) to create a uniform record. Depending on the exact circumstances, alternative
statistical tests may be more appropriate.
Statistical Power Analysis
Statistical power analyses were used to determine the minimum trend that could be
detected and therefore to determine the adequacy of the monitoring design.
-------
A-37
TABLE A2-L SEASONAL KENDALL TAU RESULT (oteO,10)
CB3.2
CB33C
CB3.3E
CB3.3W
CB4JC
CB4.1E
CB4.1W
CB4.2C
CB4.2E
CB4.2W
CB43C
CB4.3E
CB43W
CB4.4
CB5.1
CB3.2
CB33C
CB3.3E
CB3.3W
CB4JC
CB4.1E
CB4.1W
CB4.2C
CB4.2E
CB4.2W
CB4.3C
CB43E
CB4J1W
CB4.4
CB5.1
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
No trend
Negative trend
No trend
No trend
Notretid
No trend
No trend
Negative trend
No trend
No trend
Negative trend
No trend
No trend
Negative trend
Negative trend
No trend
Negative trend
No trend
No trend
Negative trend
Negative trend
No trend
Negative trend
No trend
No trend
Negative trend
Negative trend
No trend
Negative trend
Negative trend
CB3.2
CB3.3C
CB3.3E
CB3.3W
CB4.1C
CB4.1E
CB4.1W
CB4.2C
CB4.2E
CB4.2W
CB4.3G
CB4.3E
CB4.3W
CB4.4
CBS.I
CB3.2
CB3.3C
CB3.3E
CB3.3W
CB4,1C
CB4.1E
CB4.1W
CB4.2C
B4,2E
CB4.2W
CB4.3C
CB4.3E
CB4.3W
CB4.4
CB5.1
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surfaces
Surface
Surface
Surface
Surface
Surface
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
Bottom
No trend
No trend
Nofcend
No trend
No tread
No trend
No trend
No trend
No trend
No trend
No trend
No trend
No trend
No trend
No trend
No trend
No trend
Positive trend
No trend
No trend
No trend
No trend
No trend
No (rend
No trend
No trend
No tread
No trend
No trend ;
Positive trend
For some statistical tests, it is possible to develop analytical solutions to calculate
power based on the statistical characteristics estimated from data (e.g., U.S. EPA,
1987). However for nonparametric tests such as the Seasonal Kendall tau, the stan-
dard procedure for computing the test's power is to use Monte Carlo methods.
Monte Carlo simulation methods were used to evaluate the ability of the ongoing
monitoring efforts in Chesapeake Bay to detect specified levels of trend in water
quality variables.
The analyses were conducted in the following manner:
-------
A-38
1) Sampling spaces were defined to exhibit a specified magnitude of trend.
The distributional characteristics (e.g., mean and variance) of the sampling
population were otherwise similar to the monitoring data obtained from
Chesapeake Bay.
2) Random samples from the specified sampling space were generated by
computer programs. The simulated sampling frequency was the same
utilized in the Chesapeake Bay Monitoring Program. Sampling was
simulated for periods of 5 and 10 years.
3) Each sampling experiment was then repeated 500 times.
4) The simulated data in each experiment were analyzed using the Seasonal
Kendall Test, and the results of all 500 simulations were summarized to
obtain an estimate of the distribution of the test statistic. These results
were used to determine the percentage of simulations in which the null
hypothesis (no trend) was correctly rejected (i.e, the power of the test).
5) This series of steps was repeated for different magnitudes of trend, and the
results were summarized to obtain the plots of the probability of detection
versus magnitude of trend shown in Figures A2-7 and A2-8.
In Figures A2-7 and A2-8, four power curves are displayed for ammonia as nitrogen
and dissolved oxygen. For each pair of lines (the two solid or the two dashed lines),
the line to the right is the projected power for the station with the most variability of
the 15 stations and thus least powerful and the line to the left is the station with the
least variability and thus the most powerful. The results from the other 13 stations
would fall in between these two lines. The solid and dashed lines represent projections
for 10 and 20 years of data. As one would expect a significant improvement in trend
detection results from ten extra years of data.
Once the power relationships have been developed, selected results may be read from
the appropriate figure. From Figure A2-8, for example, dissolved oxygen trends in
surface waters on the order of 0.062 to 0.129 mg/t-yr can be detected with 10 years of
data at an 80% probability (using the Seasonal Kendall tau test with a = 0.10). With
20 years of data and the same group of stations, trends on the order of 0.021 to
0.042 mg/t-yr can be detected with an 80% probability.
The results of these and similar analyses may now be used as a basis to modify, if
necessary, the monitoring program. Since this study evaluated a limited selection of
-------
A-39
Figure A2-7. Power
analysis for N as NH3 for
surface stations.
Figure A2-8. Power
analysis for dissolved O2
for bottom stations.
tc
III
O
Q.
a:
LU
O
a.
1.0
0.8 -
0.6 -
0.4 -
0.2 -
0.0
_/ minimum
variance
Variable: Ammonium as N
Location: Surface
10 years of data
20 years of data
0.000
0.001 0.002 0.003
SLOPE (mg//-yr)
0.004
minimum
variance
Variable: Dissolved Oxygen
Location: Bottom
10 years of data
20 years of data
minimum detectable
trend for a power
level of 0.8
0.0
O.C
0.04
0.08 0.12
SLOPE (mg//-yr)
0.16
0.20
-------
A-40
water quality random variables, it is not appropriate to make recommendations for the
Chesapeake Bay Program. An extended analysis of other water quality random
variables would be needed and balanced against other monitoring objectives, not just
long term trend detection.
The type of results presented in Section A2.3 can be used to evaluate the performance
of individual monitoring program components. By applying this type of analysis to all
program components, the management team can then begin to assess the strengths and
weaknesses of the overall monitoring effort. This information can be useful in deter-
mining where limited financial resources should best be applied.
With the type of information generated in the previous section or similar "what-if'
power analysis scenarios, some general questions can be answered:
Which stations are most effective for detecting trends?
Which water quality variables are the most useful for detecting trends?
By answering these questions, it is possible to modify the sampling frequency (to
collect more or less data), deleting or moving stations, or narrowing or expanding the
water quality random variables sampled. By coordinating decisions with the other
monitoring objectives, improved evaluation of the physical, chemical, and biological
integrity of the estuary is possible.
A2.4 Use of
Power Analysis
Results
Batiuk, Rich. 1990. Chesapeake Bay Liaison Office. Personal communication.
Berryman, D., B. Bobee, D. Cluis, and J. Haemmerli. 1988. Nonparametric tests for
trend detection in water quality time series. Water Resources Bulletin, 24(3):545-556.
Box, G.E.P. and G.M. Jenkins. 1970. Time Series Analysis, Forecasting and Control.
Holden-Day, San Francisco, 553 p.
Brockwell, P.J. and R.A. Davis. 1987. Time Series, Theory and Methods. Springer-
Verlag,NewYork,519p.
A2.5 References
-------
A-41
Camacho, R. and H.C. Haywood. 1988. An Application of Power Analysis for
detection of trends in water quality. Interstate Commission on the Potomac River
Basin.
Chesapeake Bay Program. 1988. Basinwide Nutrient Reduction Strategy. Prepared in
Response to the 1987 Chesapeake Bay Agreement.
Chesapeake Bay Program. 1989A. Chesapeake Bay Basin Monitoring Program Atlas,
Volume I & II.
Chesapeake Bay Program. July, 1989. Water Quality Data Management Plan, Revi-
sion 1. United States Environmental Protection Agency, CBP/TRS 31/89.
Chesapeake Bay Program Monitoring Subcommittee. 1989. The State of the Chesa-
peake Bay, Third Biennial Monitoring Report -1989. 33 pages.
Elster, HJ. and W. Ohle. 1985. An Experimental Test of the Relative Significance of
Food Quality and Past Feeding History to Limitation of Egg Production of the Estua-
rine Copepod Arcartia tonsa. Arch. Hydrobiol. Beih. Ergebn. Limmol. 21:235-235.
Flemer, D.A., L. C. Davidson, K. Price, G. B. Mackiernan, and B. Johnson. 1983. A
monitoring and research strategy to meet management objectives. Appendix F, A
Framework for Action. Philadelphia: U.S. EPA.
Fulton, R.S. and H.W. Paerl. 1987. Toxic and Inhibitory Effects of the Blue-Green
Alga Microcystis aeruginosa on Herbiverous Zooplankton. J. Plankton Res.
9(5):837-855.
Gearing, J.N., PJ. Gearing, D.T. Rudnick, A.G. Requejo, and MJ. Hutchins. 1984.
Isotopic Variability of Organic Carbon in a Phytoplankton-Based, Temperate Estuary.
Geochim. Cosmochim. Acta. 48(5):1089-1098.
Grassle, J.P. and J.F. Grassle. 1984. The Utility of Studying the Effects of Pollutants
on Single Species Population in Benthos of Mesocosms and Coastal Ecosystems,
pp. 621-6432. In Concepts in Marine Pollution Measurement (HH. White, ed).
Maryland Sea Grant Publication, University of Maryland, College Park, MD.
-------
A-42
Grassle, J.F., J.P. Grassle, L.S. Brown-Leger, R.F. Petrecca, and N.J. Copley. 1985.
Subtidal Macrobenthos of Narragansett Bay, Field and Mesocosm Studies of the
Effects of Eutrophication and Organic Input on Benthic Populations, pp. 421-434. In
Marine Biology of Polar Regions and Effects of Stress on Marine Organisms (J.S.
Grey and M.E. Christiansen, eds.). John Wiley and Sons, New York, NY.
Harcum, J. B., J.C. Loftis, and R.C. Ward. 1990. Selecting trend tests for water
quality series with serial correlation and missing values. Water Resources Bulletin, in
review.
Hirsch, R.M., J.R. Slack, and R.A. Smith. 1982. Techniques of trend analysis for
monthly water quality data. Water Resources Research, 18(1):107-121.
Hirsch, R.M. and J.R. Slack. 1984. A nonparametric trend test for seasonal data with
serial dependence. Water Resources Research, 20(6):727-732.
Magnien, R.E. and M.S. Haire. 1989. Maryland's Chesapeake Bay Water Quality
Monitoring Program: An Estuarine Water Quality Information Systems. In Proceed-
ings of the International SYmposium on the Design of Water Quality Information
Systems. June 7-9, Colorado State University, Fort Collins, Colorado.
National Research Council. 1988. Marine Environmental Monitoring in Chesapeake
Bay. Marine Board, NRC, 2101 Constitution Avenue, Washington, D.C., 20418,
81 pages.
National Research Council. 1990. Managing Troubled Waters, The Role of Marine
Environmental Monitoring. National Academy Press. Washington, D.C., 125 pages.
Nixon, S.W., C.A. Oviatt, J. Frithsen, and B. Sullivan. 1986b. Nutrients and the
Productivity of Estuarine and Coastal Marine Ecosystems. J. Limmol. Soc. South Afr.
12(l/2):43-71.
Salas, J. D., J.W. Delleur, V. Yevjevich, and W.L. Lane. 1980. Applied Modeling of
Hydrologic Time Series. Water Resources Publications, Fort Collins, Colorado, 484 p.
Taylor, C.H. and J.C. Loftis. 1989. Testing for trend in lake and ground water quality
time series. Water Resources Bulletin 25(4):715-726.
-------
A-43
U.S. EPA. 1987. Bioaccumulation Monitoring Guidance: Strategies for Sample
Replication and Compositing, Volume 5. OW\OMEP EPA 430/09-87-003. Washing-
ton, D.C.
U.S. EPA. 1990. EPA nonpoint Source News-Notes. #5. June, 1990. Washington,
D.C.
Verity, P.O. 1987. Factors Driving Changes in the Pelagic Trophic Structure of
Estuaries, with Implications for Chesapeake Bay, pp. 35-56. In Perspectives on the
Chesapeake Bay: Recent Advances in Estuarine Sciences (M.P. Lynch and B.C.
Krome, eds.). Chesapeake Bay Program and Chesapeake Research Consortium, U.S.
Environmental Protection Agency, Gloucester Point, VA. CBP/TRS 16/87. CRC
Publication No. 127.
Ward, R.C., J.C. Loftis, H.P. DeLong, and H.F. Bell. 1988. Groundwater quality: a
data analysis protocol. Journal Water Pollution Control Federation, 60(11):1938-1945.
-------
A-44
-------
V. "W
. ,A '
V
Appendix B - Methods
-------
B-3
>
Methods - Introduction
This section identifies and describes the primary set of sampling and analysis methods
that will be used in estuarine monitoring programs. The primary purpose of this
section is to provide program coordinators and scientific staff of the estuary programs
with a basic understanding of the sampling methods as well as information on feasibil-
ity and the use of the monitoring program results. As stated in Section 4.2, standard-
ized sampling and analytical protocols or a performance based quality assurance
program should be developed in each estuary. This strategy is extremely important to
ensure comparability of data. The methods and recommendations described in Appen-
dix B can be used to guide decisions regarding the selection of individual techniques
as well as appropriate quality assurance and quality control procedures.
The primary sources of information for these method descriptions were documents
developed by the EPA's Office of Marine and Estuarine Protection and the Puget
Sound Estuary Program Protocols (described below). Individuals from EPA's Office
of Research and Development, the National Oceanographic and Atmospheric Admin-
istration, the U.S. Fish and Wildlife Service, university research laboratories and
ongoing National Estuary Programs were also contacted for information on latest
developments and ongoing research.
B.1 Methods The methods that are described in this section (Table B-l) are grouped for presentation
Chapter Format into four categories: water quality, sediment quality, biological resources and human
health risks.
-------
B-4
The format that is used to describe the selected sampling methods is shown in Table B-2.
Following a brief description of the monitoring method and a definition of terms, the
rationale for monitoring is described in terms of what information will be generated
||^i^i!!||iiiiHi||
:::::::1if';:ง:m0tii|Bn
ti
i||;|3|;!
||l|^||is|||l|i||||i||||^^||||^||||ft
^^^^^^^^j^^^^^^^^^^j^^^pi^-
^
:;Ji
-------
B-5
-. \ * ^ > *y- * *-f f ^ > Sr ^ ^ ^^^ J-.X s"1 *-%
and how that information can be used to address the goals and evaluate the success of
the CCMP. Major monitoring design issues such as selection of species for monitor-
ing or selection of temporal and spatial sampling strategies are addressed under the
topic of monitoring design considerations. Selected existing laboratory QA/QC and
analyses methods are described. Statistical design considerations include method-
specific statistical techniques to evaluate expected monitoring program data. The data
use section provides guidance for data interpretation and use in addressing issues of
concern. All information presented for each monitoring program component is
summarized. Each chapter emphasizes the importance of coordinating monitoring
program components to assure an effective, cost-efficient monitoring program.
As part of the task of summarizing available information, emphasis is placed on
providing references to existing documents. Currently, there is a major cooperative
effort underway within EPA, the U.S. Army Corps of Engineers, the National Oceanic
and Atmospheric Administration, and the National Bureau of Standards to develop a
Compendium of Methods for Estuarine and Marine Environmental Studies. The draft
document describes methods for nutrient measurements in seawater (U.S. EPA, 1990).
Additional sections are under development. Two other important sources of monitor-
ing program guidance include documents developed by the Office of Marine and
Estuarine Protection for implementation of the 301 (h) permit program. These docu-
ments (Table BT3) provide useful information on sampling equipment, field and
laboratory methods, statistical techniques, and the design and implementation of
marine and estuarine monitoring programs. The Puget Sound Estuary Program, in a 2-
year effort, developed standardized protocols for measuring selected environmental
variables in Puget Sound. These protocols have been developed for the monitoring
topics shown in Table B-4. The process of developing these regionally standardized
protocols is described by Becker and Armstrong (1988).
This methods section is also intended to provide a summary of available information
and to address the most important issues associated with the design and implementa-
tion of the monitoring program. Issues common to all monitoring methods include
quality assurance/quality control (QA/QC) and statistical design. These two important
issues will be discussed further.
B.2 Quality Quality assurance (QA), implemented at the management level, focuses on guidelines
Assurance/Quality and procedures, such as delegation of authority and responsibility, to assure data
Control (QA/QC) quality. Quality control (QC) focuses on the technical activities, such as calibration or
Considerations interlaboratory studies, needed to achieve specific data quality. QA/QC is not just a
series of requirements and procedures but a management discipline whose end result is
-------
B-6
||fi6jSjf^
liihilpj^
^^^ffj^jj^W^^^^:
W^ijK^^mM^^^:-^ii-^^^
'^^^$^^^-^^^^^^^^^^j^
Wi^^x^^^^$Q^^^^^^.
y^j^j^j^^i^^i^^^^^ij^^^Mj^.
^i^j^^^ja^^^^^^^jv^^.
-------
B-7
, ^ .
;"*
>x%
TABLE B-3. TECHNICAL SUPPORT AND
GUIDANCE DOCUMENTS
(continued)
ซ A Simplified Deposition Calculation (DECAL) for Organ!
Accumulation Near Marine Outfalls. 1987, EPA 430/09-88-001.
ซ Technical Support Document for ODES Statistical Power
Analysis. 1987, EPA 430/9-87-005.
ซ Evaluation of Differential Loran-C for Positioning in
Nearshore Marine and Estuarine Waters. 1988,
Tetra Tech., Inc.* Draft Report, EPA Contract No. 68-C8-001.
Bioaccumulation Monitoring Guidance Series
Bioaccuraulation Monitoring Guidance: 1) Estimating the
Potential for Bioaccumulation of Priority Pollutants and
301 (h) Pesticides Into Marine and Estuarine Waters.
1985, EPA 503/3-90-001.
Bioaccumuiation Monitoring Guidance: 2) Selection of
Target Species and Review of Available Bioaccumulation
Data. 1985, EPA 430/9-86-005.
Bioaccumulation Monitoring Guidance: 3) Recommended
Analytical Detection Limits. 1985, EPA 503/6-90-005.
Bioaccuraulation Monitoring Guidance: 4) Analytical
Methods for US, EPA Priority Pollutants and 301 (h) Pesticides
in Tissue from Estuarine and Marine Organisms.
1986, EPA 503/3-90-001.
* Bioaccumulation Monitoring guidance: 5) Strategies for
Sample Replication and Compositing.
1987, EPA 430/9-87-003.
-------
B-8
lltllllK
validated and verifiable information. QA begins with effective and conscientious
work planning and ends with a carefully constructed set of checks and balances
designed to ensure that uncertainties have been reduced to a known practical minimum
(U.S. EPA 1987a). Establishment of data quality standards is fundamental to the
selection of monitoring methods, and ensures the comparability among data collected.
-------
B-9
Data Quality Objectives
A major part of the QA effort is establishment of data quality objectives (DQOs).
DQOs are specific, integrated statements and goals developed for each data or infor-
mation collection activity to ensure that the data are of the required quantity and
quality. One function of establishing DQOs, as discussed in Section 2.3, is to ensure
that monitoring/data collection studies yield data adequate to meet their intended uses.
DQOs are also used to delineate QA/QC programs specifically geared to the data
collection activities to be undertaken. Development of DQOs usually consists of three
processes: 1) decision definition, 2) data use and needs identification, and 3) data
collection program design. DQOs should be continually reviewed during data collec-
tion activities and any needed corrective action may be planned and executed to
minimize problems before they become significant.
DQOs play an important role in the selection of sample collection, sorting, and analy-
sis methods. In fact, the selection of monitoring methods should be driven by DQOs.
Any future modifications to monitoring protocols should be considered only after these
new methods meet established data performance criteria. Data collected with different
methods should not be compared unless information exists which supports such
comparisons.
The following methods chapters provide a discussion of existing monitoring methods.
These methods provide a starting point for considering monitoring methodologies.
New, cost-effective methods are sure to develop for certain disciplines in the coming
years. A key consideration when deciding whether new methods are to be incorpo-
rated into the monitoring program, is whether they meet the QA/QC program's
performance criteria. QA/QC considerations in the following methods chapters will
primarily discuss sample collection, processing, storage, and analysis.
B.3 Statistical Prior to the collection of data, it is necessary to determine how the data will be used to
Design support decisions. It is further recommended that analytical performance criteria (e.g.,
Considerations minimally acceptable Type I and Type II error) be defined to establish quantitative
expectations for the monitoring program. The link between data, performance and
decision-making should be specified a priori to ensure that appropriate data, and
spatial and temporal coverages are addressed by the monitoring plan.
Selection of the number of replicates is an important component of program design.
The use of power analyses, examining alternative sampling and compositing strategies,
-------
B-10
J ->fv& v> v*- $#. > ฃ *$Xyfi-ff&fA$ฃy
will lead to an effective monitoring design strategy. These statistical techniques may
mitigate the high costs of collecting and processing samples. As a general recommen-
dation, equal numbers of samples should be collected whenever possible to simplify
statistical analyses.
Composite Sampling
Composite sampling consists of mixing two or more replicates and/or samples. The
chemical analysis of a composite sample provides an estimate of an average contami-
nant concentration for locations and/or times comprising the composite sample (see
space and time bulking below). Advantages of the composite sampling strategy are:
provides cost-effective strategy when individual chemical analyses are
expensive
results in a more efficient estimate of the mean at specified sampling
locations
Because of the reduced sample variance, composite sampling may result in a consider-
able increase in statistical power. If the primary objective of a monitoring program is
to determine differences in contaminant concentrations among sampling locations,
composite sampling is an appropriate strategy.
Space-bulking consists of sampling from several locations and combining samples into
one or more composite samples. Time-bulking involves taking multiple samples over
time from a single location and compositing these samples. A disadvantage of using
space- and/or time-bulking strategies is that significant information concerning spatial
and temporal heterogeneity may be lost.
Compositing of tissues provides a means to analyze bioaccumulation when the tissue
mass of an individual is insufficient for the analytical protocol. If mixing species,
tissue composites are likely to be composed of different proportions of species, and
different numbers, age, and sex of individuals. These differences among composites
tend to confound whether patterns of tissue residue concentrations are due to differ-
ences in locations or interspecific differences in bioaccumulation. Mixing of species is
not recommended because of these confounding effects.
Composite sampling is not recommended if the objective of the monitoring program is
to determine compliance with specified sediment contaminant concentration limits,
since this sampling method does not detect the true range of sediment contaminant
-------
B-11
concentrations in the environment. The adoption of composite strategies will depend
upon the objectives of individual monitoring programs.
Statistical Power
The large degree of temporal and spatial variability observed for many ecosystem state
and rate variables requires collection of sufficient replicate samples to ensure an
accurate description of the measure of interest. However, increases in replication
increase sample processing costs. Power analyses assist in the allocation of sampling
resources (stations, replication, and frequency) with regard to program finances and
design.
Power analyses may be applied to determine the appropriate number of sample repli-
cates, and/or subsamples in a replicate composite, required to detect a specified
difference (U.S. EPA, 1987b). The number of replications required to detect a speci-
fied minimum difference is a function of the statistical power and the variance in the
data. Power analyses require a prior knowledge of the variability in the data. A best
guess or variation observed in historical data are often used initially in the design of
the monitoring program.
For composite samples, statistical power increases with the increase of the number of
subsamples in each replicate composite sample (U.S. EPA, 1987c). However, there
exists a diminishing return of statistical power with the addition of successive
subsamples to each composite. For composites of greater than ten replications, the
increase of power is negligible given typical levels of data variability.
To improve the power of a statistical test, while keeping the significance level con-
stant, increase the sample size. However, due to constraints in cost and time imposed
by the CCMP this option may not be available. Power analyses have shown that for a
fixed level of sampling effort, a monitoring program's power is generally increased by
collecting more replicates at fewer locations. The number and distribution of sampling
locations required to evaluate the effectiveness of the CCMP will be depend upon the
size and complexity of the estuary.
Power-Cost Analyses
The relative power of one design with respect to another is more meaningful when the
relative costs of implementing alternative designs are taken into consideration. Analy-
ses of power-costs are fundamental in selecting appropriate sample/replicate number,
sample location, and sampling frequency.
-------
B-12
A cost function which describes the cost per replicate sample is required. Power-cost
formulations for parametric statistical analyses are of the form:
Power-costi= (1 -
where i is a replicate sample collection and analysis scheme, c is the cost per replicate
sample, n is the number of replicate samples, and fij is the Type II error. Costs will
depend upon sample collection, sorting, and analysis equipment and protocols. Type
II error (B) is a function of the number of replicate samples and the variance detected
by a sampling and analysis scheme. The measured variance may be influenced by the
sampling equipment used and the unit replicate sample size (e.g., as the unit replicate
area increases, variance increases).
Total costs are usually fixed. Therefore, iterations of power-cost formulations for
various sample processing scenarios (e.g., sample collection, sorting, and analysis
schemes) and their comparisons will result in the selection of the appropriate monitor-
ing methodology. Sokal and Rohlf (1981) provide a series of formulas for calculating
the most efficient sampling design for a given cost.
APHA. 1985. Standard methods for examination of waste and wastewater. American B.4 Literature
Public Health Association, Washington, D.C. 1268 pp.
Becker, D.S. and J.W. Armstrong. 1988. Development of regionally standardized
protocols for marine environmental studies. Mar. Poll. Bull. 19(7):310-313.
Bros, W.E. and B.C. Cowell. 1987. A technique for optimizing sample size (replica-
tion). J. Expt. Mar. Biol. Ecol. 30: 21-35.
Cuff, W. and N. Coleman. 1979. Optimal survey design: Lessons for a stratified
random sample of macrobenthos. J. Fish. Res. Board Can. 36: 351-361.
Ferraro, S.P., F.A. Cole, W.A. DeBen, and R.C. Swartz. 1989. Power-cost efficiency
of eight macrobenthic sampling schemes in Puget Sound, Washington, USA. Can. J.
Fish. Aquat. Sci. 46: 2157-2165.
Gilbert, R.O. 1987. Statistical Methods for Environmental Pollution Monitoring.
New York, NY: Van Nostrand Reinhold Co.
Cited and
References
-------
B-13
Hirsch, R.M. 1988. Statistical methods and sampling design for estimating step trends
in surface-water quality. Water Resources Bulletin 24(3): 493-503.
Millard, S.P., and D.P. Lettenmaier. 1986. Optimal design of biological sampling
programs using the analysis of variance. Estuarine Coastal Shelf Science 22: 637-656.
Sokal, R.R., and FJ. Rohlf. 1981. Biometry. San Francisco, CA: W.H. Freeman
and Co. 859pp.
U.S. EPA. 1983. Methods for Chemical Analysis of Water and Wastes, 2nd Edition.
EPA 600/4-79-020. Environmental Monitoring and Support Laboratory, Cincinnati,
OH.
U.S. EPA. 1987a. Quality Assurance/quality control (QA/QC) for 301(h) monitoimg
programs: Guidance on field and laboratory methods. EPA 430/9-86-004. Office of
Marine and Estuarine Protection, Washington, D.C. 267 pp.
U.S. EPA. 1987b. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, D.C.
34pp.
U.S. EPA. 1987c. Bioaccumulation monitoring guidance: Strategies for sample
replication and compositing. Vol.5. EPA 430/9-87-003. Office of Marine and
Estuarine Protection, Washington, D.C. 51 pp.
U.S. EPA. 1988a. U.S. EPA Statement of Work for Organics Analysis: Multi-Media,
Multi-Concentrations. SOW No. 288.
U.S. EPA. 1988b. U.S. EPA Statement of Work for Inorganics Analysis: Multi-
Media, Multi-Concentrations. SOW No. 788.
U.S. EPA. 1990. Compendium of methods for estuarine and marine environmental
studies. Draft U.S. Environmental Protection Agency, Office of Marine and Estua-
rine Protection; U.S. Department of Commerce, National Oceanic and Atmospheric
Administration; and U.S. Army, Corps of Engineers. Washington, D.C.
Ward, R.C. and J.C. Loftis. 1986. Establishing Statistical Design Criteria for Water
Quality Monitoring Systems, Review and Synthesis. Water Resources Bulletin
22(5):759-767.
-------
B-14
-------
B-15
B1.0 Water Column Physical
Characteristics
B1.1 Rationale
Physical characteristics of the water column are typically monitored to evaluate
estuarine water quality. The objective of monitoring these physical characteristics is to
identify and detect spatial and temporal variations in water quality. These results may
be used to characterize the water column, to monitor estuarine hydrodynamics, and to
verify estuarine hydrodynamic models.
Many chemical and biological processes in the environment are affected, directly or
indirectly, by the physical characteristics of the environment (Thomann and Mueller,
1987). Collection of physical data also provide ancillary information necessary in
evaluating chemical and biological data.
B1.2 Monitoring
Design
Considerations
Monitoring schemes for physical characteristics usually involve in situ methods. Data
concerning the water column's physical characteristics can be collected at relatively
inexpensive costs; the expense of different methods is generally governed by the level
of automation. Typically, the location and frequency of sampling water column
physical characteristics are determined by other, more expensively collected water
column parameters. However, it is essential to standardize the monitoring design in
order to ensure the comparability of monitoring data throughout the program.
Sampling Depths
When using bottle sampling techniques, samples for temperature, salinity, and turbid-
ity should be taken at a minimum of four depths in the vertical profile: 1) one meter
below the surface, 2) one meter above the bottom, 3) one meter above the pycnocline,
and 4) one meter below the pycnocline. If the waters are too shallow or no stratifica-
tion occurs, it would be appropriate to take the later two samples at evenly spaced
distances between the top and bottom samples. However, features of water masses
recorded in the historical data (i.e. historical profiles of salinity, temperature, and
turbidity) and the collection of other data types (e.g. plankton community structure and
water chemistry) should be considered when establishing sample depths (Pond and
Pickard, 1983).
When using in situ methods [e.g., Conductivity Temperature Depth (CTD)] tempera-
ture and salinity measurements should be taken at one meter intervals over the entire
depth profile (to within 1 meter of the surface and bottom). Little additional cost is
incurred for this detailed characterization of the water column once the CTD is de-
ployed. In areas of high stratification, a smaller interval would be appropriate.
-------
B-16
A consistent vertical sampling design is recommended in order that comparisons may
be conducted between data collected at different locations and over time. For ex-
ample, a set of depths may be specified by the monitoring design; at a minimum, any
sampling of the water column must include data from this set of specified depths.
Spatial and Temporal Scale
Generally, as the scale and complexity (spatially and/or temporally) of circulation
increases, more current meters need be deployed to adequately describe the environ-
ment. The need for macroscale and detailed descriptions of circulation will depend
upon the objectives of the CCMP and the physical structure of the estuary.
The length of the sampling period will be determined by monitoring objectives. If
descriptions of seasonal variations and/or high frequency natural events (e.g. tides,
storms, etc.) are necessary, measurements should be collected over a period not less
than two years (Fischer et a/., 1976). Frequent sampling in fish and shellfish areas,
and other biologically sensitive areas may be required in order to detect rapid changes
in those habitats which may endanger these resources.
Methods available for monitoring physical characteristics include instruments that B1.3 Existing
range from simple mercury-filled thermometers that provide one observation of Analytical
temperature to state-of-the art CTD (Conductivity Temperature Depth) equipment that Methods
provide vertical profiles of salinity and temperature. A list of methods (and/or equip-
ment) discussed in this section are summarized in Table B1-1.
Temperature
Surface temperature measurements may be made with any good grade of mercury-
filled or dial type centigrade thermometer, or a thermistor taking care not to expose the
water sample to the sun or to the evaporative cooling effect of the wind (APHA, 1987).
For measuring subsurface temperatures the basic instrument is the "protected reversing
thermometer"; a mercury-in-glass thermometer which is attached to the water sam-
pling bottle (Pond and Pickard, 1983).
Bathythermographs (BTs) consist of a thermistor and an electronic pressure trans-
ducer. Wire extending from the BT to the vessel relay measures of temperature and
pressure. BTs may be operated from stationary or slow moving (< 4 knots) vessels.
Expendable bathythermographs (XBTs) are based on tracking an expendable, free-
falling bathythermograph. The system is designed to operate from a stationary or
-------
B-17
Ifllflllp^^
ฃttii;
t|$|ซ^^
||jซ;;|jdif^ :
llS^
;if:|;r^ei^:|||i|ify;;?I||||
S^^|^|;^SS|l^f
:l|:;::f|ป|; ||^i^Sri::illt>|:-
":f:|||--|||:|||^S:||*K;::
?iliifiii^:infiiiil
-------
B-18
moving vessel (< 15-20 knots) typically in deep waters. Wires extending from the
free-falling transducer to the vessel relay temperature and depth recordings. The wires
are cut when the probe reaches its maximum depth. BTs and XBTs do not record
measures of salinity. If it has been determined that density is not related to salinity,
bathythermographs may be a cost-effective tool. Monitoring of temperature and
pressure using XBTs is higher in cost since the probe is lost after each vertical profile.
Conductivity-Temperature-Depth (CTDs) meters are the tool of choice, since density
measures are typically dependent on salinity, temperature, and depth (see Density
below). Measures of density are important in identifying and tracking the movements
of water bodies and in determining the depth at which a water mass will settle at
equilibrium. CTDs coupled with a computer can continuously integrate temperature
and conductivity measurements with depth. CTD/computer provides real-time read-
outs which allow immediate modifications in CTD position and ensures comprehen-
sive water column coverage during the survey.
Salinity
The Knudsen titration method determines chlorinity by titration with a standard silver
nitrate solution and salinity is then computed from a standard formula (Pond and
Pickard, 1983). This titration method is practical but not convenient to use on board a
ship (APHA, 1987).
Conductive cells are used for in situ use and a variety of such instruments are now
available from several manufacturers (Pond and Pickard, 1983). A CTD unit, consist-
ing of conductivity, temperature, and depth sensors, is lowered through the water on
the end of an electrical conductor cable which transmits the information to recording
units on board ship (UNESCO, 1988).
Density
Density is an important identifying property of sea water because there is no process
by which this property changes, except by mixing, below the surface (Pickard and
Emery, 1982). Direct measures of density are slow (Pond and Pickard, 1983). De-
vices such as hydrometers, pycnometers, and magnetic float densimeters may be used
when great accuracy is not required.
In practice, the density - or the specific gravity - of sea water can be determined from
temperature, salinity, and pressure. Temperature, salinity, and pressure are measured
directly and the density is read from tables which have been prepared from laboratory
-------
B-19
determinations (e.g., Eugene, 1966). Algorithms for automated data processing have
been developed by UNESCO and are applicable for salinities as low as 2 ppt (Fofonoff
and Millard, 1983). For lower salinities, see Hill et al. (1986).
Depth
The simplest method of determining the depth of an instrument is to pass the wire
attached to the instrument over a meter wheel with a known circumference and a
counter. In calm conditions and negligible currents, the length of wire passed over the
meter wheel will be the actual depth.
Depth may also be estimated using a bourdon tube moving the slider of an electrical
potentiometer or a strain-gauge pressure transducer. The hydrostatic pressure exerted
on gauge or bourdon tube may then be converted to a depth (Pickard and Emery,
1982).
The water depth at a particular station is typically determined with acoustics. Depth-
sounders have accuracies of 0.3 to 0.7 ft (Clausner et al., 1986).
Turbidity
To determine the transmission of visible light through the water, the simplest device is
the Secchi disc (Pond and Pickard, 1983). Measurements of Secchi disc depth are
probably the most widely used means of estimating light penetration. The Secchi disc
is easy to use and can be used to estimate the attenuation coefficients for collimated
and diffuse light. As a result, the depth of euphotic zone may be estimated. Secchi
disk readings vary with the observer because of interpersonal differences in visual
ability and therefore, cautious must be exercised when comparing Secchi disk readings
taken by different observers.
ป
Turbidity is the measure of paniculate matter (plankton and suspended sediments) in
the water column. Measurements, either in situ or remote, are made by spectrophoto-
metric methods. In remote measurements, water samples are obtained from discrete
depths and then analyzed on deck. This method of sampling is quick and accurate but
is labor intensive and is restricted to sampling only at those depths where water
samples were taken. In situ sampling involves passing a light transmissometer (usu-
ally as part of a CTD system) through the water column. This provides for much
higher sampling rates and less labor. To correlate data taken from different instru-
ments all values should be expressed as a percent transmittance for a standard
pathlength.
-------
B-20
Current Measurements
There are two basic ways to describe fluid flow, the Eulerian method in which the
velocity (i.e., speed and direction) is stated at every point in the fluid, and the Lagrang-
ian method in which the path followed by each fluid particle is stated as a function of
time.
Eulerian Methods - Mechanical meters include Savonious rotors, ducted impellers,
drag inclinometers, and propeller-type meters. Although Savonious rotors have been
used for many years in estuarine work, they are not suited for operation in shallow
water environments which are exposed to wave and swell activity. Ducted impeller
current meters have been designed to reduce the problems associated with current
measurements in the presence of waves (Brainard and Lukens, 1975). Davis-Wellar
propeller type meters are also designed to operate in the presence of waves.
Electromagnetic current meters measure the instantaneous horizontal and vertical
velocity components at a flow sensor. The characteristics of these meters make them
suitable for use in the presence of waves and shallow water (McCullough, 1977).
Most recent electromagnetic current meters are equipped with several probes to
measure other physical characteristics of the receiving water.
Acoustic current meters determine the current velocity by emitting a short acoustic
pulse and measuring the Doppler frequency shift of the backscattered signal reflected
from small, passively transported particles in the water column (Appell and Curtin,
1990; Appell and Woodward, 1986). This instrument allows for non-intrusive, remote
measures of current velocity throughout the water column from a depth of 3 to over a
1000 meters. The current meter can be placed on a ship and sampling conducted while
the ship is underway, thereby allowing for a synoptic measurement of the current field
over a large area with the use of only one instrument. Acoustic current meters are
suitable for use in the presence of waves but the accuracy of the measurements de-
creases close to the sea surface. These instruments are also expensive and require
highly trained technicians to maintain them.
Another technology employs the use of land based radar systems sending out electro-
magnetic waves and measures the reflected energy from the surface waves (Appell and
Curtin, 1990; Appell and Woodward, 1986). Since this system is set up on the shore it
only works out to a few dozen kilometers offshore and only measures the current
velocity in the top meter of the water column. This system does have the advantage in
that a ship is not needed to deploy or retrieve the instrument.
-------
B-21
Lagrangian Methods - Most drogues are drogue-buoy systems consisting of a small
marker buoy which is tracked at the surface and a larger submerged drogue portion
which is set at the desired depth by connecting line between the two portions (U.S.
EPA, 1982). The drogue portion must be weighted and ballasted so that the drogue
assembly has sufficient negative buoyancy to keep both the drogue and connecting line
in their intended vertical orientation, and keep the buoy mast upright. Drogues are
intended to passively drift with the currents at a specified depth. In reality, some error
is introduced in the drogue trajectories by wind drag on the exposed portion of the
marker buoy, by the relative surface current drag on the submerged portion of the
surface buoy, and, for deep drogues with long lines, by the relative current drag on the
connecting line. If possible, it is best to avoid drogue studies under high wind condi-
tions, especially when measuring lower current speeds in deeper waters. Drogues can
be used from just a couple of days to many months and can be tracked by satellite.
Most drogues can be classified into one of the following four categories: 1) Parachute
drogues, 2) Cruciform drogues, 3) Window Shade drogues, and 4) Cylindrical drogues.
Of these type, parachute and cruciform drogues are the most widely used (U.S. EPA,
1982).
Surface drifters are used to measure the average path of currents at the surface. Drift
bottles and drift cards are the types most commonly used. Vertical drift bottles and
drift cards are useful for evaluating the movement of effluent which reach the surface.
Horizontal drift cards, which float on the surface, may be used to determine the
potential movement of surface slicks due to oils or other floatables which form a
surface film. Because these movements are influenced largely by the wind, horizontal
drift cards should be released under several different conditions (Grace, 1978).
Seabed drifters measure the average path of currents near the sea floor. They are
useful for determining the fate of waste materials subject to transport by bottom
currents. This includes settleable solids and any portion of the effluent which remains
near the bottom. The drifters provide information on the net movement of a waste
field along the bottom, including where the waste field may reach the shoreline, and a
rough estimate of how long it may take. If a sufficient number of drifters are recov-
ered, they may indicate areas of possible shoreline contamination (Grace, 1978).
B1.4 QA/QC All environmental monitoring programs should have a written and approved quality
Considerations assurance project plan (U.S. EPA, 1987a). Water column sampling devices should be
checked for proper functioning prior to the survey. It is prudent to have a backup
-------
B-22
O"" '"'"%
sampler onboard the survey vessel in case the primary sampler is found to be unsuit-
able during the cruise.
Tables Bl-2 and Bl-3 lists sample preservation and storage requirements and recom-
mended methods for salinity and temperature (U.S. EPA, 1987a). When several
methods are available, the selection should be made by comparing the accuracy and
precision of the candidate methods for the parameter range expected at the site. All
monitoring programs should follow manufacturers recommended calibration proce-
dures.
Each temperature-measuring instrument should be calibrated against a precision
thermometer certified by the National Bureau of Standards at least every week (U.S.
EPA, 1987b). It is recommended that calibration be calculated daily when temperature
variation is suspected. Temperature probes may, at best, be accurate to within one-
tenth of a degree Celsius. Temperature probe systems are rarely linear over large
temperature ranges and must be checked against research grade laboratory thermom-
eters (APHA, 1987).
'^^^j^g^j^^^f^y^^^^^j^.]^iM&
^^^^BS^fffl^KSS^KBWISiS
lit)
'Note: :^;|pi:!^^
-------
B-23
TABLE Bl-3
RECOMMENDED ANALYTICAL METHODS
Parameter
Method
Precision
Significant
Minimum Digits
Detection Desired
Salinity Induction Salinometer titration;
or titration, (SM 14th ฑ 0.05 ppt
ed.,p,107,Sec,209C)
Temperature Bathytfiermograph or ฑ 0,05"C
Thermometric. EPA
Method 170,1. (SM 14th
ed., p. 125, Sec. 212)
Ippt
The water column, being a 3-D medium, requires greater numbers of samples to be
collected (vy, a 2-D medium) in order to be adequately described.
Salinity probe systems offer moderate accuracy but should be cross-checked by
discrete water samples analyzed by induction type laboratory salinometers (APHA,
1987). Two standard determinations should be made before the start of each series of
samples. In addition, one standard sample should be analyzed, to monitor instrument
drift. It is recommended that duplicate determinations be made for at least 10 percent
of the samples analyzed (U.S. EPA, 1987a).
Many multi-probe in situ measurement systems incorporate depth measurements by
use of pressure transducers. The accuracy and precision of such system must be
periodically checked. A pressure calibration can be done in a laboratory pressure
chamber. Daily calibration may be done in calm waters of a known depth (e.g., inside
a lagoon or a marina). Precision is determined by multiple measurements at the same
depth. Accuracy is evaluated by comparison to measurements made with a heavily
weighted line.
A standard suspension of Formalin, prepared under closely defined conditions, is used
to calibrate the nephelometer (U.S. EPA, 1974). The nephelometer should be cali-
brated at the start of each series of analysis and after each group of 10 successive
-------
B-24
samples. Duplicate analyses should be conducted on at least 10 percent of the total
number of samples (U.S. EPA, 1987b).
Furthermore, the following recommendations should be considered:
current meters should be calibrated before and after each major deployment
of instruments
UNESCO (1988) presents details on the calibration of CTD instruments
transmissometers can be calibrated in air and must be kept strictly cleaned
fluorometer calibration prior to use with a series of prepared concentra-
tions, blanks and spikes are recommended quality control checks (Wilson
era/., 1986)
In general, one expects to find temporal and spatial patterns in physical characteristics. B1.5 Statistical
The primary purpose of monitoring the physical characteristics of the water column is Design
to describe these patterns (e.g., range, seasonal variations) and to determine the Considerations
primary physical mechanisms affecting the monitored area.
As a result, statistical comparisons between sampling locations and sampling periods
are not common except for turbidity. In this instance, the reader is referred to the
Water Chemistry section for analysis approaches that would be appropriate.
Temporal A nalyses
Tidal and seasonal variability in the distribution and magnitude of several water
column physical characteristics are typically observed (Day et al., 1989). Time series
analyses - e.g. temporal autocorrelation, spectral analyses - may be necessary to
examine the effects of these cyclical influences and/or to filter out these cyclic forcing
functions in order to examine long-term temporal trends.
Graphical Representations
The physical characteristics data described in this section are typically summarized in
graphical form; the most common being temperature-depth profiles, salinity-depth
profiles, and temperature-salinity plot (T-S plot). Alternatively, horizontal sections
(e.g., contour maps) of properties are useful for displaying geographical distributions.
-------
B-25
Multiple physical characteristics may be displayed by displacing one graph above the
other. More details about analyzing characteristic diagrams can be found in Pickard
and Emery (1982).
Current data can be displayed in numerous fashions including stick plots, a time series
of rectangular components, and horizontal maps (Pickard and Emery, 1982). These
figures are typically plotted after periodic components are removed through spectral
analysis leaving only the residual non-periodic components.
B1.6 Physical characteristics may be used to determine water column stability, to character-
Use Of Data ize estuarine conditions, and to identify existing and potential problems affecting
habitats offish and sensitive life-stages of marine organisms (e.g., upwelling areas).
Turbidity can be used to estimate the reduction of light transmittance. Reduction of
the depth to which sunlight penetrates, turbidity increase, causes a decrease in the
compensation depth of phytoplankton resulting in reduced biological growth
(Thomann and Mueller, 1987).
Physical characteristics data are used to calibrate and validate hydrodynamic models
(Davis, 1988; Leighton et al, 1988; U.S. EPA, 1982). If mathematical models will be
used to assess the effects of ambient physical characteristics, the monitoring program
should be designed to provide the model with the required boundary condition data. It
may also be necessary to collect meteorological data (e.g., wind speed and direction,
temperature, pressure) for modeling efforts since estuary circulation can be affected by
wind effects.
Data concerning salinity, temperature, and currents may be used in interpreting the
temporal and spatial variability of distributions of particular species and benthic
infaunal communities.
B1.7 Rationale
Summary and
Recommendations Most chemical and biological processes in the marine environment are
affected, directly or indirectly, by physical characteristics of the marine
environment (Thomann and Mueller, 1987)
Physical characteristics may be used to evaluate hydrodynamics and to
provide ancillary information to interpret other water chemistry variables
-------
B-26
Monitoring Design Considerations
Temperature, salinity, density, turbidity, and depth (depth of sample and
depth to bottom) should be collected at one meter depth intervals at every
sampling location; sufficient data should be collected in the vertical profile
for temperature and salinity
Sampling frequency should be determined by the estuary program's
requirement for detecting seasonal or high frequency events (e.g. tides,
storms, etc.); measurements should be collected over a period of not less
than two years in order to examine seasonal effects (Fischer et al., 1976)
More frequent sampling in fish and shellfish areas, and other biologically
sensitive areas may be required
Existing Analytical Methods
Temperature measurements may be made with:
- mercury-filled or dial-type centigrade thermometer
- thermistor
- reverse thermometer
- bathythermograph
- CTD instrumentation
Salinity measurements are preferably made with conductivity cells (CTD
instruments)
Density measurements are to be calculated from salinity, temperature, and
pressure
Depth measurements are preferably made with hydrostatic pressure tech-
niques
Turbidity measurements may be made with:
- light transmissometer
- Secchi disc
-------
B-27
Current measurements may be made with:
- Eulerian methods - e.g. propeller-type meters, electromagnetic
meters, acoustic-doppler meters, etc.
- Lagrangian methods - e.g. drogues, surface drifters, etc.
QAIQC Considerations
Calibrate all equipment regularly
Blank and replicate analyses are recommended QC checks
Statistical Design Considerations
The primary purpose is to describe what the patterns (e.g., range, seasonal
variations) of physical characteristics are and to determine the primary
physical mechanisms affecting the monitored area
Statistical comparisons between sampling locations and sampling periods
are not common except for turbidity.
Common graphical summaries include:
- temperature-depth, salinity-depth profiles, and temperature-
salinity (T-S) plots
- horizontal sections (i.e., contour maps) of properties with mul-
tiple physical characteristics displayed by displacing one graph
above the other
Current data can be displayed with stick plots, a time series of rectangular
components, and horizontal maps (Pickard and Emery, 1982)
Use of Data
Physical characteristics are used to determine hydrodynamics and to
provide ancillary information to interpret other variables
Turbidity can be used to estimate the reduction of light transmittance.
Reduction of the depth to which sunlight penetrates, due to turbidity
increase, can reduce biological community growth (Thomann and Mueller,
1987)
-------
B-28
APHA. 1987. Standard methods for the examination of water and wastewater, 17th B 1.8 Literature
Edition. American Public Health Association, Washington, D.C. Cited and
References
Appell, G.F., and W.E. Woodward. 1986. Proceedings of the IEEE Third Working
Conference on Current Measurement. Institute of Electrical and Electronic Engineers,
New York.
Appell, G.F., and T.B. Curtin. 1990. Proceedings of the IEEE Fourth Working
Conference on Current Measurement. Institute of Electrical and Electronic Engineers,
New York.
Becker, D.S. and J.W. Armstrong. 1988. Development of regionally standardized
protocols for marine environmental studies. Marine Pollution Bulletin. 19(7) 310-
313.
Brainard, B.C. and RJ. Lukens. 1975. A comparison of the accuracies of various
continuous recording current meters for offshore use. Offshore technology confer-
ence, Paper No. OTC 2295.
Clausner, I.E., W.A. Birkemeier, and G.R. Clarke. 1986. Field comparison of four
nearshore survey systems. Miscellaneous Paper CERC-86-6, US Army Engineer
Waterways Experimental Station, Vicksburg, MS.
Davis, R.E. 1988. Modeling eddy transport of passive tracers. J. Mar. Res., 45, 635.
Day, J.W., C.A.S. Hall, W.M. Kemp, and A. Yanez-Arancibia. 1989. Estuarine
Ecology. New York, NY: John Wiley & Sons. 558pp.
Demmann, W.P., J.R. Proni, J.F. Craynock, and R. Fargen. In Press. Oceanic waste-
water outfall plume characterization measured acoustically.
Eugene, L.B. 1966. Handbook of Oceanographic Tables. Naval Oceanographic
Office, Rep. SP-68.
Fofonoff, N. P. and R.C. Millard, Jr. 1983. Algorithms for Computation of Funda-
mental Properties of Seawater. United Nations Educational, Scientific, and Cultural
Organization, UNESCO, Technical papers in marine science #44 Paris, France.
Goldstein, RJ. 1983. Fluid Mechanics Measurements. New York, NY: Hemisphere
Publ. Corp.
-------
B-29
Grace, R.A. 1978. Marine Outfall Systems Planning. Design, and Construction.
Englewood Cliffs, NJ: Prentice Hall, Inc. 600pp.
Hill K.D., T.M. Dauphinee, and D. J. Woods. 1986. Extension of the practical salinity
scale 1978 to low salinities. IEEE Journal of Oceanic Engineering Vol OE-11, #1, Jan
1986, pp 109-112.
Johns, W.E. 1988. Near-surface current measurements in the Gulf Stream using an
upward-looking acoustic doppler current meters. J. Atmos. Ocean. Technol., 5, 602.
Kennish, M.J. 1989. CRC Practical Handbook of Marine Science. CRC Press, Inc.,
Boca Raton, FL.
Koh, R.C.Y. 1988. Shoreline impact from ocean waste discharges. J. Hydraul. Div.
Am. Soc. Civ. Eng. 114:361.
Leighton, J.P. 1988. Verification/Calibration of a thermal Discharge Model. Proc. Int.
Symp. Model-Prototype Correlation of Hydraul. Structures, 148.
McCullough, J.R. 1977. Problems in measuring currents near the ocean surface.
Oceans '77, Marine Tech. Soc. and Inst. of Electrical and Electronics Engineering.
Muellenhoff, W.P., A.M. Soldate, Jr., D.J. Baumgarthner, M.D. Schuldt, L.R. Davis,
and W.E. Frick. 1985. Initial mixing characteristics of municipal ocean discharges.
Volume I- Procedures and applications. EPA-600/3-85-073a. U.S. Environmental
Protection Agency, Narragansett, RI. 90 pp.
Pickard, G. L. and W. J. Emery. 1982. Descriptive Physical Oceanography. An
Introduction. 4th (SI) Enlarged Edition. New York, NY: Pergamon Press, Inc.
Pond S., and G. L. Pickard. 1983. Introductory Dynamic Oceanography. 3rd Edition.
New York, NY: Pergamon Press, Inc.
Riggs, N.P. and K.A. Thompson. In Press. Current Profilers: Comparing three
different techniques.
Thomann, R.V. and J.A. Mueller. 1987. Principles of Surface Water Quality Model-
ing and Control. New York, NY: Harper and Row Publ. 644 pp.
-------
B-30
U.S. EPA. 1974. Methods for Chemical Analysis of Water and Wastes. Methods
Development and Quality Assurance Research Laboratory, National Environmental
Research Center, EPA-625-/6-74-003.
U.S. EPA. 1979. Methods for chemical analysis of water and wastes. Rept. No.
600/4-79-020 US EPA, Environmental Support Laboratory. Cincinnati, Ohio. 430 pp.
U.S. EPA. 1982. Design of 301(h) Monitoring Programs for Municipal Wastewater
Discharges to Marine Waters. EPA 430/9-82-010. Washington, D.C.
U.S. EPA. 1986. Quality Criteria for Water. Office of Water Regulations and Stan-
dards. EPA 440/5-86-001. Washington, D.C.
U.S. EPA. 1987a. Quality Assurance/Quality Control (QA/QC) for 301(h) monitoring
program: Guidance on field and laboratory methods. EPA, Office of Marine and
Estuarine Protection. EPA 430/9-86-004. Washington D.C..
U.S. EPA. 1987b. Puget Sound protocols. Final Report. 31pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1987c. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
UNESCO. 1988. The acquisition, calibration, and analysis of CTD data. UNESCO
technical paper in marine science 54. France. 92 pp.
Wallace, J.W., and J. W. Cox. 1976. Design, Fabrication and System Integration of a
Satellite Tracked, Free-Drifting Ocean Data Buoy. NASA Technical Memorandum
X-72817, January, 1976.
Wilson. 1986. Techniques of water-resources investigations of the United States
Geological Survey. Fluorometric procedures for dye tracing. Department of the
Interior, US Geological Survey. Washington, D.C.
Wright, S.J. 1988. Outfall Plume Dilution in Stratified Fluids. Proc. Int. Symp.
Model-Prototype Correlation of Hydraul. Structures, 148.
-------
B-31
" f
-------
B-32
Selection of sampling locations and choice of sampling techniques are crucial deci-
sions to be made before the sampling effort is initiated. Random sampling may not
detect "hot spots" associated with point source discharges. Too much sampling close
to a point source discharge will, however, bias the evaluation of overall ambient
conditions.
Water Column Sampling Equipment
Water column samples are frequently collected using water bottles. They are simple
devices, usually consisting of a cylindrical tube with stoppers at each end and a closing
device that is activated by an electrical signal. Each samples a discrete parcel of water
at any designated depth. Multiple samplers are fixed on a rosette frame in order that
several depths may be sampled during one cast and/or that replicate samples may be
collected at a particular depth. The most commonly used bottle samplers include the
Kemmerer, Van Dorn, Niskin, and Nansen samplers (U.S. EPA, 1987a). Alterna-
tively, a pump may be used to sample the water column (U.S. EPA, 1987a).
Samples should be taken at a minimum of four depths in the vertical profile: 1) one
meter below the surface, 2) one meter above the bottom, 3) one meter above the
pycnocline, and 4) one meter below the pycnocline. If the waters are too shallow or no
stratification occurs, it would be appropriate to take the later two samples at evenly
spaced distances between the top and bottom samples.
Caged Organisms
The California Mussel Watch Program and the National Oceanic and Atmospheric
Administration (NOAA) Status and Trends Program have employed the use of caged
transplant mussels to monitor bioaccumulation of toxic chemicals over space and time
(Ladd et al., 1984; Goldberg et al., 1978). Caged sentinel species offer several advan-
tages:
they concentrate contaminants (up to loMo5 fold vs. estuarine waters),
facilitating laboratory analyses of contaminant concentrations
they provide a means of temporally integrating water quality conditions
they provide an assessment of biological availability of water column
contaminants
they may be deployed and maintained in a number of diverse locales
-------
B-33
The Long Island Sound Estuary Program has shown that chemical residue analyses on
caged indicator species appears to be a promising approach for identifying sources of
pollution (U.S. EPA, 1982).
However, the disadvantages of caged organisms include:
the cost of transplanting organisms
the possibility of loss of the cage/buoy system
the possibility of introducing a "nuisance" species
Species have different bioaccumulation potentials for various contaminants. It is
recommended that multiple sentinel species be deployed in order to ensure that a
number of contaminants are sufficiently evaluated and a comprehensive characteriza-
tion of water contaminants be conducted. Unfortunately, formulations which would
allow comparisons of bioaccumulation data between different species and/or different
tissues types have not yet been developed; comparisons of tissue burdens among
species is not acceptable. Thus, it is essential that monitoring design elements be
standardized to allow for comparisons among estuarine studies.
B2.3
Existing Analytical
Methods
Chemical Analyses
Questions to be considered during the choice of an appropriate analytical method
include the parameters of interest, desired detection limits, sample size requirements or
restrictions, methods of preservation, technical and practical holding times, and matrix
interferences (especially from saline water). Saline waters cause additional problems
due to matrix effects. D'Elia et al. (1989) discuss the common analytical problems
encountered during monitoring analyses of water samples.
Selection of appropriate methods should consider a compromise between full-scan
analyses, which are economical but cannot provide optimal sensitivity for all com-
pounds, and alternate methods that are more sensitive for specific compounds, but can
result in higher analytical costs. Successive analyses can be accomplished in less time
than would be required by manual methods because each analysis is not carried to
completion, but is brought to the same stage of development and exposure by the
timing of the stream flow through the system. Possible interferences include sus-
pended solids, metal ions, and residual chlorine. A list of existing analytical tech-
niques is presented in Table B2-1.
-------
B-34
Dissolved Oxygen - The titrimetric, or Winkler, method is the first method of choice
for the measurement of dissolved oxygen (U.S. EPA, 1983). The membrane electrode
method is recommended for samples containing interferents such as sulfur compounds,
chlorine, free iodine, color, turbidity, or biological floes, or when continuous monitor-
ing is planned (U.S. EPA, 1983).
Nutrients - Nutrients such as ammonia nitrogen, total Kjeldahl nitrogen, nitrate-nitrite
nitrogen, and phosphorus are determined by spectrophotometric measurements using a
ifMWHiilSfll
Isffiffl^^^^^Bffl^^HjS
if". Ii||||dvel|||]|i||
If ; i^ipecfcil^l^C^ii
ijjjjjj^-/with mass .sp^l||fi^|j|งSSii^Sl
lllil ^:;^iy|iiii^
-------
B-35
---,-., - , , ,
'k* "v. Sv.*- v,XVปซ. -av^ j^O" \vi.
segmented continuous flow analyzer where samples and reagents are continuously
added in sequence separated by air bubbles and pumped through glass tubing (U.S.
EPA, 1983).
Ammonia nitrogen is determined by treating the samples with alkaline phenol and
hypochlorite to produce indophenol blue which is intensified with sodium
nitroprusside. Standard solutions should be made up using substitute sea water to
approximate the matrix of the samples (U.S. EPA, 1983). Total Kjeldahl nitrogen is
defined as the sum of free ammonia and organic nitrogen compounds which are
converted to ammonium sulfate under the conditions of digestion (U.S. EPA, 1983).
Sulfates react with nitrogen compounds of biological origin, but may not convert
nitrogenous compounds of some industrial wastes. For the determination of nitrate-
nitrite nitrogen, a filtered sample is passed through a granulated copper-cadmium
column to reduce any nitrate to nitrite. The nitrite is then transformed to a highly
colored azo dye which is measured spectrophotometrically.
Total phosphorus is determined by heating the sample in the presence of sulfates, then
cooling the sample and measuring phosphorous spectrophotometrically.
The fluorometric method for chlorophyll a. is more sensitive than the spectrophotomet-
ric method. The sample is subjected to an excitation wavelength and the fluorescence
is measured at a second emission wavelength (APHA, 1989). The greatest uncertainty
in the method is the choice of reference standard. HPLC is the most accurate of the
methods used for the analysis of chlorophyll (since all forms can be separately quanti-
fied), however it is also the most expensive.
Trace Metals - The choice of analytical method for trace metal analysis is determined
by the required detection limit. ICP allows the simultaneous measurement of several
elements. However, the achievable detection limits are usually not as low as those
obtained by graphite furnace or hydride AA. The combination of AA and ICP is the
recommended analytical method for detection of metals since no technique is best for
all elements. Cold vapor AA is the recommended technique for the analysis of mer-
cury (U.S. EPA, 1987a).
Graphite furnace AA is more sensitive than flame AA or ICP, but is more subject to
matrix and spectral interferences, which results in potential QC problems during the
analyses and uncertainties in the resulting data. Because of the lower concentrations
which can be seen by graphite furnace AA, particular caution must be taken with
regard to laboratory contamination. The concentration of each element is determined
-------
B-36
by a separate analysis, making the analysis of a large number of contaminant metals
both labor-intensive and relatively expensive compared to ICP. However, AA methods
may be cost-effective for the analysis of a few metals over a large number of samples.
Semi volatile Organic Compounds - Analysis of semivolatile organic compounds
involves a solvent extraction of the sample, cleanup of the extract, GC separation, and
quantitation (U.S. EPA, 1987a). There are two GC/MS options for detecting extract-
able organic compounds: internal standard technique and isotope dilution. The isotope
dilution technique is recommended because reliable recovery corrections can be made
for each analyte with a labeled analog or a chemically similar analog. This method is,
however, more expensive and less widely employed than the internal standard method,
which is the current method of choice in the Contract Laboratory Program. Mass
spectrometry provides positive compound identification by comparison of both reten-
tion time and spectral patterns with standard compounds.
The identification of pesticides and PCBs can be made by GC/ECD analysis (U.S. EPA,
1987a). GC/ECD provides greater sensitivity (lower detection limits) relative to
GC/MS, however GC/ECD does not provide positive compound identification. Since
the identity of a compound is determined solely by matching retention time with that of
a standard, confirmation of pesticides and PCBs on a second GC column is required for
confidence in the reliability of the qualitative identification of the compound. When a
compound is present at a high enough concentration, its identity should be confirmed
by the use of GC/MS.
Glass capillary GC achieves a much higher degree of resolution in the analysis of PCB
congeners than the standard packed column methods; however, few labs regularly use
this specialized technique. The LC/MS technique is being developed for the detection
and quantitation of nonvolatile organic compounds.
Volatile Organic Compounds - Analysis of volatile organic compounds also involves
a solvent extraction of the sample, cleanup of the characteristically complex extract, GC
analysis, and quantification (U.S. EPA, 1987a). The purge and trap GC/MS technique
is commonly employed for detecting volatile organic compounds in water. As in the
case of semivolatiles, GC/ECD may be used to achieve lower detection limits, although
this introduces a level of uncertainty to the qualitative identification of compounds.
The isotope dilution technique is recommended if the DQOs of the monitoring program
require accurate quantitation of each compound. This technique, however, carries
additional analytical time and expense.
-------
B-37
B2.4 QA/QC Sample Collection
Considerations
Sampling Gear - The primary criterion for an adequate sampler is that it consistently
collect undisturbed and uncontaminated samples. Water column sampling devices
should be inspected for war and tear leading to sample leakage upon ascent. It is
prudent to have a backup sampler onboard the survey vessel in case the primary
sampler is found to be unsuitable during the cruise.
In the field, sources of contamination include sampling gear, lubricants and oils,
engine exhaust, airborne dust, and ice used for cooling samples. Samples designated
for chemical analyses require all sampling equipment - e.g., siphon hoses, scoops,
containers - be made of noncontaminating material and be cleaned appropriately prior
to use. Potential airborne contamination - e.g., stack gases, cigarette smoke - should
be avoided. Furthermore, samples and sampling containers should not be touched with
ungloved fingers.
Sample Handling and Storage
Splitting water samples for chemical analyses should only be conducted with
noncontaminating tools under "clean room" conditions. Recommended container
materials, sample sizes, preservation techniques, and storage lifetimes for all metals of
concern in water are summarized in Table B2-2.
Laboratory Analyses
Laboratory performance as measured by method QA/QC protocols should be used to
evaluate and select appropriate analytical methods. Changes to laboratory protocols
should only be considered if new protocols meet established performance criteria.
Suggested QA/QC protocols and analytical performance criteria are discussed below.
QA/QC reports should describe the results of quantitative QA/QC analyses, as well as
other elements critical to the laboratory analyses of wate"r, sediments, and tissues to
ensure proper interpretation of the results. It is recommended that these reports be
recorded and stored in a database for future reference.
Field QA/QC Checks - Transfer blanks will indicate whether any contamination was
introduced or reagents in the field or introduced during shipping of samples. Cross-
contamination blank is designed to verify the absence of contamination carried over
from one sample to another due to inadequate cleaning of field equipment.
-------
B-38
TABLE B2-2.
SAMPLE PRESERVATION AND STORAGE PARAMETERS
Sample
Storage
Anatyte
Container* Size
Preservative Lifetime
Total or dissolved metals
(except Hg)
Total or dissolved Hg
Paniculate Metals
P,O(TFEb 1L HNO^pH<2 6 mo
G.TFB 250 mL HNO3pH<2 28 days
P,<3 Igal c *
a P = linear polyethylene, 0 = borosilicate glass, TFE = tetrafluoroethylene.
b If aliquot for Hg taken from this 1 L sample, cannot use linear polyethylene,
c Samples should be filtered as soon as possible and always with 24 h. Workshop
attendees recommend that filtering be done shipboard rather than in lab on shore.
Blind replicates, splits treated and identified as separate samples, may be sent to the
same laboratory for analysis or have one sample sent to a "reference" laboratory for
comparison. In addition, standard reference material may also be placed in a sample
container at the time of collection and sent "blind" to the laboratory.
Instrument QA/QC Checks - Calibration standards should be analyzed at the begin-
ning of sample analysis, and Method QA/QC Checks - Analysis of preparation blanks
should be conducted to demonstrate the absence of contamination from sampling or
sample handling in the laboratory. At least one method blank must be included with
each batch of samples and should constitute at least 5% of all samples analyzed.
Spike recovery analyses are required to assess method performance for the particular
sample matrix. Spike recoveries serve as an indication of analytical accuracy, whereas
analysis of standard reference materials measure extraction efficiency. Recommended
control limits include 75-125 percent recovery for spikes, and 80-120 percent recovery
for the analysis of standard reference materials (SRM).
Replicates must be analyzed to monitor the precision of laboratory analyses. A
minimum of 5% of the analyses should be laboratory replicates. The control limits are
-------
B-39
ฑ20 percent relative percent difference for duplicates. Triplicates should be analyzed
on one of every 20 samples or on one sample per batch if less than 20 samples are
analyzed.
Laboratory performance and calibration should be verified at the beginning and end of
each 12-hr shift during which analyses are performed.
Currently, there exists no universally agreed upon convention for reporting detection
limits of analytical procedures. Table B2-3 lists definitions of various detection limits
used by the American Chemical Society's Committee on Environmental Improvement
(CEI). The IDL does not address possible blank contaminants or matrices interfer-
ences and is not a good standard for complex environmental matrices. The LOD and
LOQ account for blanks, but not matrix interference. The MDL provides high statisti-
cal confidence but, like the LOQ, may be too stringent. Detection limits must be
specified and reported to ensure adequate data quality and comparability among
protocols.
l^
i(
:-:ij^
|fง|^^
-------
B-40
The statistical analysis of water chemistry data typically involves estimating summary
statistics and testing for spatial and temporal trends. Summary statistics are useful for
spatial displays (e.g. plume contours), load estimations, and general reporting. Testing
for trends is useful for evaluating whether water chemistry conditions have improved
or degraded over time or space. In comparison to other variables (see Bioaccumula-
tion Methods), it is not common to collect replicated or composite data except for QA/
QC procedures. This is the usual case since most monitoring designs typically opt for
collecting water chemistry data more frequently or at more stations.
Summary Statistics
To summarize data, the analyst typically estimates statistics for central tendency (e.g.
mean, median) and variability (e.g. standard deviation, interquartile range) of the data.
Uncertainty should be indicated by reporting estimates with confidence limits or
percentiles (Ward and Loftis, 1986). Data may also be summarized graphically with
plots (e.g. box and whisker plots, time series plots, scatter plots, contours, profiles,
etc.).
Another problem associated with different analytical methods is left censoring (i.e.,
data less than or equal to equipment detection limit). Porter et al. (1988) and Gilliom
and Helsel (1986) provide a discussion on the implications of analyzing left censored
data and alternative procedures for estimating summary statistics when the data are
censored. For the purposes of estimating summary statistics, it is generally recom-
mended to not perform statistics on left censored data.
B2.5 Statistical
Design
Considerations
The results may be used to:
establish spatial and temporal trends in the transport of pollutants dis-
charged into ambient waters
calculate nutrient budgets
calibrate and verify hydrodynamic models
develop water quality standards for receiving waters
In addition, these results may be used to monitor rates of recovery following environ-
mental interventions and provide early warnings of potential impacts to the estuarine
ecosystem.
B2.6
Use of Data
-------
B-41
Monitoring of pollutant levels in the water column is a widely-accepted means of
measuring the condition of the aquatic habitat. However, the singular use of pollutant
loading data to assess the condition of the water column or to guide the decision-
making process in not recommended. Data acquired from monitoring water contami-
nant levels, in conjunction with the water's physical properties, may be used to assess
the health risks to human populations. In addition, ecological risks to individuals,
populations, and communities living in the water column may be evaluated. Monitor-
ing of water column chemistry remains a powerful tool in the evaluation of spatial and
temporal effects of anthropogenic and natural disturbance.
B2.7 Rationale
Summary and
Recommendations Monitoring water quality parameters will provide information on ambient
water conditions and the potential for transport and persistence of contami-
nants in the aqueous environment
Monitoring Design Considerations
The objectives of the monitoring program must be clearly defined in order
that:
- the analytical methods show appropriate selectivity, specificity
and sensitivity for the contaminants of concern
- sampling and sample handling procedures are well-defined and
consistent so as not to compromise the integrity and representa-
tiveness of the samples
- sampling numbers and locations are appropriate to the level of
information required
Existing Analytical Methods
Dissolved Oxygen
- The Winkler (titrimetric) method is the first method of choice for
the measurement of dissolved oxygen, unless the samples contain
interferents such as sulfur compounds, chlorine, free iodine,
color, turbidity, or biological floes, or when continuous monitor-
ing is planned. In those instances, a membrane electrode should
be used
-------
B-42
Nutrients
- nutrients such as ammonia nitrogen, total Kjeldahl nitrogen,
nitrate-nitrite nitrogen, and phosphorus are determined by spec-
trophotometric measurements using a segmented continuous flow
analyzer
- the usual method for chlorophyll determination is fluorometric,
however the choice of method is determined by the need to
separate the different forms. HPLC is the most accurate method,
however it is also the most expensive
Trace Metals
- a combination of AA spectroscopy and ICP spectrometry is the
recommended method for the detection of metals
- cold vapor AA spectroscopy is the recommended protocol for the
detection of mercury
Organics
- for monitoring efforts where the most accurate quantitation is
important, the isotope dilution GC/MS method is recommended
for volatiles and semivolatiles. GC/ECD is the method of choice
for pesticides and PCBs
QA/QC Considerations
Calibration standards and blank, matrix spike, and replicate analyses are
recommended quality control checks
A report describing the objectives of the analytical effort, methods of
sample collection, handling and preservation, details of the analytical
method, problems encountered during the analytical process, any necessary
modifications to the written procedures, and results of the QC analyses
should be included with the quantitative data
Statistical Design Considerations
The analyst typically estimates statistics for central tendency (e.g. mean,
median) and variability (e.g. standard deviation, interquartile range) of
grouped data. Uncertainty should be indicated by reporting estimates with
confidence limits or percentiles
-------
B-43
V \s *
For the purpose of estimating summary statistics, it is generally recom-
mended not to censor data
Prior to the collection of data, the statistical test (and significance level)
used to analyze the data should be specified
Use of Data
Monitor ambient levels of pollutants in the environment
Establish spatial and temporal trends in the accumulation and transport of
pollutants discharged into ambient waters
Calibrate and verify mathematical models
Calculate nutrient budgets
Develop water quality standards for receiving waters
Identify noncompliant discharges
B2.8 Literature APHA. 1989. Standard Methods for the Examination of Water and Wastewater. 17th
Cited and Edition. American Public Health Association. Washington, DC.
References
Armstrong, J.W. and A.E. Copping. 1989. Comparing the Regional Puget Sound
marine monitoring with the NOAA National Status and Trends Program. Coastal
Zone Proceedings 3:2421-2435.
Becker, D.S. and J.W. Armstrong. 1988. Development of regionally standardized
protocols for marine environmental studies. Mar. Poll. Bull. 19(7):310-313.
D'Elia, C.F., P. A. Steudler, and N. Corwin. 1977. Determination of total nitrogen in
aqueous samples using persulfate digestion. Limnol. Oceanogr. 22:760-764.
D'Elia, C.F. et al. 1987. Nitrogen and phosphorus determinations in estuarine waters:
A comparison of methods used in Chesapeake Bay monitoring. Final Report to the
Chesapeake Bay Program, U.S. EPA Region III. US Government Printing Office,
Washington, DC.
-------
B-44
D'Elia, C.F. et al. 1988. Nutrient analytical services laboratory standard operating
procedures. Chesapeake Biological Laboratory, University of Maryland.
D'Elia, C.F., J.G. Sanders, and D.G. Capone. 1989. Analytical chemistry for environ-
mental sciences: A question of confidence. Environ. Sci. Technol. 23(7).
Gilliom, R.J. and D.R. Helsel. 1986. Estimation of distributional parameters for
censored trace level water quality data 1. Estimation Techniques. Water Resources
Research 22(2):135-146.
Gilbert, P.M. and T.C. Loder. 1977. Automated analysis of nutrient seawater:
Manual of techniques. Woods Hole Oceanographic Institute Technical Report No.
WH01-77-47.
Goldberg, E.D., V.T. Bowen, G.H. Farrington, J.H. Martin, P.L. Parker, R.W.
Risebrough, W. Robertson, E. Schneider and E. Gamble. 1978. The mussel watch.
Environ. Conserv. 5:101-125.
Hirsch, R.M. 1988. Statistical methods and sampling design for estimating step trends
in surface-water quality. Water Resources Bulletin 24(3):493-503.
Knezovich, J.P. and F.L. Harrison. 1987. A new method for determining the concen-
tration of volatile organic compounds in sediment interstitial water. Bull. Environ.
Contam. Toxicol. 38:837-940.
Ladd, J.M., S.P. Hayes, M. Martin, M.D. Stephenson, S.L. Coale, J. Linfield, and M.
Brown. 1984. California state mussel watch: 1981-1983. Trace metals and synthetic
organic compounds in mussels from California's coast, bays and estuaries. Biennial
Report. Sacramento, CA: Water Quality Monitoring Report No. 83-6TS.
Plumb, R.H. 1981. Procedure for handling and chemical analysis of sediment and
water samples. Technical Report EPA/CE-81-1. U.S. EPA and Corps of Engineers,
U.S. Army Engineers Waterways Experimental Station, Vicksburg, MS.
Porter, P.S., R.C. Ward, and H.F. Bell. 1988. The detection limit. Env. Sci. Tech.
22:856-861.
Salley, B.A., J.G. Bradshaw, and BJ. Neilson. 1986. Results of comparative studies
of preservation techniques for nutrient analysis on water samples. Virginia Institute of
Marine Science Report to the Chesapeake Bay Liaison Office, September 1986.
-------
B-45
Strickland, J.D.H. and T.R. Parsons. 1968. A Practical Handbook of Seawater
Analysis. Fisheries Research Board of Canada, Ottawa, Canada.
Tetra Tech. 1986. Recommended protocols for measuring selected variables in Puget
Sound. Final report submitted to Puget Sound Estuary Program by Tetra Tech, Inc.
Bellevue, Washington.
U.S. EPA. 1979. Handbook for analytical quality control in water and wastewater
laboratories. EPA 600/4-79-019. Environmental Monitoring and Support Laboratory,
Cincinnati, OH.
U.S. EPA. 1982. Methods for organic chemical analysis of municipal and industrial
wastewater. EPA 600/4-82-057. Environmental Monitoring and Support Laboratory,
Cincinnati, OH.
U.S. EPA. 1983. Methods for chemical analysis of water and wastes, 2nd Edition.
EPA 600/4-79-020. Environmental Monitoring and Support Laboratory,
Cincinnati, OH.
U.S. EPA. 1986. Test methods for evaluating solid waste. EPA Publication SW-846,
3rd Edition. Office of Solid Waste and Emergency Response, Washington, DC.
U.S. EPA. 1987a. Puget Sound protocols. Final Report. 31 pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1987b. Quality Assurance/Quality Control (QA/QC) for 301(h) Monitor-
ing Programs: Guidance on field and laboratory methods. EPA 430/9-86-004. Office
of Marine and Estuarine Protection, Washington, DC.
U.S. EPA. 1987c. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1988a. Compendium of methods for marine and estuarine environmental
studies. EPA 503/2-89/001. Office of Water, Washington, DC.
U.S. EPA. 1988b. US EPA Statement of Work for Organics Analysis: Multi-Media,
Multi-Concentration. SOW No. 288.
-------
B-46
U.S. EPA. 1988c. US EPA Statement of Work for Inorganics Analysis: Multi-Media,
Multi-Concentration. SOW No. 788.
U.S. EPA. 1990. Compendium of methods for estuarine and marine environmental
studies. Draft. U.S. Environmental Protection Agency, Office of Marine and Estua-
rine Protection; U.S. Department of Commerce, National Oceanic and Atmospheric
Administration; and U.S. Army, Corps of Engineers. Washington, D.C.
Valderama, J.C. 1981. The simultaneous analysis of total nitrogen and total phospho-
rus in natural waters. Mar. Chem. 10:109-122.
Ward, R.C. and J.C. Loftis. 1986. Establishing statistical design criteria for water
quality monitoring systems: Review and synthesis. Water Resources Bulletin
22(5):759-767.
-------
B-47
B3.0 Sediment Grain Size
B3.1 The objective of monitoring sediment grain size composition is to detect and describe
Rationale spatial and temporal changes of the benthic environment. The availability of sediment
contaminants are often correlated with the grain size composition of the benthic
medium; sediment contaminants are absorbed to small grain sediment surfaces.
Likewise, grain size information may explain the temporal and spatial variability in
biological assemblages; changes in sediment grain size often affect an infaunal
organism's ability to build tubes, capture food, and escape predation. These results
may be used to monitor rates of recovery following environmental interventions, to
evaluate the condition of benthic habitats, and to assist in providing early warnings of
potential impacts to the estuarine ecosystem.
B3.2
Monitoring Design
Considerations
Sediment Sampling Devices
Sampling crews have been given a wide latitude in how sediment samples are col-
lected. In fact, the protocols required to collect an acceptable surficial sediment sample
for subsequent measurement of physical characteristics have generally been neglected.
However, because sample collection protocols influence all subsequent laboratory and
data analysis, it is key that sediment samples be collected using acceptable and stan-
dardized techniques.
Collection of undisturbed sediment requires that the sampler:
create a minimal bow wake when descending
form a leakproof seal when the sediment sample is taken
prevent winnowing and excessive sample disturbance when ascending
allow easy access to the sample surface in order that undisturbed
subsamples may be taken
Penetration well below the desired sampling depth is preferred to prevent sample
disturbance as the device closes. It is optimal to use a sampler that has the means of
weight adjustment in order that penetration depths may be modified.
Several types of devices can be used to collect sediment samples: grabs and box
corers (Mclntyre et al., 1984). Many of these devices sample the benthic habitat in a
unique manner. Accordingly, conducting comparisons among data collected using
different devices is inadvisable.
-------
B-48
Grab samplers - Grabs are capable of consistent sampling of bottom habitats. De-
pending on the size of the device, areas 0.02 to 0.5 m2 and depths ranging from 5 to 15
cm may be sampled. Limitations of grab samplers include:
variability among samples in penetration depth depending on sediment
properties
oblique angles of penetration which result in varying penetration depths
within a sample
the sample is inevitably folded resulting in the inability to section the
sample and the loss of information concerning the vertical structure in the
sediments
However, the careful use of these devices will provide quantitative data. Grab sam-
plers are the tools of choice for a number of estuarine and marine monitoring programs
due to their ability to provide quantitative data at a relatively low cost (Fredette et al.,
1989; U.S. EPA, 1987a).
Core samplers - Box corers utilize a surrounding frame to ensure vertical entry;
vertical sectioning of the sample is possible (U.S. EPA, 1987a). These devices are
capable of maximum penetration depths of 15 cm and may collect volumes 5 to 10
times that of grab samplers. Limitations of box corers include:
its large size and weight require the use of cranes or winches and a large
vessel for deployment
higher construction expenses
lack of calibration studies to permit comparisons to grab samples
The Hessler-Sandia box corer uses dividers to section the core into subsections,
facilitating subsampling of the core. Box corers are recognized as the tools of choice
for maximum accuracy and precision when sampling soft bottom habitats.
Sediment Profiling Camera - The sediment profiling camera allows vertical in situ
imaging of the water-sediment interface from which grain size determinations may be
conducted. The sediment grain size determinations may be made at a maximum depth
of 18 cm, however, penetration depth of the viewing prism may be a limited due to the
-------
B-49
physical characteristics of the sediment (i.e. penetration depths are greater in silt than
sand).
Distinction between finer silts and clays are not possible. Furthermore, it is recom-
mended that quantitative calculations of sediment grain size from the image be
ground-truthed by the results of laboratory analysis of collected sediment samples.
The singular use of the sediment profiling camera is not recommended. However, the
sediment profile camera proves to be effective as a reconnaissance tool. Delineation
of habitats with similar physical characteristics may aid in the selection of appropriate
sampling stations.
Recommendations - Collection of sediments and benthic organisms should be done
concurrently in order to mitigate the costs of field sampling and to permit sound
correlation, regression, and multivariate analyses (see Benthic Community Structure).
Therefore, it is recommended that the sampling device also be suitable for benthic
sampling. Grab and core sampling devices permit adequate sampling of both sediment
and benthic infaunal communities with the one sampling device.
Sample Depth
It is recommended that the upper 2 cm of the sediment column be examined to charac-
terize surficial sediments. Although the 2 cm specification is arbitrary, it will ensure
that:
relatively recent sediments are sampled
adequate volumes of sediments are readily obtained for laboratory analyses
data from different studies may be compared
Sampling of depths other than 2 cm or vertical stratification of deeper sediment cores
may be appropriate, depending upon the objectives of the monitoring program. For
example, if information concerning only the most recent sedimentation events is
required, examination of the upper 1 cm may be appropriate. Stratification of deeper
cores will provide historical data of sediment grain size and depositional events.
Comparison of data from studies analyzing different sediment depths is not advised.
Penetration well below the desired sampling depth is preferred to prevent sample
disturbance as the device closes.
-------
B-50
Selection of Time of Sampling
Sediment grain size compositions are often temporally stable, although some slight
seasonal variability may be present. Changes are usually associated with seasonal
patterns of benthic turbulent mixing and sediment transport phenomena. The fre-
quency of sampling should be related to the expected rate of change in grain size
compositions. A consistent sampling period is recommended in order that spatial and
temporal comparisons may be conducted.
If seasonal variations are exhibited, it is recommended that direct comparisons be-
tween samples collected during different seasons be avoided. Studies investigating
interannual variation in the percent composition of grain sizes should be conducted
during the same season (preferably the same month) each year. Furthermore, it is
recommended that grain size be sampled when contaminant concentrations are ex-
pected to be at their highest level in order to evaluate worst-case scenarios.
True vs. Apparent Particle Size
Particle size determination can either include or exclude organic material. If organic
material is removed prior to analysis, the "true" particle size distribution is determined.
If organic material is included in the analysis, the "apparent" particle size distribution
is ascertained. Because true and apparent distributions differ, detailed comparisons
between samples analyzed by these different methods are questionable. It is therefore
recommended that measures of sediment grain size be examined using only one of
these methods. A standardized grain size analysis will allow all comparisons between
samples within each study and among different studies.
Sieving
Sieving separates the sediment sample into size fractions. Wet sieving separates the
sample into two fractions: grain size greater than 62.5 um (i.e., sand and gravel) and
less than 62.5 um (i.e., silt and clay). Aggregates should be gently broken and wet
sieving should continue until only clear water passes through the sieve.
The gravel-sand fraction can be further divided by dry sieving through a graded
series of screens (U.S. EPA, 1987a). The sample should be shaken on the mesh for a
pre-determined, standard amount of time. The last sieve should have a mesh size of
62.5 um (4 phi) and any material passing through this sieve should be added to the silt-
clay fraction of that sample. The silt-clay fraction can be subdivided using a pipet
B3.3 Existing
Analytical
Methods
-------
B-51
technique (U.S. EPA, 1987a). This technique depends upon differential settling rates
of different sized particles. Withdrawal time for pipet analysis, as a function of
particle size and water temperature, is given in Table B3-1 (U.S. EPA, 1987a). The
total weight of phi-size interval must then be calculated (U.S. EPA, 1987a).
It is key that sieving techniques and the desired number of subtractions be specified
and standardized to allow for comparisons between samples.
TABLE B3-1. SEDIMENT GRAIN SIZE:
WITHDRAWAL TIMES FOR PIPET ANALYSIS AS A FUNCTION OF
PARTICLE SIZE AND WATER TEMPERAT0REa>b
Diameter Diameter
Finer Finer \
than . than
(phi)
-------
B-52
It is critical that each sample be homogenized thoroughly in the laboratory before a B3.4 QA/QC
subsample is taken for analysis. Laboratory homogenization should be conducted Considerations
even if samples were homogenized in the field. In addition, after dry-sieving a sample
all material must be removed form the sieve.
The total amount of fine-grained material used for pipet analysis should be 5-25
grams. If more material is used, particles may interfere with each other during settling
and the possibility of flocculation may be enhanced. If less material is used, the
experimental error in weighing becomes large relative to the sample size. Once the
pipet analysis begins, the settling cylinders must not be disturbed, as this will alter
particle settling velocities.
It is recommended that triplicate analyses be conducted on one of every 20 samples, or
on one sample per batch if less than 20 samples are analyzed. It is recommended that
the analytical balance, drying oven, and temperature bath be inspected daily and
calibrated at least once per week.
Statistical strategies may mitigate the high costs of collecting sufficient quantities of
sediment. See also Statistical Design Considerations: Composite Sampling, Power
Analysis, and Power Cost Analysis (Appendix B Introduction; Section B.3).
B3.5 Statistical
Design
Considerations
Sediment grain size provides evidence essential in the evaluation of spatial and tempo- B3.6 Use Of Data
ral effects of anthropogenic and natural disturbance.
Grain size information may explain the temporal and spatial variability in biological
assemblages. Because many infaunal organisms are sensitive to changes in grain size
(e.g., affects their ability to burrow, to construct tubes), changes in benthic community
structure may be explained by simply examining grain size characteristics.
The detection of a change in sediment grain size composition may provide an early
warning of a potential threat to sensitive resources, such as a threat to commercial
clam beds. For many commercially and recreationally important clams, the ability to
quickly burrow, and to escape predation, is dependent upon the sediment grain size
composition of the bottom. Thus, changes of sediment grain size may be a good
indicator of impending loss of important resources.
However, the singular use of sediment grain size to assess the condition of the benthic
habitat or to guide the decision-making process is not recommended. The data gar-
-------
^ B-53
\ INJVA, Af - V* 0*- S x-> * >". ... ^ A. . \ \ ^ x " v.
nered from monitoring of the sediment's physical characteristics should be used in
conjunction with that collected from chemical and biological monitoring. Subsequent
analyses of health risks to human populations, as well as ecological risks to benthic
individuals, populations, and communities may be evaluated.
B3.7 Rationale
Summary and
Recommendations The objective of monitoring the sediment's physical characteristics is to
detect and describe spatial and temporal changes in sediment grain size
Results may be used to monitor rates of recovery following environmental
interventions, to evaluate the condition of benthic habitats, and to assist in
providing early warnings of potential impacts to the estuarine ecosystem
Monitoring Design Considerations
It is recommended that consistent types of sampling gear, and location and
timing of sample collection be implemented to allow for comparisons
among studies
Collection of undisturbed sediment requires that the sampler:
- create a minimal bow wake when descending
- form a leakproof seal when the sediment sample is taken
- prevent winnowing and excessive sample disturbance when
ascending
- allow easy access to the sample surface in order that undisturbed
subsamples may be taken
Analysis of the upper 2 cm is advised in order to examine recent deposi-
tional events
Penetration well below the desired sampling depth is preferred to prevent
sample disturbance as the device closes.
Existing A nalytical Methods
Because true and apparent distributions differ, detailed comparisons
between samples analyzed by these different methods are questionable
-------
B-54
It is therefore desirable that all samples within each study and among
different studies be analyzed using one of these two methods
QAIQC Considerations
Laboratory homogenization should be conducted even if samples were
homogenized in the field
It is recommended that the analytical balance, drying oven, and tempera-
ture bath be inspected daily and calibrated at least once per week.
Statistical Design Considerations
Compositing sediment sampling consists of mixing samples from two or
more replications collected at a particular location and time period
Space- (combining composites from several locations) and/or time- (com-
bining several composites over time from one location) bulking strategies
should be used judiciously since information concerning spatial and
temporal heterogeneity may be lost
Power analyses have shown that for a fixed level of sampling effort, a
monitoring program's power is generally increased by collecting more
replicates at fewer locations
Use of Data
Grain size information is may explain the temporal and spatial variability
in biological assemblages
The singular use of sediment grain size to assess the condition of the
benthic habitat or to guide the decision-making process in not recom-
mended; data garnered from monitoring of the sediment's physical charac-
teristics should be used in conjunction with that collected from chemical
and biological monitoring
-------
B-55
B3.8 Fredette, T.J., D.A. Nelson, T. Miller-Way, J.A. Adair, V.A. Setter, J.E. Clausner,
Literature Cited E.B. Hands, and F.J. Anders. 1989. Selected tools and techniques for physical and
and References biological monitoring of aquatic dredged material disposal sites. Final Report. U.S.
Army Engineer Waterways Experiment Station, Vicksburg, MS.
Mclntyre, A.D., J.M. Elliot, and D.V. Ellis. 1984. Introduction: design of sampling
programs. IBP Handbook No. 16. In: Methods for the Study of Marine Benthos.
(N.A. Holme and A.D. Mclntyre, eds.). Oxford: Blackwell Scientific Publications.
pp. 1-26.
Plumb, R.H. 1981. Procedure for handling and chemical analysis of sediment and
water samples. Technical Report EPA/CE-81-1. U.S. EPA and Corps of Engineers,
U.S. Army Engineers Waterways Experimental Station, Vicksburg, MS.
U.S. EPA. 1987a. Puget Sound protocols. Final Report. 31 pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1987b. Bioaccumulation monitoring guidance: Strategies for sample
replication and compositing, vol.5. EPA 430/9-87-003. Office of Marine and Estua-
rine Protection, Washington, DC. 51 pp.
U.S. EPA. 1987c. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
-------
B-56
-------
B-57
B4.1
Rationale
B4.0 Sediment Chemistry
The sediments represent the ultimate sink for many chemical contaminants in the
estuarine environment (U.S. EPA, 1989). The objective of monitoring bulk sediment
chemistry is to detect and describe spatial and temporal changes of these sediment
pollutants. The results may be used to monitor rates of recovery following environ-
mental interventions, to evaluate the condition of benthic habitats, and to provide early
warnings of potential impacts to the estuarine ecosystem.
Monitoring of pollutant levels in sediments is a widely-accepted means of measuring
the condition of the benthic habitat and is a powerful tool in the evaluation of spatial
and temporal effects of anthropogenic and natural disturbance. The singular use of
sediment pollutant loading data to assess the condition of the benthic habitat or to
guide the decision-making process is not recommended since other factors, such as
water quality and sediment grain size, can also affect habitat quality.
B4.2
Monitoring Design
Considerations
Sediment Sampling Devices
The protocols required to collect an acceptable surficial sediment sample for subse-
quent measurement of chemical variables have generally been neglected. In fact,
sampling crews have been given wide latitude in determining how samples are col-
lected. However, because sample collection protocols influence all subsequent labora-
tory and data analysis, it is key that sediment samples be collected using acceptable
and standardized techniques.
Collection of undisturbed sediment requires that the sampler:
create a minimal bow wake when descending
form a leakproof seal when the sediment sample is taken
prevent winnowing and excessive sample disturbance when ascending
allow easy access to the sample surface in order that undisturbed
subsamples may be taken
Penetration well below the desired sampling depth is preferred to prevent sample
disturbance as the device closes. It is optimal to use a sampler that has the means of
weight adjustment in order that penetration depths may be modified.
-------
B-58
Several types of devices can be used to collect sediment samples: dredges, grabs, and
box corers (Mclntyre et al., 1984). Many of these devices sample the benthic habitat
in a unique manner. Accordingly, conducting comparisons among data collected using
different devices is inadvisable.
Grab samplers - Grabs are capable of consistent sampling of bottom habitats. De-
pending on the size of the device, areas 0.02 to 0.5 m2 and depths ranging from 5 to
15 cm may be sampled. Limitations of grab samplers include:
variability among samples in penetration depth depending on sediment
properties
oblique angles of penetration which result in varying penetration depths
within a sample
the sample is inevitably folded resulting in the inability to section the
sample and the loss of information concerning the vertical structure in the
sediments
However, the careful use of these devices will provide quantitative data. Grab sam-
plers are the tools of choice for a number of estuarine and marine monitoring programs
due to their ability to provide quantitative data at a relatively low cost (Fredette et al.,
1989; U.S. EPA, 1987a).
Core samplers - Box corers utilize a surrounding frame to ensure vertical entry;
vertical sectioning of the sample is possible (U.S. EPA, 1987a). These devices are
capable of maximum penetration depths of 15 cm and may collect volumes 5 to 10
times that of grab samplers. Limitations of box corers include:
its large size and weight require the use of cranes or winches and a large
vessel for deployment
higher construction expenses
lack of calibration studies to permit comparisons to grab samples
The Hessler-Sandia box corer uses dividers to section the core into subsections,
facilitating subsampling of the core. Box corers are recognized as the tools of choice
for maximum accuracy and precision when sampling soft bottom habitats.
-------
B-59
Recommendations- Collection of sediments and benthic organisms should be done
concurrently in order to mitigate the costs of field sampling and to permit sound
correlation, regression, and multivariate analyses (see Benthic Community Structure).
Therefore, it is recommended that the sampling device also be suitable for benthic
sampling. Grab and core sampling devices permit adequate sampling of both sediment
and benthic infaunal communities with the one sampling device.
Sample Depth
It is recommended that the upper 2 cm of the sediment column be examined to charac-
terize surficial sediments. Although the 2 cm specification is arbitrary, it will ensure
that:
relatively recent sediments are sampled
adequate volumes of sediments are readily obtained for laboratory analyses
data from different studies may be compared
Sampling of depths other than 2 cm or vertical stratification of deeper sediment cores
may be appropriate, depending upon the objectives of the monitoring program. For
example, if information concerning only the most recent sediment contamination is
required, examination of the upper 1 cm may be appropriate. Stratification of deeper
cores will provide historical data of sediment contaminant levels and depositional
events. Comparison of data from studies analyzing different sediment depths is not
advised. Penetration well below the desired sampling depth is preferred to prevent
sample disturbance as the device closes.
Total Organic Carbon and Acid Volatile Sulfides Normalization
Total organic carbon (TOC) and acid volatile sulfides (AVS) are considered by many
to be the most important parameters in defining organic and metal concentrations in
sediments. Toxic hydrophobic contaminant concentrations have been found to be
related to the TOC content of the sediment (Karickhoff et al., 1979). Likewise, toxic
metals concentrations have been found to be related to the AVS concentration of the
sediment (DiToro et al., in press). The AVS pool is available to bind metals, thereby
reducing their bioavailability.
TOC and AVS normalizations have been conducted in order to estimate the
bioavailability of inorganic and organic contaminants. Furthermore, normalization of
-------
B-60
sediment contaminant concentrations to TOC and AVS appear to account for some of
the variability found in bioaccumulation rates and biological community structures.
TOC/lipid normalized accumulation factors (AF) have also been used to predict tissue
residue concentrations (Ferraro et al., 1990; Lake et al., 1987). AVS promises to assist
in the development of similar accumulation factors for metals.
These normalizations assume the following:
organic contaminants partition predominantly to sediment organic carbon;
metal contaminants partition predominantly to sediment AVS
rapid steady-state kinetics of contaminants
Development of standardized methods for measuring sediment TOC and AVS in the
laboratory are required before normalized contaminant concentrations may be com-
pared among studies.
The decision to normalize for TOC and AVS will depend upon monitoring program
objectives. For example, if the objective is to identify the "footprint" of a discharge,
normalization may not be appropriate. However, if the monitoring objective is to
identify "hot spots" - i.e. those areas where bioavailable contaminants represent a risk
to human and ecological health - normalization may be justified.
Selection of Sampling Period
Sediment contaminant levels display temporal patterns, commonly seasonal variabil-
ity. These changes are associated with seasonal patterns of benthic turbulent mixing
and sediment transport phenomena. The frequency of sampling should be related to
the expected rate of change in sediment contaminant concentrations. A spatially and
temporally consistent sampling design (i.e., time of sampling event and sampling
locations do not change over time) is recommended in order that comparisons between
areas over time may be conducted.
If seasonal variations are exhibited, it is recommended that direct comparisons be-
tween samples collected during different seasons be avoided. Studies investigating
interannual variation in the concentrations of sediment contaminants should be con-
ducted during the same season (preferably the same month) each year. Furthermore, it
is recommended that sediments be sampled when contaminant concentrations are
expected to be at their highest level in order to evaluate worst-case scenarios.
-------
B-61
B4.3 Questions to be considered during the choice of an appropriate analytical method
Existing include the parameters of interest, desired detection limits, sample size requirements or
Analyti cal restrictions, methods of preservation, technical and practical holding times, and matrix
Methods interferences. Several U.S. EPA documents (e.g., 1986a and 1987a) discuss the
common analytical problems encountered during monitoring analyses of sediment
samples.
Chemical Residue Analyses
Several factors determine achievable detection limits for a specific contaminant,
regardless of analytical procedure - a list of analytical procedures and U.S. EPA
method numbers is given in Table B4-1. These factors include:
sample size; 50-100 g (wet weight) with a minimum final dilution volume
of 0.5 ml is considered adequate (U.S. EPA, 1987a)
presence of interfering substances
IllllliIIll!ii;งiII!lllll
IlllilSrtiell
||||^&^aj^||iil;i
lllaiiiiiiiiiiiEfiiPi
l^lyiS^
i||R|||!||<:;|ll
"""'"
lg|l|||?i|||||||ji
l!l||iii;;illlill(i
-------
B-62
range of pollutants to be analyzed - an optimal method may be developed
without regard to potential effects on other parameters
level of confirmation - qualitative (e.g. presence or absence) or quantita-
tive (e.g. residue concentrations) analyses
level of pollutant found in the field and in analytical blanks
Selection of appropriate methods ought to be based on a trade-off between full-scan
analyses, which are economical but cannot provide optimal sensitivity for some
compounds, and alternate methods that are more sensitive for specific compounds but
can result in higher analytical costs.
Metals and Metalloids - Digestion methods for sediment samples are reviewed by
Plumb (1981). The U.S. EPA Contract Laboratory Program (CLP) requires the use of
HNO3:H2O2 (U.S. EPA, 1985).
Trace element analyses by ICP (US EPA method 6010) allows for several elements to
be measured simultaneously. Detection limits of ICP for most metals are generally
comparable to those achieved by GFAAS; however, detection limits for several metals
are significantly lower using AAS: arsenic, selenium, and mercury.
The combination of atomic absorption spectrophotometry techniques (AAS, U.S. EPA
methods 7000 series) and inductively coupled plasma emission spectrometry (ICP) is
the recommended analytical method for detection of metals and metalloids since no
technique is best for all elements (U.S. EPA, 1986a). Cold vapor AAS (U.S. EPA
method 7470) analysis is the only recommended technique for mercury (U.S. EPA,
1986a).
Graphite furnace AAS (GFAAS) is more sensitive (i.e. lower detection limits) than
flame AAS, but is more subject to matrix and spectral influences - GFAAS requires
particular caution with regard to laboratory contamination. GFAAS requires more
skilled laboratory technicians. Both AAS methods require that the concentration of
each element be determined by a separate analysis, making the analysis of a large
number of contaminant metals both labor-intensive and relatively expensive compared
to ICP. However, AAS methods may be cost-effective for the analysis of a few metals
over a large number of samples.
Semi-Volatile Organic Compounds - Analysis of semi-volatile organic compounds
involves a solvent extraction of the sample, cleanup of the characteristically complex
-------
B-63
' - \ , *?
* ^ ". ' $
extract, GC analysis, and quantification (U.S. EPA, 1987b and 1986a). There are two
GC/MS options for detecting extractable organic compounds: internal standard
technique and isotope dilution. The isotope dilution technique, which requires spiking
the sample with a mixture of stable isotope labeled analogs of the analytes, is recom-
mended because reliable recovery corrections can be made for each analyte with a
labeled analog or a chemically similar analog (U.S. EPA, 1986a). The isotope dilution
method is more expensive and less widely employed than the internal standard tech-
nique, which is the current method of choice in the Contract Laboratory Program
(CLP). Mass spectrometry provides positive compound identification by comparison
of both retention time and spectral patterns with standard compounds.
The identification of pesticides and PCB cogeners can be made by GC/ECD analysis
(U.S. EPA method 8080). GC/ECD provides greater sensitivity relative to GC/MS,
however GC/ECD does not provide positive compound identification. Confirmation
of pesticides and PCBs on an alternative GC/ECD or preferably by GC/MS, when
sufficient concentrations occur, is recommended for reliable results (U.S. EPA,
1986a). All other organic compound groups are recommended for analysis by GC/MS
(U.S. EPA, 1986a).
Volatile Organic Compounds - Analysis of volatile organic compounds also in-
volves a solvent extraction of the sample, cleanup of the characteristically complex
extract, GC analysis, and quantification (U.S. EPA, 1986a and 1987b). The purge and
trap GC/MS technique is employed for detecting volatile organic compounds in water.
GC/ECD may be used to achieve lower detection limits, although identification of the
compounds may not be as accurate.
A successful variation for detection of volatile organic residues in sediments involves
a device that vaporizes volatile organic compounds from the sediment sample under
vacuum and then condenses the volatiles in a super-cooled trap (Hiatt, 1981). The trap
is then transferred to a purge and trap device where it is treated as a water sample. The
isotope dilution option is recommended as it provides reliable recovery data for each
analyte (U.S. EPA, 1986a). However, this technique costs additional analytical time
and expense.
BAA Sample Collection
QA/QC
Considerations Sampling Gear - The primary criterion for an adequate sampler is that it consistently
collect undisturbed and uncontaminated samples. Benthic sampling devices should be
inspected for wear and tear leading to sample leakage upon ascent. It is prudent to
-------
B-64
have a backup sampler onboard the survey vessel in case the primary sampler is found
to be unsuitable during the cruise.
In the field, sources of contamination include sampling gear, lubricants and oils,
engine exhaust, airborne dust, and ice used for cooling samples. Samples designated
for chemical analyses require all sampling equipment - e.g., siphon hoses, scoops,
containers - be made of noncontaminating material and be cleaned appropriately prior
to use. Potential airborne contamination - e.g., stack gases, cigarette smoke - should
be avoided. Furthermore, samples and sampling containers should not be touched with
ungloved fingers.
Sample Condition - Benthic samples should satisfy the following sample acceptabil-
ity criteria (U.S. EPA. 1987):
sampler is not over-filled with sample so that the sediment surface is
pressed against the top of the sampler
overlying water is present, indicating minimal leakage
the overlying water is not excessively turbid indicating minimal sample
disturbance
the sediment surface is relatively flat indicating minimal disturbance or
winnowing (Figure B4-1)
the desired penetration depth is achieved
If the sample does not meet all the criteria, the sample should be rejected. Likewise,
water column samples from leaking bottles should not be included in the analyses.
Sample Handling
For analyses of metals, samples should be frozen and kept at -20ฐC. Although specific
holding times have not been recommended by U.S. EPA, a maximum of 6 months (28
days for mercury) would be consistent with that for water samples (Tables B4-2 and
B4-3).
For analyses of volatile compounds, samples should be stored in the dark at 4ฐC (U.S.
EPA, 1987a). Analyses of volatile compounds should be performed within 14 days of
-------
B-65
Figure B4-1.
Examples of acceptable
and unacceptable samples
(US. EPA, 1987b).
Acceptable if Minimum
Penetration Requirement Met
and Overlying Water is Present
Unacceptable
(Canted with Partial Sample)
Unacceptable
(Washed, Rock Caught in Jaws)
Unacceptable
(Washed)
collection as recommended by U.S. EPA (1984). If analyses of semivolatile com-
pounds will not be performed within the recommended 7-day holding time, freezing of
the samples at -20ฐC is advised. Holding times for frozen samples has not been
established by the U.S. EPA. A general guideline of a maximum of 6 months would
be consistent with that for water samples (U.S. EPA, 1987a).
Resectioning sediment samples for chemical analyses should only be conducted with
noncontaminating tools under "clean room" conditions.
Laboratory Analyses
Laboratory performance as measured by method QA/QC protocols should be used to
evaluate and select appropriate analytical methods. Changes to laboratory protocols
should only be considered if new protocols meet established performance criteria.
Suggested QA/QC protocols and analytical performance criteria are discussed below.
QA/QC reports should describe the results of quantitative QA/QC analyses, as well as
other elements critical to the laboratory analyses of water, sediments, and tissues to
-------
B-66
. v- ^ ~f. : :
TABLE B4-2,
SUMMARY OF SAMPLE COLLECTION AND PREPARATION
QA/QC REQUIREMENTS FOR ORGANIC COMPOUNDS
Variable Sample Size3 Container'5 Preservation
Maximum
Holding Time
Semivotatiles 50*100g
Volaliles 40 ml
I^reexe
Cool, 4ฐ
lyear0
14 days
a Recommended field sample sizes for one laboratory analysis. If additional
laboratory analyses are required (i.e., replicates). The field sample size should be
adjusted accordingly.
b C = Glass.
c This is a suggested holding time. No U.S. EPA criteria exist for the preservation
of this variable.
d No headspace or airpockels should remain,
c Freezing these samples will likely cause breakage of the sample container, because
no airspace for expansion is provided.
ensure proper interpretation of the results. It is recommended that these reports be
recorded and stored in a database for future reference.
Field QA/QC Checks - Transfer blanks will whether any contamination was intro-
duced of reagents in the field or introduced during shipping of samples. Cross-
contamination blank is designed to verify the absence of contamination carried over
from one sample to another due to inadequate cleaning of field equipment.
Blind replicates, splits treated and identified as separate samples, may be sent to the
same laboratory for analysis or have one sample sent to a "reference" laboratory for
comparison. In addition, standard reference material may also be placed in a sample
container at the time of collection and sent "blind" to the laboratory.
Instrument QA/QC Checks - Calibration standards should be analyzed at die begin-
ning of sample analysis, and should be verified at the end of each 12-hr shift during
-------
B-67
Table B4-3.
SAMPLE PRESERVATION AND STORAGE PARAMETERS
Sample
Storage
Analyte
Container9 Size
Preservative Lifetime
All metals
Elutriate Studyd
Fractionation Study*1
P,G
P,G
50 gb
12 L
3L
1L
Freeze
4'C
4'C
4'C
a P = linear polyethylene, G = borosilicaie glass, TFE = tetrafluoroethylene.
b Wet weight
c Suggested. No EPA criteria exist Hg holding time 28 days.
4 Plumb, 1981. Storage time "as short as possible," analyses to be completed "within 1
week of sample collection."
which analyses are performed (U.S. EPA, 1987b). The concentrations of calibration
standards should bracket the expected sample concentrations, or sample dilutions or
sample handling modifications (i.e., reduced sample size) will be required.
Method QA/QC Checks- Analysis of preparation blanks should be conducted to
demonstrate the absence of contamination from sampling or sample handling in the
laboratory. At least one method blank must be included with each batch of samples
and should constitute at least 5% of all samples analyzed.
Spike recovery analyses are required to assess method performance for the particular
sample matrix. Spike recoveries serve as an indication of analytical accuracy, whereas
analysis of standard reference materials measure extraction efficiency. Recommended
control limits include 75-125 percent recovery for spikes, and 80-120 percent recovery
for the analysis of standard reference materials (SRM).
-------
B-68
Replicates must be analyzed to monitor the precision of laboratory analyses. A
minimum of 5% of the analyses should be laboratory replicates. The control limits are
ฑ20 percent relative percent difference for duplicates. Triplicates should be analyzed
on one of every 20 samples or on one sample per batch if less than 20 samples are
analyzed.
Tables B4-4 and B4-5 provide a brief summary of QA/QC for laboratory analyses.
For information concerning detection limits, see Water Column Chemistry QA/QC
Considerations (Section B2.2).
ijj^m^MynKS
S Standard reference
iipjfe^
; Matrix spikes
lii^^jji^^^^^^^^^^^^
^^^^ii^s^^^^jiiiiliA^^^j^^^^^^^^^^^.
"^^^^^^^^j^P^^^^^^^^^^^^^^^^^^
^^^^^j^^^tj^^^^^^^^^^^^^^^^^^
-------
B-69
B4.5 Statistical strategies may mitigate the high costs of collecting sufficient quantities of
Statistical Design sediment. See statistical design considerations: Composite sampling, power analyses,
Considerations and power-cost analysis (Appendix B Introduction).
B4.6 The results may be used to monitor rates of recovery following environmental inter-
Use Of Data ventions, to evaluate the condition of benthic habitats, and to provide early warnings of
potential impacts to the estuarine ecosystem. The singular use of sediment pollutant
loading data to assess the condition of the benthic habitat or to guide the decision-
making process is not recommended. Data acquired from monitoring sediment
TABLE B4-5, SUMMARY OF WARNING
AND CONTROL LIMITS FOR QUALITY CONTROL SAMPLE
Analysis Type
Recommended
Warning Limit
Recommended
Control Limit
Surrogate Spikes
Method Blank
Phthalate,
Acetone
Other Organic
Compounds
Standard
Reference Materials
10 percent recovery
30 percent of the analyte
1 jig total or 5 percent
oftheanalyte
i
95 percent
confidence interval
Matrix spikes (50-65 percent recovery)
Spiked Method Blanks (5045 percent recovery)
Analytical Replicates
Held Replicates -
Ongoing Calibration
50 percent recovery
5 ug total or 50 percent
of the analyte
2.5 }ig total or 5 percent
of the analyse
95 percent confidence
interval for Certified
Reference Material
(50 percent recovery)
(50 percent recovery)
ฑ100 percent
coefficient of variation
25 percent of
initial calibration
-------
B-70
contaminant levels, in conjunction with the sediment's physical properties, may be
used to assess the bioavailability of these pollutants (SCCWRP, 1986). Subsequent
analysis of this information, in accord with biological data, may be used to assess the
impacts and risks of sediment pollutants to human populations, benthic populations/
communities, and the estuarine ecosystem.
Monitoring of pollutant levels in sediments is a widely-accepted means of measuring
the condition of the benthic habitat and is a powerful tool in the evaluation of spatial
and temporal effects of anthropogenic and natural disturbance.
Rationale B4.7
Summary and
The sediments represent the ultimate sink for many chemical contaminants Recommendations
infiltrating the estuarine environment (U.S. EPA, 1989)
The objective of monitoring bulk sediment chemistry is to detect and
describe spatial and temporal changes of these sediment pollutants
Monitoring Design Considerations
It is recommended that consistent types of sampling gear, and location and
timing of sample collection be implemented to allow for comparisons
among studies
Collection of undisturbed sediment requires that the sampler:
- create a minimal bow wake when descending
- form a leakproof seal when the sediment sample is taken
- prevent winnowing and excessive sample disturbance when
ascending
- allow easy access to the sample surface in order that undisturbed
subsamples may be taken
Analysis of the upper 2 cm is advised in order to examine recent sediment
contamination events
Penetration well below the desired sampling depth is preferred to prevent
sample disturbance as the device closes.
-------
B-71
Total organic carbon and acid volatile sulfide normalization is recommend
to allow comparisons of chemical residue concentrations between locations
- normalizing the data will depend upon the objectives of the individual
monitoring program
It is recommended that sediments be sampled when contamination concen-
trations are expected to be at their highest level
Existing Analytical Methods
It is recommended that consistent types of analytical protocols be imple-
mented to allow for comparisons among studies
Metals/Metalloids
- The combination of GFAA and ICP are the recommended
methods for the detection of metals and metalloids (U.S. EPA,
1987b)
- cold vapor AAS is the recommended protocol for mercury
detection
Organics
- GC/MS in conjunction with isotope dilution is recommended for
the detection of semi-volatile organic compounds
- vacuum super-cooled trap in conjunction with a purge and trap
device is recommended for the detection of volatile organics
(Hiatt, 1981)
- isotope dilution option is recommended as it provides reliable
recovery data for each analyte (U.S. EPA, 1986a)
QA/QC Considerations
Blank, spike recovery, and replicate analyses are recommended quality
control checks
Reports delineating the essential elements of the bioaccumulation compo-
nent of the program should be included with the quantitative QA/QC
analyses
-------
B-72
Statistical Design Considerations
Compositing sediment sampling consists of mixing samples from two or
more replications collected at a particular location and time period
Space- (combining composites from several locations) and/or time- (com-
bining several composites over time from one location) bulking strategies
should be used judiciously since information concerning spatial and
temporal heterogeneity may be lost
Use of Data
The data garnered from monitoring of the pollutant levels in estuarine
sediments should be used in conjunction with that collected from other
physical and biological monitoring to monitor rates of recovery following
environmental interventions, to evaluate the condition of benthic habitats,
and to provide early warnings of potential impacts to the estuarine ecosys-
tem
The singular use of sediment pollutant loading data to assess the condition
of the benthic habitat or to guide the decision-making process in not
recommended
DiToro, D.M., J.D. Mahony, D.J. Hansen, K.J. Scott, A.R. Carlson, and G.T. Ankley. B4.8
In Press. Acid volatile sulfide predicts the acute toxicity of cadmium and nickel in Literature Cited
sediments. and References
DiToro, D.M., J.D. Mahony, D.J. Hansen, K.J. Scott, M.B. Hicks, S.M. Mayr, and
M.S. Redmond. In Press. Toxicity of cadmium in sediments: the role of acid volatile
sulfide.
Ferraro, S.P., H. Lee, R.J. Ozretich, and D.T. Specht. 1990. Predicting bioaccumula-
tion potential: A test of a fugacity-based model. Arch. Environ. Contam. Toxicol.
19:386-394.
Fredette, T.J., D.A. Nelson, T. Miller-Way, J.A. Adair, V.A. Sotler, J.E. Clausner,
E.B. Hands, and F.J. Anders. 1989. Selected tools and techniques for physical and
biological monitoring of aquatic dredged material disposal sites. Final Report. U.S.
Army Engineer Waterways Experiment Station, Vicksburg, MS.
-------
B-73
3sซ:r'*?
Hiatt, M.H. 1981. Analysis of fish and sediment for volatile priority pollutants. Anal.
Chem. 53:1541-1543.
Karickhoff, S.W., D.S. Brown, and T.A. Scott. 1979. Sorptionofhydrophobic
pollutants on natural sediments. Wat. Res. 13:241-248.
Knezovich, J.P. and F.L. Harrison. 1987. A new method for determining the concen-
tration of volatile organic compounds in sediment interstitial water. Bull.
Lake, J.L., N.I. Rubinstein, and S. Parvignano. 1987. Predicting bioaccumulation:
Development of a partitioning model for use as a screen tool in regulating ocean
disposal of wastes. In: Fate and Effects of Sediment-bound Chemicals In Aquatic
Systems. (Dickson, K.L., A.W. Maki, and W.A. Brungs, eds). Florissant, CO: Sixth
Pellston Workshop.
Landrum, P.P. and J.A. Robbins. In Press. Bioavailability of sediment-associated
contaminants to benthic invertebrates. In: Sediments: Chemistry and Toxicity of In-
Place Pollutants (Giesy, J.P., R. Baudo, and H. Muntau, eds). Lewis Publishers.
Mclntyre, A.D., J.M. Elliot, and D.V. Ellis. 1984. Introduction: design of sampling
programs. IBP Handbook No. 16. In: Methods for the Study of Marine Benthos.
(N.A. Holme and A.D. Mclntyre, eds.). Oxford: Blackwell Scientific Publications.
pp. 1-26.
Plumb, R.H. 1981. Procedure for handling and chemical analysis of sediment and
water samples. Technical Report EPA/CE-81-1. U.S. EPA and Corps of Engineers,
U.S. Army Engineers Waterways Experimental Station, Vicksburg, MS.
SCCWRP. 1986. Polynuclear aromatic hydrocarbon contamination in sediments form
coastal waters of southern California. Southern California Coastal Water Research
Program, pp. 13-16.
U.S. EPA. 1984. US EPA contract laboratory program statement of work for organics
analysis, multi-media, multi-concentration. IFB WA 85-T176, T177, T178. Washing-
ton, DC.
U.S. EPA. 1985. Contract laboratory program statement of work, inorganic analysis,
multi-media, multi-concentration. SOW No. 785. U.S. EPA, Washington, DC.
-------
B-74
U.S. EPA. 1986a. Analytical methods for U.S. EPA priority pollutants and 301 (h)
pesticides in estuarine and marine sediments. Prepared for the Office of Marine and
Estuarine Protection, Washington, DC.
U.S. EPA. 19865. Test methods for evaluating solid wastes, physical/chemical
methods. SW-846, 3rd Edition. Environmental Protection Agency, Washington, DC.
U.S. EPA. 1987a. Puget Sound protocols. Final Report. 31 pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1987b. Quality Assurance/Quality Control (QA/QC) for 301(h) Monitor-
ing Programs: Guidance on field and laboratory methods. EPA 430/9-86-004. Office
of Marine and Estuarine Protection, Washington, DC.
U.S. EPA. 1987c. Bioaccumulation monitoring guidance: Strategies for sample
replication and compositing, vol. 5. EPA 430/9-87-003. Office of Marine and Estua-
rine Protection, Washington, DC. 51 pp.
U.S. EPA. 1987d. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1989. Sediment classification methods compendium. Prepared for the
Office of Water Regulations and Standards.
-------
B-75
B5.0 Plankton: Blomass, Productivity and
Community Structure/Function
B5.1 Rationale
Although increased primary production resulting from intentional nutrient inputs has
been shown to increase fish stocks in some experimental systems, the possible increase
in fisheries in naturally productive systems is generally considered insignificant when
compared to the deleterious effects of eutrophication. The most conspicuous and
potentially detrimental effects of eutrophication to estuarine environments are in-
creased turbidity, phytotoxins, and creation of hypoxic or anoxic conditions. Further-
more, changes in nutrient concentrations in an estuary may result in the potential for
long-term biological changes in the estuarine community structure and function (e.g.,
changes in species distribution and abundance of both primary producers and consum-
ers).
If eutrophic conditions are suspected of occurring, monitoring of the plankton commu-
nity may assist in detecting and defining the spatial and temporal scales of these
conditions. The objectives of monitoring plankton characteristics are to detect spatial
and temporal changes in plankton community structure and function. These results
may be used to assess the consequences of spatial and temporal variability in plankton
community structure and function to higher trophic levels which depend upon plank-
tonic resources. In addition, this information may be used to further the understanding
of the relationship between water quality conditions and planktonic community
structure and function.
B5.2 Monitoring Sampling Methods
Design
Considerations Phytoplankton - Plankton samples are frequently collected using water bottles. They
are simple devices, usually consisting of a cyndrical tube with stoppers at each end and
a closing device that is activated by an electrical signal. Each samples a discrete
parcel of water at any designated depth. Multiple samplers are fixed on a rosette frame
in order that several depths may be sampled during one cast and/or that replicate
samples may be collected at a particular depth. The most commonly used bottle
samplers include the Kemmerer, Van Dorn, Niskin, and Nansen samplers (U.S. EPA,
1987a).
Phytoplankton samples should be collected at a variety of depths throughout the water
column some above and some below the pycnocline. A minimum of four depths in the
vertical profile should be sampled: 1) one meter below the surface, 2) one meter
above the bottom, 3) one meter above the pycnocline, and 4) one meter below the
-------
B-76
pycnocline. If the waters are too shallow or no stratification occurs, it would be
appropriate to take the later two samples at evenly spaced distances between the top
and bottom samples. For example, composite water column samples collected for the
Virginia Chesapeake Bay Plankton Monitoring Program are taken at five different
depths above and five different depths below the pycnocline (U.S. EPA, 1989).
Vertical measures of chlorophyll a. (an indirect measure of phytoplankton standing
stocks) can be made by fluorometric or spectrophotometric determination of chloro-
phyll a from these depth samples.
If available, a pump station may be used with a flow through fluorometer for a con-
tinuous profile of chlorophyll a concentration with depth (Lorenzen, 1966). Samples
to be used for taxonomy analysis should be collected with water bottles, as agitation
associated with pumping may damage cells making them unidentifiable. Pumps
should only be used for determination of chlorophyll a concentrations.
Zooplankton - Because zooplankton possess varying degrees of swimming ability,
they have the potential for aggregating in patches or in a narrow depth strata. This
introduces additional complications into quantitative sampling, such as avoidance of
certain types of gear.
The sampling methods to be used for collecting zooplankton will vary depending on
the size of the organisms. Microzooplankton (size range of 20-200um) can be col-
lected with water bottles at various depths similar to those used for phytoplankton, or
small (for example 44 um) mesh nets can be used (Jacobs and Grant, 1978). Pumping
systems can also be used which have the advantage of being able to take samples
integrated over depth and of collecting samples while the ship is underway (Beers
et al., 1967). However, pumps may damage soft-bodied organisms and they are more
expensive and complicated than water bottles. Triplicate samples should be collected
from each station depth, thus allowing for statistical analysis of intra-station variabil-
ity.
For small mesozooplankton (greater than 200 um) nets are generally used (UNESCO,
1968). Additional tows may have to be made with larger nets in order to collect
representative samples of larger zooplankton and larval fish. All tows should be
replicated. The number of replicates necessary for the desired precision of estimation
should be determined during a preliminary or pilot sampling program. A number of
other considerations, including net mouth diameter, towing speed and ship board
handling of samples will effect sampling results. Some problems associated with the
use of nets for zooplankton sampling include avoidance and clogging which may result
-------
B-77
in underestimating abundance and diversity and loss of filtration efficiency (McGowan
and Fraundorf, 1966; Wiebe and Holland, 1968).
Conduction-Temperature-Depth (CTD) Systems
Continuous vertical profiles of chlorophyll a concentrations can be obtained by
attaching a fluorometer to a conductivity, temperature and depth (CTD) system. An
additional sensor measuring dissolved oxygen could be attached to the CTD system.
This combined instrument package will make direct measurements of chlorophyll a
levels and the degree of eutrophication and water column stratification.
The advantage of this measurement system is that instead of pulling water up from
depth to a measuring device on the deck of a ship, the instrument is lowered through
the water column taking samples in undisturbed water. Sampling rates can be as high
as 24 samples per second with a locating speed of 1 meter per second. Data is usually
collected by a computer based data acquisition system and in advanced systems can be
displayed in real-time as it is being collected. A rosette water sampler can be lowered
at the same time as the CTD package to collect water samples at discrete depths.
These samples will be used for taxonomy analysis and to verify and calibrate chloro-
phyll a measurements from the CTD fluorometer. The calibration will be carried out
using laboratory spectrophotometer methods.
Selection of Temporal Sampling Strategies
The time of the year should be controlled or stratified in the design - the use of annual
averages is seldom good practice. Given the seasonal variation characteristic of
phytoplankton abundances, it is recommended that direct comparisons between
samples collected during different seasons be avoided. Studies investigating
interannual variation in the characteristics of phyto- and zooplankton communities
should be conducted during the same season (preferably the same month) each year.
Due to their short turnover times, phytoplankton communities are capable of respond-
ing to perturbations much more rapidly than other biotic groups. Therefore, phyto-
plankton samples should be collected relatively more frequently. In those situations
where phytoplankton communities display pronounced seasonal variations in standing
stock or production, it may be appropriate to use a temporally stratified sampling
approach. For example, in the Maryland Chesapeake Bay Phytoplankton Monitoring
Program, sampling takes place once monthly from October through March and twice a
month from April through September (U.S. EPA, 1989).
-------
B-78
The lifespan of zooplankton on the other hand is longer than phytoplankton, so the
capacity for responding to perturbations is less than that of phytoplankton. Therefore
less frequent sampling is required. As with the phytoplankton community, the zoop-
lankton monitoring program should consider the natural temporal fluctuations in
abundance and species composition.
In many cases, regular monitoring of the zooplankton community may not be neces-
sary unless changes in the phytoplankton community are observed which would induce
changes in the herbivore community. Because many zooplankton graze on phytoplank-
ton, in areas where the phytoplankton community has been affected, alterations of the
zooplankton community are a distinct possibility. In other cases, monitoring of
zooplankton may be desirable only when there is evidence of previous impact on the
zooplankton community, or in those situations where point and nonpoint source
discharges are located in areas where there is high potential impact on zooplankton.
For example, it may be useful to monitor zooplankton in estuarine environments with
macroplanktonic larvae of important commercial or recreational species.
Taxonomic Identification
Subsamples drawn from water collected in water sampling bottles should be preserved
for later microscopic analysis to determine phytoplankton community composition.
The choice of fixation will depend on the dominant types of phytoplankton known to
inhabit a given area (buffered formaldehyde and Lugol's solution are two common
fixatives). Preserved phytoplankton samples normally must be concentrated for
quantitative microscopic analysis.
The taxonomic analysis should include identification and enumeration of dominant
phyto- and zooplankton taxa and measures of population abundance whenever pos-
sible. Particular attention should be given to meroplankton larvae of commercially,
recreationally or ecologically important species.
Perturbations to the phyto- and zooplankton communities should be analyzed in
relation to other potential impacts on other biological communities which include:
potential primary and secondary impacts to higher trophic level communi-
ties (e.g., food web impacts)
occurrence of toxic or nuisance phytoplankton
B5.3 Existing
Analytical
Methods
-------
B-79
\ ^
V*** v
Some CCMPs may require that both larval and adult zooplankton community structure
and function should be analyzed in order to adequately assess impacts to high trophic
levels. Standardized methods are recommended to permit comparisons between
studies.
Biomass and Productivity
Methods for estimating phytoplankton biomass used in earlier monitoring surveys
include cell counts, total cell volume estimates, protein estimates, and dry weight.
These methods have certain disadvantages related to speed of the technique (e.g., time-
consuming) and the degree of accuracy for the estimation of phytoplankton biomass
(D'Elia et al., 1986). The use of chlorophyll a measurements, especially fluorometric,
has become widespread primarily because the method is relatively fast, simple and
reproducible.
Phytoplankton biomass can be indirectly measured through the measurement of the
concentration of chlorophyll & in the water. This is done through fluorometric or
spectrophotometric measurements. Within these methods there are also differences in
extraction techniques for chlorophyll determination, including various methods of
filtration, solvents, temperature and/or physical treatment (sonication or grinding)
(D'Elia et al., 1986).
However, with both the fluorometric and spectrophotometric determinations, the
presence of accessary pigments may interfere with determination of chlorophyll a-
Several researchers, however, have successfully used chromatographic procedures to
separate interfering substances prior to determination (D'Elia et al., 1986). The High
Performance Liquid Chromatography (HPLC) is generally acknowledged as the most
accurate (and most expensive) method for chlorophyll determinations (Zimmerman,
C.F., 1990, personal communication).
Phytoplankton primary productivity should be measured by the 14C light-dark bottle
technique (UNESCO, 1973).
Biological Indices
Biological indices allow large amounts of multivariate data to be reduced to a single
number. The numbers of individuals and the numbers of species have been found to
be good indicators of anthropogenic disturbance, as well as of other environmental
stresses. Furthermore, these simple biological indices are less ambiguous and are often
as informative as diversity indices (U.S. EPA, 1985; Green, 1979; Hurlbert, 1971).
-------
B-80
More complicated indices e.g. species diversity, species richness, dominance,
evenness have found varying degrees of acceptance. Diversity indices have the
following limitations (Green, 1984):
often lack biological meaning
are not robust empirical indicators of any important correlates of "estuarine
health"
do not incorporate information of form and function of resident species
are susceptible to biases associated with well-described taxa
However, species diversity indices are a widely used measure of community structure.
The dominance index is a measure of the degree to which one or a few species domi-
nate the community. The dominance index, herein defined as the minimum number of
species required to account for 75 percent of the total number of individuals, has been
useful in describing community structure (Swartz et al., 1985). It is easily calculated,
does not assume an underlying distribution of individuals among species, and is
statistically testable.
Indicator Species
Abundances of selected indicator species are used to evaluate response of the commu-
nity as a whole. Examination of abundances of individual indicator species are
generally informative and may reduce the cost of the analysis. The absence of pollu-
tion sensitive species and the enhancement of opportunistic and pollution tolerant
species may assist in defining the spatial and temporal extent and magnitude of
impacts. However, indicator variables must possess the following characteristics
(Green, 1984):
provide a sufficiently precise and accurate appraisals of:
- species of concern
- anthropogenic disturbances to planktonic communities
- presence/absence or the magnitude of anthropogenic perturbance
to the estuarine ecosystem
cost-effective and a statistically reliable alternative to monitoring all
critical planktonic community measures of habitat perturbance
-------
'<
B-81
appropriate for the spatial and temporal scale demanded by the study
objectives
It is recommended that multiple pollution-tolerant and pollution-sensitive indicator
species are selected by clearly defined criteria. Multiple indicator species often
provide a more complete representation of environmental conditions.
B5.4 QA/QC Variability in measurements caused by field heterogeneity is quantitatively determined
Considerations by the analysis of replicate field samples. Replicate sampling should be conducted at
all field stations where measurements are to be used in comparisons. Analysis of
replicate sample data is necessary for assessing the reliability of such comparisons.
Laboratory performance and calibration should be verified at the beginning and
periodically during the time analyses are performed. Commercially available chloro-
phyll is available and recommended for use in calibration. Chlorophyll quality control
samples are available from EPA's Environmental Monitoring and Support Laboratory
in Cincinnati, Ohio. Use of blind, split or other control samples can be used to evalu-
ate performance. The "Interim Guidance on Quality Assurance/Quality Control (QA/
QC) for the Estuarine Field and Laboratory Methods" (U.S. EPA, 1985) provides a
standard operating procedure for chlorophyll measurements.
B5.5 Statistical Consideration of statistical strategies will mitigate the high costs of collecting and
Design processing samples. Also see Statistical Design Considerations: Power Analysis and
Considerations Power Cost Analysis (Appendix B Introduction; Section B.3).
Temporal Stratification of the Data
Temporal stratification of the data should not be attempted until sufficient knowledge
of long-term natural cycles is attained. Initially, simple regression analyses may be
conducted on seasonally stratified data in order to identify monotonic temporal trends.
Further examinations of whether conditions are improving or degrading over time may
be examined using various statistical time series analyses (e.g., temporal
autocorrelation, spectral analyses, etc.)
Information derived from water quality monitoring will be important in interpreting
the results of plankton sampling. For example, phosphate concentrations may indicate
the cause of a plankton "bloom," while measures of dissolved oxygen levels may
describe the consequences of the "bloom" to other living estuarine resource. There-
-------
B-82
ป f^ ** ฃ :-..
.,"ซ,V%'J;,^t^. '
fore, the selection of water quality and plankton sampling criteria must not be done
independently. Also, alternations to the plankton community should be analyzed in
relation to other impacts on biological resources such as food web impacts on fish
communities.
Plankton monitoring strategies should be able to delineate between natural variability B5.6 Use Of Data
in plankton stocks and those caused by anthropogenic changes in nutrient concentra-
tions.
Characterization of phytoplankton taxonomic abundance and distribution and primary
productivity provide indications of water quality conditions. Monitoring changes in
phytoplankton population composition and densities are critical for the interpretation
and evaluation of long-term trends in water and habitat quality. Further understanding
of the causes of excessive water column and sediment oxygen demand requires
tracking of photosynthetic activity and metabolic rates overtime (Chesapeake Execu-
tive Council, 1988). Zooplankton abundance and distribution are affected both by
changes in phytoplankton and changes in predator populations. Therefore, population
characteristics of this group can indicate symptoms of water quality problems, fishing
pressure and other habitat problems for predator species (Chesapeake Executive
Council, 1988).
Rationale B5.7
Summary and
Track phytoplankton and herbivore populations if eutrophication is sus- Recommendations
pected
Purpose of monitoring is to assess the effectiveness of the CCMP in
mitigating the potential impacts caused by changes in the planktonic
community biomass and structure/function
Monitoring Design Considerations
Collect both phytoplankton and zooplankton samples at specific depths
throughout the water column
Phytoplankton samples taken with water bottles and pumps used for
chlorophyll concentrations
-------
B-83
Zooplankton sampling methods based on sized
Monitoring programs should consider natural and temporal fluctuations in
plankton biomass and species composition
Taxonomic analysis should include identification and counts of dominant
species
Other components of the overall monitoring program, including water
quality monitoring and fish and shellfish communities should be analyzed
relative to plankton community data to establish relationships and trends
Existing Analytical Methods
It is recommended that consistent types of sampling gear, and location and
timing of sample collection be implemented to allow for comparisons
among studies
Phytoplankton biomass can be determined using fluorometric or spectro-
photometric methods using a variety of filtration and extraction techniques
Phytoplankton productivity measured using the 14C light-dark bottle
technique
Selected biological indices should retain biological meaning, be robust
indicators of estuarine "health", and incorporate species form and function
Indicator species should possess the following characteristics:
- sensitive to planktonic perturbances of concern
- cost-effective and statistically reliable alternative to measuring all
species in a monitoring program
- statistically reliable indicative measures of habitat perturbance
- appropriate for the spatial and temporal scale dended by the study
objectives
Selection of reference sites is key to the evaluation of environmental
impact due to anthropogenic perturbances; several reference sites may be
required to proper control for sampling sites
-------
B-84
QA/QC Considerations
Replicate samples taken at all field stations where applicable
Laboratory performance evaluations and calibrations must be done on a
regularly scheduled basis
Standard chlorophyll samples, obtained from the EPA, should be used for
calibrations
Statistical Design Considerations
Power analyses may be applied to determine the appropriate number of
sample replicates required to detect a specified difference
Use of Data
Detect short-and long-term spatial and temporal trends in overall biomass
and productivity
Detect short- and long-term spatial and temporal trends in species abun-
dance, distribution and composition
Examine relationship between water quality conditions and trends in
plankton community characteristics
Examine relationship between plankton community characteristics and
impacts on other living resources (e.g., fish and shellfish communities)
Abaychi, J.K., and J.P. Riley. 1979. The determination of phytoplankton pigments
by high-performance liquid chromatography. Anal. Chim. Acta.
Ahlstrom, E. 1969. Recommended procedures for measuring the productivity of
plankton standing stock and related oceanic properties. Nat. Acad. Sci. 59 pp.
APHA. 1985. Standard methods for the examination of water and wastewater.
American Public Health Association, Washington, DC 1268 pp.
B5.8 Literature
Cited and
References
-------
ฃ>' ' *y
B-85
ASTM. 1979. Water. In: Annual Book of ASTM Standards. Part 31. Philadephia,
PA: Amer. Soc. Test. Mat.
Beers, J.R. 1978. Pump sampling, pp. 41-49. In: Monographs on oceanographic
methodology. 6. Phytoplankton manual. United Nations Educational, Scientific, and
Cultural Organization, Paris.
Beers, J.R., G.L. Stewart, and J.D.H. Strickland. 1967. A pumping system for sam-
pling small plankton. J. Fish. Res. Board Can. 24:1811-1818.
Boesch, D.F. 1977. Application of numerical classification in ecological investiga-
tions of water pollution. Rep. No. 600/3-77-033. U.S. EPA, Corvallis, OR. 115pp.
Brown, L.M., B.T. Margrave, and M.D. Mackinnon. 1981. Analysis of chlorophyll a.
in sediments by high-pressure liquid chromatography. Can. J. Fish. Aquat. Sci.
38:205-214.
Bumison, B.K. 1980. Modified dimethyl sulfoxide (DMSO) extraction for chloro-
phyll analysis of phytoplankton. Can. J. Fish. Aquat. Sci. 37:729-733.
Burrell, V.G., W. A. Van Engel, and S.G. Hummel. 1974. A new device for
subsampling plankton species. J. Cons. Perm. Int. Explor. Mer. 35:364-366.
Chesapeake Executive Council. 1988. Living Resources Monitoring Plan. Chesa-
peake Bay Program, Annapolis, MD. 94 pp.
Clifford, H.T., and W. Stephenson. 1975. An Introduction to Numerical Classifica-
tion. San Francisco, CA: Academic Press. 229 pp.
Cochran, W.G. 1977. Sampling Techniques. 3rd edition. New York, NY: John
Wiley & Sons, Inc. 428pp.
Cooley, W.W., and P.R. Lohnes. 1971. Multivariate Data Analysis. New York, NY:
John Wiley & Sons, Inc. 364 pp.
Grossman, J.S., R.L. Kaisler, and J. Cairns, Jr. 1974. The use of cluster analysis in the
assessment of spills of hazardous materials. American Midland Naturalist 92:94-114.
-------
B-86
D'Elia, C.F., K.L. Webb, D.V. Shaw, and C.W. Keefe. 1986. Methodological com-
parisons for nitrogen and chlorophyll determinations in estuarine water samples.
Submitted to the Power Plant Siting Program, Maryland Department of Natural
Resources, Annapolis, MD and Chesapeake Bay Liaison Office, U.S. Environmental
Protection Agency, Annapolis, MD.
Dunn, O.J. 1964. Multiple comparisons using rank sums. Technometric
6(3):241-252.
Garside, C, and J.P. Riley. 1969. A thin-layer chromatographic method for the
determination of plant pigments in sea water and cultures. Anal. Chim. Acta
46:179-191.
Gauch, H.G., and R.H. Whittaker. 1972. Comparison of ordination techniques.
Ecology 53:868-875.
Gibbs, C.F. 1979. Chlorophyll a and 'phaeo-pigments.' Aust. J. Mar. Freshwater
Res. 30:596-606.
Gieskes, W.W.C., and G.W. Kraay. 1983. Dominance of Cryptophyceae during the
phytoplankton spring bloom in the central North Sea detected by HPLC analysis of
pigments. Mar.Biol. 75:179-185.
Goeyens, L., E. Post, F. Deharis, A. Vandenhoudt, and W. Baeyens. 1982. The use of
high pressure liquid chromatography with fluorimetric detection for chlorophyll a.
determination in natural extracts of chloropigments and their degradation products.
Intern. J. Environ. Anal. Chem. 12:51-63.
Gordon, D.C. and Sutcliffe, W.H., Jr. 1974. Filtration of seawater using silver filters
for paniculate nitrogen and carbon analysis. Limnol. Oceanogr. 19:989-993.
Grasshoff, K., M. Ehrhardt, and K. Kremling. 1973. Methods of Seawater Analysis,
Second edition. Verlag-Chimie. Weinheim. 419pp.
Green, R.H. 1979. Sampling Design and Statistical Methods for Environmental
Biologists. New York, NY: John Wiley & Sons. 257pp.
Green, R.H. 1980. Multivariate approaches in ecology: The assessment of ecologic
similarity. Ann. Rev. Ecol. Syst. 11:1-14.
-------
B-87
Green, R.H. 1984. Some Guidelines for the Design of Biological Monitoring Pro-
grams in the Marine Environment pp. 647-655. In: Concepts of Marine Pollution
Measurements. (White, H.H., ed.). College Park, MD: University of Maryland Sea
Grant.
Green, R.H., and G.L. Vascotto. 1978. A method for the analysis of environmental
factors controlling patterns of species composition in aquatic communities. Water
Res. 12:583-590.
Heck, K.L. Jr. and R. J. Horwitz. 1984. Statistical Analysis of Sampling Data to
Assess Impact on Marine Environments, pp. 233-245. In: Concepts of Marine Pollu-
tion Measurements. (White, H.H., ed.). College Park, MD: University of Maryland
Sea Grant.
Hollander, M., and D.A. Wolfe. 1973. Nonparametric Statistical Methods. New
York, NY: John Wiley & Sons. 503pp.
Hurlbert, S.H. 1971. The nonconcept of species diversity: A critique and alternative
parameters. Ecol. 52:577-586.
Inskeep, W.P., and P.R. Bloom. 1985. Extinction coefficients of chlorophyll 1 and b,
in N, N-dimethylformamide and 80% acetone. Plant Physiol. 77:483-485.
Jacobs, F., and G.C. Grant. 1978. Guidelines for zooplankton sampling in quantita-
tive baseline and monitoring programs. Rep. No. 600/3-7-78-026. U.S. EPA Corval-
lis, OR. 51 pp.
Jeffrey, S.W., and G.M. Hallegraeff. 1980. Studies of phytoplankton species and
photosynthetic pigments in a warm core eddy of the East Australian Current. I.
Summer populations. Mar.Ecol.Prog.Ser. 3:285-295.
Jeffrey, S.W., and G.F. Humphrey. 1975. New spectrophotometric equations for
determining chlorophyll a, b_, c! and fl in higher plants, algae and natural phytoplank-
ton. Biochem. Physiol. Pflanzen. 167:191-194.
Jeffrey, S.W., M. Sielicki, and F.T. Hazo. 1975. Chloroplast pigment patterns in
dinoflagellates. J.Physcol. 11:374-384.
-------
B-88
Johnson, B.D., and P.J. Wangersky. 19??. Seawater filtration: Particle and impaction
considerations. Limnol. Oceanogr. 30(5):966-972.
Knight, R., and R.F.C. Mantoura. 1985. Chlorophyll and carotenoid pigments in
Foraminifera and their symbiotic algae: analysis by high performance liquid chroma-
tography. Mar. Ecol. Prog. Ser. 23:241-249.
Lorenzen, C.J. 1966. A method for the continuous measurement of in vivo chloro-
phyll concentration. Deep-Sea Res. 13:223-227.
Lorenzen, C.J. 1967. Determination of chlorophyll and pheo-pigments: spectrophoto-
metric equations. Limnol oceanogr. 12:343-346.
Magnien, R.E. 1986. A Comparison of Estuarine Water Quality Chemistry Analysis
on the Filtrate from Two Types of Filers. Maryland Office of Environmental Pro-
grams, Ecological Modeling and Analysis Division, Technical Report.
Baltimore, MD.
Mantoura, R.F.C., and C. A. Llewellyn. 1983. The rapid determination of algal
chlorophyll and carotenoid pigments and their breakdown products in natural waters
by reverse-phase high-performance liquid chromatography. Anal. Chim. Acta
151:297-314.
McEwen, G.F., M.W. Johnson, and T.R. Folsom. 1954. A statistical analysis of the
performance of the Folson plankton sample splitter based upon test observations.
Arch. Meteorol. Geophys. Biolkimatoll., Ser. A. Meteorol. Geophys. 7:502-527.
McGowan, J.A., and D.M. Brown. 1966. A new opening-closing paired zooplankton
net. Univ. Calif. Scripps Inst. Oceanogr. (Ref. 66-23).
McGowan, J.A., and V.J. Fraundorf. 1966. The relationship between size of net used
and estimates of zooplankton diversity. Limnol. Oceanogr. 11:456-469.
Moran, R., and D. Porath. 1980. Chlorophyll determination in intact tissues using
N,N-Dimethylformamide. Plant Physiol. 65:478-479.
Parsons, T.R., Y. Maita, and C.M. Lalli. 1984. A Manual of Chemical and Biological
Methods for Seawater Analysis. New York, NY: Pergamon Press. 173pp.
-------
B-89
Pielou, E.G. 1966. The measurement of diversity in different types of biological
collections. J. Theroet. Biol. 13:131-144.
Pielou, E.G. 1970. An Introduction to Mathematical Ecology. New York, NY:
Wiley-Interscience. 286 pp.
Pielou, E.G. 1977. Mathematical Ecology. New York, NY: Wiley-Interscience.
385 pp.
Puget Sound Monitoring Management Committee. 1988. Puget Sound Ambient
Monitoring Program. Puget Sound Water Quality Authority, Seattle, WA. 145 pp.
Saila, S.B., D. Chen, VJ. Pigoga, and S.D. Pratt. 1984. Comparative Evaluation of
Some Diversity Measures for Assessing Environmental Changes in an Estuarine
Community, pp. 217-233. In: H. H. White (ed.), Concepts of Marine Pollution
Measurements. College Park, MD: University of Maryland Sea Grant.
SCORE. 1966. Monographs on oceanographic methodology. Vol.1. Determination
of photosynthetic pigments in seawater. UNESCO Press, Paris. 69pp.
Seely, G.R., MJ. Duncan, and W.E. Vidaver. 1972. Preparative and analytical
extraction of pigments from brown algae with dimethyl sulfoxide. Mar Biol.
12:184-188.
Shioi, Y., R. Fukae, and T. Sasa. 1983. Chlorophyll analysis by high-performance
liquid chromatography. Biochim. Biophy. Acta 722:72-79.
Shoaf, W.T., and B.W., Lium. 1976. Improved extraction of chlorophyll a and b from
algae using dimethyl sulfoxide. Limnol. Oceanogr. 21:926-928.
Siegel.S. 1956. Nonparametric Statistics for the Behavioral Sciences. New York,
NY: McGraw-Hill Book Co. 312pp.
Smith, R.W., and C.S. Greene. 1976. Biological communities near submarine outfall.
J. Water Poll. Control Fed. 48(8):1894-1912.
Smith, W., V.R. Gibson, L.S. Brown-Leger, and J.F. Grassle. 1979. Diversity as an
indicator of pollution: cautionary results from microcosm experiments, pp. 269-277.
In: Ecological Diversity in Thp.nrv and Practice (Graale et al, eds). Fairland, MD:
Internal. Coop. Publ. House.
-------
B-90
Sneath, P.,andR. Sokal. 1973. Numerical Ta'xonomv. San Francisco, CA:W.H.
Freeman and Co. 573 pp.
Snedecor, G.W., and W.G. Cochran. 1967. Statistical Methods. Sixth ed. The Iowa
State University Press.
Sokal, R.R., and FJ. Rohlf. 1969. Biometry. San Francisco, CA: W.H. Freeman and
Co. 776pp.
Speziale, B.J., S.P. Schreiner, P.A. Giammatteo, and J.E. Schindler. 1984. Compari-
son of N, N-dimethylformamide, dimethyl sulfoxide, and acetone for extraction of
phytoplankton chlorophyll. Can. J. Fish. Aquat. Sci. 41:1519-1522.
Sprules, W.G. 1977. Crustacean zooplankton communities as indicators of limnologi-
cal conditions: an approach using principal component analysis. J. Fish Res. Board
Can. 34:962-975.
Stauffer, R.E., G.F. Lee, and D.E. Armstrong. 1979. Estimating chlorophyll extrac-
tion biases. J. Fish. Res. Bd. Can. 36:142-157.
Stofan, P.F., and G.C. Grant. 1978. Phytoplankton sampling in quantitative baseline
and monitoring programs. Rep. No. 600/3-78-025. U.S. EPA, Corvallis, OR. 83 pp.
Strickland, J.D.H., and T.R. Parsons. 1968. A practical handbook of seawater analy-
sis. Fish. Res. Bd. Can. Bull. No. 167, 311 p.
Strickland, J.D.H., and T.R. Parsons. 1972. A Practical Handbook of Seawater
Analysis. Bull. Fish. Res. Board. Can. 310 pp.
Swartz, R.C., D.W. Schultz, G.R. Ditsworth, W.A. DeBen, and F.A. Cole. 1985.
Sediment toxicity, contamination, and macrobenthic communities near a large sewage
outfall. In: Validation and Predictability of Laboratory Methods for Assessing the Fate
and Effects of Contaminants in Aquatic Ecosystems. (T.T. Boyle, ed.) Philadelphia,
PA: American Society for Testing and Materials (ASTM). pp. 152-175.
Tnrondsen, J. 1978. Preservation and storage, pp. 69-71. In: Monographs on
Oceanographic Methodology. 6. Phytoplankton Manual. United Nations Education,
Scientific, and Cultural Organization, Paris.
-------
B-91
U.S. EPA. 1919. Handbook for Analytical Quality Controls in Water and Wastewater
Laboratories. Environmental Monitoring and Support Laboratory,
EPA-600/4-79-019, Cincinnati, OH.
U.S. EPA. 1982. Design of 301 (h) Monitoring Programs for Municipal Wastewater
Discharges to Marine Waters. U.S. Environmental Protection Agency, Office of
Water, Washington, D.C. 135 pp.
U.S. EPA. 1985. Interim Guidance on Quality Assurance/Quality Control (QA/QC)
for the Estuarine Field and Laboratory Methods. U.S. Environmental Protection
Agency, Office of Marine and Estuarine Protection. 154 pp.
U.S. EPA. 1987a. Puget Sound protocols. Final Report. 31pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1987b. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1989. Chesapeake Bay Basin Monitoring Program Atlas (Vol. 1 and 2).
U.S. Environmental Protection Agency, Chesapeake Bay Liaison Office,
Annapolis, MD.
U.S. EPA. 1989 Draft. Compendium of Methods for Marine and Estuarine Environ-
mental Studies. U.S. Environmental Protection Agency, Office of Water,
Washington, D.C.
UNESCO. 1968. Monographs on oceanographic methodology. 2. Zooplankton
sampling. United Nations Education, Scientific, and Cultural Organization, Paris.
174 pp.
UNESCO. 1973. Monographs on oceanographic methodology. 3. A guide to the
measurement of marine primary production under some special conditions. United
Nations Education, Scientific, and Cultural Organization, Paris. 73 pp.
UNESCO. 1978. Monographs on oceanography methodology. 6. Phytoplankton
manual. United Nations Education, Scientific, and Cultural Organization, Paris.
337pp.
-------
B-92
Venrick, E.L. 1978. Water-bottles, pp. 33-40. In: Monographs on Oceanographic
Methodology. 6 Phytoplankton Manual. United Nations Education, Scientific, and
Cultural Organization, Paris.
Venrick, E.L. 1978. The implications of subsampling. pp. 75-87. In: Monographs on
Oceanographic Methodology. 6. Phvtoplankton Manual. United Nations Education,
Scientific, and Cultural Organization, Paris.
Wiebe, P.M., and W.R. Holland. 1968. Plankton patchiness: Effects on repeated net
tows. Limno. Oceanogr. 13:315-321.
Williams, W.T. 1971. Principles of clustering. An. Rev. Ecolo. Syst. 2:303-326.
Winer, B.J. 1971. Statistical Principles in Experimental Design. New York, NY:
McGraw-Hill Book Company.
Wood, A.M. 1979. Chlorophyll a:b_ in marine planktonic algae. J. Phycol.
15:330-332.
Wood, L.W. 1985. Chloroform-methanol extraction of chlorophyll a. Can. J. Fish.
Aquat. Sci. 42:38-43.
Yentsch, C.S., and D.W. Menzel. 1963. A method for the determination of phyto-
plankton chlorophyll and phaeophytin by fluorescence. Deep-Sea Res. 10:221-231.
Zar, J.H. 1974. Biostatistical Analysis. Englewood Cliffs, NJ: Prentice-Hall, Inc.
620 pp.
Zimmerman, C.F. 1990. Personal Communication. University of Maryland, Center
for Environmental and Estuary Studies, Chesapeake Biological Laboratory,
Solomons, MD.
-------
B-93
B6.0 Aquatic Vegetation
Human activities that contribute the greatest direct impacts to vegetation loss and
modification are those causing physical alterations to the habitat. Physical alterations
may cause habitat loss directly by removal or indirectly by modifying natural pro-
cesses that significantly affect aquatic vegetation, such as freshwater inflow, hydrol-
ogy, sedimentation, and sea-level changes.
Marine and estuarine vegetation may be organized into the following groups:
Emergent vegetated wetlands
- marshes
- mangroves
Submerged aquatic vegetation (SAV)
- seagrasses
- other submerged plant communities
Although both of these habitat types provide essential functions for living marine
resources, the focus of this discussion is primarily on emergent vegetated wetlands due
to the recognized ecological value of these habitat types and the rapid rate of their
deterioration and loss in many coastal areas.
lilllllllliliilllllilli
:li||||^
lif|||||||^
Itiiiil^^
Jiilprt^
/y^BSmSIJ;!^^
|ff||||rp^
-------
B-94
Declines in the quality and quantity of marine habitats are thought to reduce produc-
tion of living marine/estuarine resources and to diminish other important values of
habitats. Marine and estuarine habitats perform important functions for all living
marine resources. Aquatic vegetation is used by these species for spawning, rearing,
feeding, migration, and shelter from predators. A positive correlation generally exists
between wetland and submerged aquatic vegetation (SAV) acreage and abundance of
commercially and recreationally important living marine resources (e.g., fish and crabs
in estuarine areas). Populations of living marine resources therefore appear to be
limited by the quantity and quality of aquatic vegetation. In fact, the condition of
wetlands may prove to be an effective ecological indicator of the condition of selected
bird and mammal populations.
Furthermore, marine and estuarine habitats, in particular vegetated wetlands (tidal
marshes) and submerged aquatic vegetation, perform other valuable functions to man
and the environment. These functions include erosion and flood protection, water
quality control, and waterfowl and wildlife utilization. Pressures brought on by
population shifts to coastal areas and associated industrial and municipal expansion
will continue to jeopardize coastal wetland resources. Monitoring changes in aquatic
vegetation will permit assessment of impacts to the habitat and the organisms it
supports.
B6.1 Rationale
It is essential to understand the effect of monitoring design characteristics on the
results and to standardize them as much as possible to ensure the comparability of
monitoring effects throughout the estuary.
Transects and Quadrants
Physical and biological attributes of wetland vegetation may be taken at regular
intervals along a transect: point line intersects (SCS, 1976). Critical biological
characteristics include species name, plant and mulch biomass, and foliar and basal
cover. Physical characteristics (e.g., salinity, water depth, flow rates) should also be
collected to assist in the interpretation of the biological data.
Plants may be harvested from standard-sized quadrats (e.g., 1 m2) along a transect
(SCS, 1976). Simple factors are used to convert plant species data from biomass per
plot to more common measures of productivity (e.g., pounds/acre or kilograms/
hectare). Measures should specify green weight or air-dry weight; factors for convert-
ing between green weight to air-dry weight exist.
B6.2
Monitoring Design
Considerations
-------
B-95
Measures of cover are fundamental to managing living resources of wetlands. Basal
cover at the erosion interface provides information concerning retention of detritus
material and potential for erosion. Basal cover complements measures of productivity
by providing information concerning the stability of the habitat. For example, al-
though a plant community may be highly productive, it may be susceptible to erosion
if there exists little basal cover at the erosion interface. This information may be used
to guide implementation of erosion control measures.
The US Department of Agriculture (USDA) has developed the Range Data System
(RDS), a nationwide database system for grass-based ecosystems. RDS is capable of
storing, retrieving, and analyzing data on wetland plant productivity and species
composition. RDS also provides a remarks field where data such as salinity and water
depth may be recorded.
Maps, Aerial and Remote Photography
Losses of aquatic vegetation acreage in an area of particular concern may be the
compelling reason for the evaluation of habitat within the estuary. The primary
sources of spatial information are maps, aerial photography, and remote or satellite
imagery. These tools can serve as a simple visual aid for communicating the problem
to the public as well as a main data source for wetland structural and functional
characteristics and assessment of cumulative impacts and trends. Aerial or remote
monitoring can measure changes in acreage, cover types, drainage, configuration and
other spatial parameters related to overall habitat condition and health.
Maps - Maps are the least expensive spatial data source and are also the most simpli-
fied. Generally, maps do not capture all visible characteristics that can be observed
from aerial photographs, rather selectively portray a subset of what is in an area.
However, maps such as U.S. Geological Survey Topographic Quadrangle maps or
U.S. Fish and Wildlife Service National Wetlands Inventory (NWI) maps can be used
as a base of information from which to compare changes in wetland acreage.
The main benefits of maps for identifying habitats is in their ease of use and low cost.
Their main disadvantage is that habitats are often not clearly or accurately delineated
on maps, and if they are, they are generally not updated frequently enough to provide
timely information on trends. For instance, although the FWS has mapped over 90%
of the wetlands in the coastal zone of the lower 48 states, many of these maps are over
10 years old.
-------
B-96
There have been efforts by some states and the Federal Government to develop high-
resolution automated databases of habitat types through a process of digitizing NWI
maps at a resolution of approximately one acre. The utility of high-resolution and
geographically referenced databases for resource managers has been evaluated by
FWS (Dahl, 1987). Digitizing NWI maps would make updating information on the
status and trends of habitat types easier and more accurate; however, the progress
toward digitizing all coastal wetland maps has been hampered by the time-consuming
and costly nature of the process involved.
Aerial Photography - The most broadly used source of habitat spatial information is
aerial photography. Aerial photographs can be produced at a range of scales and
spatial resolution. Small scales can be used when, for example, very fine detail is
desired for mapping vegetation cover types or complicated drainage patterns. Large
scale coverage can be used when a broad area of coverage is desired. In addition, color
infrared photography can be used to detect differences in soil moisture and subtle
differences in vegetation types accentuating the upland/wetland boundary.
Existing aerial photographs are inexpensive. However, new overflights can be expen-
sive and are often complicated, requiring professional assistance. The advantage of
aerial photographs is that they generally contain more detail than maps of the same
scale or than satellite imagery. When planning overflights for aerial photography, the
following factors must be considered:
tides
weather conditions
time of day (affects sun angle, atmospheric haze and cloud cover)
turbidity
Photographs from new overflights can be compared to base maps or older aerial
photographs to record changes over time.
Remote Sensing/Satellite Imagery - Satellite imagery is a useful source of spatial
data when very large study areas are involved and fine detail is not important. Al-
though expensive, when large study areas are involved, satellite imagery may become
more economical than aerial photography. Other than cost, a significant constraint in
the use of satellite imagery is the lack of resolution which could result in missing small
habitats or features; remote sensing provides a coarse estimate of size or boundaries.
-------
B-97
However, if the area under investigation is characterized by a very large homogenous
habitat, this constraint may not be as important.
Because images are acquired by scanners and stored in digital form, the spectral
properties, brightness values and other digitally stored data can be manipulated to
enhance or reveal characteristics (such as productivity) not always visible from aerial
photography. In addition, factors affecting water clarity, such as suspended sediments,
may affect these spectral properties. These factors, and the amount of cloud cover
over the estuary, should be considered when appraising the use of remote sensing.
Ground-truthing of satellite images is highly recommended in order that accurate
interpretations of the images may be achieved.
Sampling Period
The aerial coverage and bed density for SAVs vary from year to year due to cata-
strophic storms, exceptionally high precipitation and turbidity, and other poorly
understood natural phenomena. Therefore, short term monitoring will not provide an
accurate assessments of the trend in SAV coverage, but rather may reflect the impacts
of infrequent or periodic events. In addition, wetland losses in acreage occur at low
enough frequencies that yearly measurements are impractical. Furthermore, the types
of wetland losses that are now occurring appear to be moving away from larger
increments to a smaller form of wetland loss that may fall within the margin of error
for current detection limits.
Seasonal/Daily Considerations - In the conduct of habitat functional assessments,
seasonal variations have to be considered. SAV and wetland sampling should be
conducted during the period of peak biomass. Other factors combined with a limited
growing season also make for a finite window of opportunity for data capture: tides,
weather conditions, turbidity.
Skills and Training Required
When analyzing remotely sensed images, the availability and effectiveness of appro-
priate algorithms may be limited (Platt and Sathyendranath, 1988). Interpretation,
manipulation, and analyses of remotely sensed images requires persons with special-
ized skills handling image analyzing software and hardware. Furthermore, some
habitat assessment methods require a high degree of advanced understanding of the
science involved in the assessment, such as plant taxonomy or hydrology. The selec-
tion of the assessment method will therefore depend on the level of specificity required
to conduct the assessment and the available technical expertise.
-------
B-98
Spatial vs. Functional Trends
Spatial Trends - The most straightforward approach for assessing trends in the
aquatic habitat - and the one most typically used - is to document the loss of habitat
acreage (Table B6-2; France and Hedges, 1989). Spatial data alone are particularly
important because of the public concern over declining habitat acreage (both wetlands
and submerged aquatic vegetation) in the nation's estuarine areas. Good current
information on habitat locations and boundaries form a baseline for monitoring future
changes. However, delineating the extent of habitats alone does not provide a com-
plete measure of the quality of the habitat in terms of its value for fish, bird and marine
mammal populations.
Functional Trends - The purpose of assessing habitat functional trends is to compare
the relative value of one habitat to another or to assess changes in values of the same
habitat over time (Table B6-2). Detailed, site-specific, functional assessment studies
can be expensive, time-consuming, and often impractical when time or budgetary
constraints exist. Although more expensive, correlating functional habitat losses to
acreage lost is a more meaningful measure of the condition of estuarine resources than
acreage loss alone.
Where time and budget allow, implementation of detailed, carefully designed monitor-
ing programs or use of quantitative computer models usually give superior-quality
results. All functional assessment methods, however, are limited by the level of
understanding of the actual processes taking place in habitats, which vary with habitat
type, region of the country, and physical, chemical and biological conditions. Func-
tional assessments usually describe trends in water quality, hydrology, and biota that
are potentially attributable to habitat loss and impacts.
B6.3 Existing
Analytical
Methods
-------
B-99
Standardized, relatively rapid functional evaluation procedures generally provide
stronger replicability and technical comprehension. Two of the most frequently
standardized evaluation techniques are the Wetland Evaluation Technique (WET),
originally developed for the Federal Highway Administration and later revised for the
U.S. Army Corps of Engineers (Adamus et al., 1987) and the Habitat Evaluation
Procedure (HEP), developed by the U.S. Fish and Wildlife Service (U.S. FWS, 1980).
Many of the more recently developed habitat evaluation methods are modified ver-
sions of these techniques.
Minimum Habitat Matrix
As part of the Chesapeake Bay Program, minimum habitat guidelines for various
species are developed with the ultimate goal of reestablishing a balanced ecosystem.
This method is designed to provide information on the minimum habitat quality
needed by a target species and identifies those factors (both environmental and eco-
logical) required for the species. This information is formatted into a habitat require-
ment matrix that defines the habitat parameters needed for successful reproduction and
survival of the indicated species.
The matrices can indicate the vital environmental parameters that should be monitored
thus facilitating the setting up of monitoring programs. This process is used to esti-
mate the feasibility, benefits and potential costs of maintaining and protecting an
estuarine environment suitable for the successful reproduction and survival of aquatic
vegetation (Chesapeake Executive Council, 1988).
One potential problem with this method is that both the primary indicated species and
those organisms that the target species depend on for food are both tracked with the
intention of maintaining habitat quality for both. This method may not completely
recognize the complex species interdependence within estuarine environments.
Wetland Evaluation Technique (WET)
The Wetland Evaluation Technique outlines the procedure for conducting an assess-
ment of the following wetland functions and values:
Ground Water Recharge
Ground Water Discharge
Flood Flow Alteration
Sediment Stabilization
Production Export
Wildlife Diversity/Abundance
Aquatic Diversity/Abundance
Recreation
-------
B-100
Sediment/Toxicant Retention Uniqueness/Heritage
Nutrient Removal/Transformation
WET also assesses the suitability of wetland habitat for 14 waterfowl species groups, 4
freshwater fish species groups, 120 species of wetland-dependent birds, 133 species of
saltwater fish and invertebrates, and 90 species of freshwater fish. WET does not
evaluate any other important wildlife resources (e.g., game and furbearing mammals).
Other evaluation methods must be used to evaluate these other wildlife resources
(Adamus ef a/., 1987).
WET evaluates functions and values in terms of social significance, effectiveness, and
opportunity (Adamus et al., 1987). Social significance assesses the value of a wetland
to society due to its special designations, potential economic value, and strategic
location. Effectiveness assesses the capability of a wetland to perform a function due
to its physical, chemical or biological characteristics. Opportunity assesses the oppor-
tunity of a wetland to perform a function to its level of capability.
WET assesses functions and values by characterizing a wetland in terms of its physi-
cal, chemical, and biological processes and attributes (Adamus et al., 1987). This
characterization is accomplished by identifying threshold values for predictors. Pre-
dictors are simple, or integrated, variables that directly, or indirectly, measure the
physical, chemical, and biological processes or attributes of a wetland and its sur-
roundings.
Habitat Evaluation Procedure (HEP)
The Habitat Evaluation Procedure is a method that can be used to document the quality
and quantity of available habitat for selected wildlife species. HEP provides informa-
tion for two general types of wildlife habitat comparisons: (1) the relative value of
different areas at the same point in time; and (2) the relative value of the same area at
future points in time. By combining the two types of comparison, the impact or
improvement in habitat quality as a result of proposed or anticipated land and water
changes on wildlife habitat can be quantified (Lonard and Clairain, 1986).
In the evaluation procedure, the habitat quality of selected evaluation species are
documented based on an evaluation of the ability of key habitat components to supply
the life requisites of selected species. The evaluation involves using the same key
habitat components to compare existing habitat conditions and the optimum conditions
for the species of interest (U.S. FWS, 1980).
-------
B-101
There are a number of limitations to using a habitat approach in an evaluation system
as was pointed out by the Fish and Wildlife Service in HEP. Using habitat quality as
an evaluation standard limits the application of the methodology to those situations in
which measurable and predictable habitat changes are an important variable. It also
forces a long-term "averaging" type of analysis. There is no assurance that popula-
tions will exist at the potential levels predicted by habitat analysis, as the analysis may
not include all of the environmental or behavioral variables that may limit populations
below the habitat potential. In addition, socio-economic or political constraints by
man may prevent the actual growth of certain populations to these potential levels
(U.S. FWS, 1980).
B6.4 QA/QC
Considerations
Whichever assessment method is selected, the researcher should ensure that it has a
high degree of replicability. All functional assessment methods contain assumptions
which will impact the results of the assessment. Some assessment methods also
require excessive judgement on the part of the researcher and as a result, different
users may arrive at different results. In selecting a method, the researcher should first
determine whether assumptions are based on scientific knowledge or on the opinion of
the author.
Before applying a method extensively within a region, the researcher should field test
the method by comparing its ratings to those indicated independently by long-term
research studies of representative habitat sites. Ground-truthing remotely sensed and
aerial images is highly recommended in order to ensure accurate interpretation of the
images.
B6.5 Statistical In designing an aquatic vegetation monitoring program the researcher must first
Design determine the resources available. This will determine the extent of the monitoring
Considerations program and the level of detail it can provide. If resources are limited, the extent of
the monitoring program may only involve an assessment of the aerial extent of wet-
lands. If more resources are available and it is decided to undertake a functional
assessment, the researcher must decide which functions are of greatest interest in the
area of study or which will provide the most information concerning the habitat
attributes perceived to be of greatest value to the area.
Due to time and monetary constraints, it may not be possible to evaluate all of the
habitats within the region of concern, for example, an estuary. In such cases, the
researcher may choose to only evaluate a select few habitats for evaluation. These
-------
B-102
may be those with potentially the greatest ecological value, those undergoing
remediation, or those which are undergoing significant loss or modification.
Most habitat evaluation techniques involve some form of ranking, either quantitative
or qualitative. The purpose of the monitoring in a estuary program is to periodically
compare baseline conditions to conditions after the management plan has been imple-
mented. In terms of habitat monitoring this means comparing the baseline acreage or
functional rating of the habitat to conditions some time after the management plan has
been implemented. Therefore, replicability of results are imperative.
See also Statistical Design Consideration: Power Analysis and Power Cost Analysis
(Appendix B Introduction; Section B.3).
The results of aquatic vegetation evaluation will track trends in habitat quality and B6.6
quantity so that they may be projected. Additionally, changes in land use or other Usฉ Of Data
anthropogenic activities over time can be measured via their impact upon aquatic
habitats. It is a primary concern to accumulate an adequate set of data to get a baseline
so that any changes in habitat quantity and quality that are derived from mitigation,
regulation or estuary management program can be assessed. Historical data can chart
changes in land use leading up to the present state.
The tracking of habitats may be an effective method of monitoring for estuarine birds
and mammals. These organisms are highly mobile and their presence or absence at a
particular tag and recapture monitoring stations may not be significant - individuals
may be situated in another similar location. However, monitoring habitats which these
organisms require (e.g. to nest, breed, feed, use for shelter from predators, etc.) may be
an effective means of monitoring the threat to their well-being in the estuary. Loss of
biologically critical habitats will indicate the potential decline or loss of estuarine bird
and mammal populations.
Although single coastal development projects that result in modification or loss of
estuarine vegetation may not have serious immediate environmental repercussions
beyond the point of initial impact, many such projects in succession over time and over
a broad geographic area may have far-reaching ecological consequences (Day et al.,
1989). Without reasonably accurate information on long-term and cumulative impacts
of human activities on marine and estuarine habitats, good decisions on permitting,
mitigation, or management of fisheries and wildlife will be difficult.
-------
B-103
B6.7
Summary and
Recommendations
Rationale
Monitoring habitat quantity and quality will provide information to assess impacts on
living resources and to gauge the effectiveness of mitigation, regulatory and manage-
ment programs
Monitoring Design Conditions
Maps have the following properties
- least expensive
- easy to use and least cost
- most simplified: do not capture all features
- habitats not always clearly or accurately delineated
- often not updated on regular basis
Aerial photography has the following properties
- range of scales and spatial resolution
- color infrared photography to detect differences in soil moisture
and vegetation types
- relatively inexpensive
- contain more detail than maps or satellite imagery
Satellite imagery has the following properties
- useful when large study areas involved and great detail not
important
- spectral properties, brightness values and other digitally stored
data can be manipulated to enhance characteristics
- expensive
- lack of resolution found in aerial photographs
Possible resilience of habitat functions should be considered; need to be
able to assess long-term and cumulative impacts
Seasonal and daily variations may have an impact on habitat functions and
the ability to identify habitat boundaries and thus evaluation results
Some methods require high degree of technical understanding while other
methods can be easily understood and implemented with more limited
technical understanding
-------
B-1041
Existing Analytical Methods
Analyses of spatial trends is the most straight forward approach for assess-
ing trends in the aquatic habitat. Typically, documentation of habitat
acreage loss is assessed.
Function analysis considerations
- provide more detailed information concerning the quality of
habitat
- can be expensive and time consuming
- limited by the level of understanding of actual processes
- WET and HEP two of the most frequently used standardized
evaluation techniques
QAIQC Considerations
Need high degree of replicability within a method
Method should be field tested before being applied extensively
Statistical Design Considerations
Available resources
Determine which functional assessments will provide information concern-
ing attributes of greatest value to the area
Determine areal extent of monitoring program
- entire watershed
- select habitats with greatest potential value or undergoing signifi-
cant loss or modification
Use of Data
Project future trends in habitat quality and quantity
Evaluate severity of habitat impacts from changes in land use or other
anthropogenic activity over time
-------
s $
B-105
Assess improvements in habitat quality and quantity resulting from mitiga-
tion, regulation, and estuary management programs
Ecological indicator of potential decline or loss of estuarine bird and
mammal populations
B6.8 Literature Ackleson, S.G. and V. Klemas. 1987. Remote sensing of submerged aquatic vegeta-
Clted and tion in the lower Chesapeake Bay: A comparison of Landsat TM and MSS imagery.
References Rem. Sen. Env. 22:235-248.
Adamus, P.R. 1987. Wetland evaluation technique for bottomland hardwood func-
tions. US EPA Office of Wetlands Protection, Washington, D.C. 59 pp.
Adamus, P.R. 1988. The FHWA/Adamus (WET) method for wetland functional
assessment, pp. 128-133 in: D.D. Hook and Others (Eds.), The Ecology and Manage-
ment of Wetlands. Croom Helm Publishers.
Adamus, P.R., L.T. Stockwell, EJ. Clairain, Jr., M.E. Morrow, L.P. Rozas, and R.D.
Smith. 1987. Wetland Evaluation Technique (WET). Volume I. Literature Review
and Evaluation Rationale. U.S. Army Corps of Engineers, Waterways Experiment
Station, Vicksburg, MS.
Adamus, P.R., E.J. Clairain, Jr., D.R. Smith, and R.E. Young. 1987. Wetland Evalua-
tion Technique (WET). Volume II. Technical Report Y-87. U.S. Army Corps of
Engineers, Waterways Experiment Station, Vicksburg, MS.
Bain, M.B. and J.L. Bain. 1982a. Habitat suitability index model: coastal stocks of
striped bass. Rep. Nad. Coastal Ecosystems Team, U.S. Fish. Wild. Serv. Rep. No.
FWS/OBS 82/10.1, Washington, D.C. 47 pp.
Beauchamp, R.G. 1974. Marine Environment Planning Guide for die Hampton
Roads/Norfolk Naval Operating Area. Naval Oceanographic Office. Spec. Pub.
No. 250. Naval Ocn. Off. Washington, D.C. 262 pp.
Bovee, K.D. and T. Cochnauer. 1977. Development and evaluation of weighted
criteria, probability-of-use curves for instream flow assessments: fisheries. Instream
How Information Paper 3. U.S. Fish Wildl. Serv. FWS/OBS-77/63. 39 pp.
-------
B-1061
Chesapeake Executive Council. 1988. Habitat Requirements for Chesapeake Bay
Living Resources. Chesapeake Bay Program, Annapolis, Md.
Christensen, E.J., J.R. Jensen, E.W. Ramsey, and H.F. Mackey. 1988. Aircraft MSS
data registration and vegetation classification for wetland change detection. Int. J.
Rem. Sens. 9:23-28.
Dahl,T.E. 1987. Wetlands Mapping in the Coastal Zone: Progress Towards a
National Digital Data Base. pp. 465-476. In: Coastal Zone '87. WWDiv/ASCE,
Seattle, WA, May 26-29.
Day, J.W., C.A.S. Hall, W.M. Kemp, and A. Yanez-Arancibia. 1989. Estuarine
Ecology. New York, NY: John Wiley & Sons. 558pp.
Deane, G.C., P.A. Bradbury, and M.W.C. Barr. 1989. The choice of imagery and
interpretation techniques for operational remote sensing. In: Remote Sensing for
Operational Applications. (Barrett, E.G., and K.A. Brown, eds.). Technical Contents
of the 15th Annual Conference of the Remote Sensing Society, University of Bristol,
September 13-15,1989. pp. 79-85.
Dee, N. Baker, J., Drobner, N., Duke, K. Whitman, I., and D. Fahrigner. 1973.
Environmental Evaluation System for Water Resources Planning. Water Resources
Research 9(3):523-534.
Ferraro, S.P., F.A. Cole, W.A. DeBen, and R.C. Swartz. 1989. Power-cost efficiency
of eight macrobenthic sampling schemes in Puget Sound, Washington, USA. Can. J.
Fish. Aquatic. Sci. 46(10):2157-2165.
France, M. J., and P.D. Hedges. 1989. The appropriate use of remotely senesed
satellite data for water resource assessment and modelling. In: Remote Sensing for
Operational Applications. (Barrett, E.G., and K.A. Brown, eds.). Technical Contents
of the 15th Annual Conference of the Remote Sensing Society, University of Bristol,
September 13-15, 1989. pp. 147-153.
Galloway, G.E. 1978. Assessing Man's Impact on Wetland. Sea Grant Publications
Nos. UNC-SG-78-17 or UNC-WRRI-78-136, University of North Carolina, Raleigh,
North Carolina.
Gosslink, J. and L. Lee. 1986. Cumulative impact assessment in bottomland hard-
wood forests. U.S. EPA Office of Wetlands Protection, Washington, D.C.
-------
B-107
-.5,..
Gross, M.F., and V. Klemas. 1986. The use of airborne imaging spectrometer (AIS)
to differentiate marsh vegetation. Rem. Sen. Env. 19:97-103.
Gupta, T.R., and J.H. Foster. 1973. Valuation of Visual-Cultural Benefits From
Freshwater Wetlands in Massachusetts. Journal of the Northeastern Agricultural
Council 2(l):262-273.
Homer, R.R. and R.w. Mar. 1982. Guide for water quality impact assessment of
highway operations and maintenance. WA-RD39.14. Dept. Civil Eng., Univ. Wash-
ington, Seattle.
Jensen, J.R. et al. 1986. Remote sensing inland wetlands: A multispectral approach.
Photo. Eng. Rem. Sen. 52:87-100.
Kibby, H.V. 1978. Effects of Wetlands on water Quality. Proceedings of the sympo-
sium on Strategies for Protection and Management of Floodolain Wetlands and other
Riparian Ecosystems. General Technical Report No. GTR-WO-12, U.S. Department of
Agriculture, Forest Service, Washington, D.C.
Klein, R. and J.C. O'Dell. 1987. Physical Habitat Requirement for Fish and Other
Living resources Inhabiting Class I and II Waters. Internal Document, Md. Dept. of
Nat. Res., Tidewater Administration.
Leonard, R.I., E.J. Clairain, Jr., R.T. Huffman, J.W. Hardy, L.D. Brown, P.E. Ballard,
and J.W. Watts. 1981. Analysis of methodologies for assessing wetlands value. U.S.
Water Resources Council, Washington, D.C., and Corps of Engineers, Vicksburg, MS.
Lippson, A.J., M.S., Haire, A.F. Holland, F. Jacobs, J. Jensen, R.L. Moran-Johnson,
T.T. Polgar and W.A. Richkus. 1979. Environmental Atlas of the Potomac Estuary.
Prepared for Power Plant Siting Program, Maryland Department of Natural Resources
by Environmental Center, Martin Marietta Corporation, 1450 S. Rolling Rd., Balti-
more, MD 21227.
Lonard, R.I. and E.J. Clairain, Jr. 1986. Identification of Methodologies for the
Assessment of Wetland Functions and Values. In: Proceedings: National Wetlands
Assessment Symposium. Association of State Wetland Managers, Inc., Portland, ME,
June 17-20. pp. 66-72.
-------
B-1081
MacDonald, P.O., W.E. Prayer, and J.K. Clauser. 1978. Documentation, chronology,
and future projects on of bottomland hardwood habitat loss in the Lower Mississippi
Alluvial Plain. US Fish. Wildl. Serv., Vicksburg, MS.
Marble, A.D. and M. Gross. 1984. A method for assessing wetland characteristics
and values. Landscape Planning 2:1-17.
Meyer, J.L. 1987. A simple method for the estimation of highway storm water
pollutant loadings to receiving waters. Proc. Svmps Wetland Hydrology. Assoc. State
Wetland Managers, Berne, N.Y.
Platt, T. and S. Sathyendranath. 1988. Oceanic primary production: Estimation by
remote sensing at local and regional scales. Science 241:1613-1619.
Overcash, M.R., S.C. Bingham, and P.W. Westerman. 1981. Predicting runoff
pollutant reduction in buffer zones adjacent to land treatment sites. Trans. ASCE
24:430-435.
Reppert, R.T., Siglero, W., Stakhiv, E., Messman, L., and Meyers, C. 1979. Wetland
values: Concepts and methods for wetlands evaluations. IWR Research Report
79-R-l, U.S. Army Engineer Institute for Water Resources, Fort Belvoir, Virginia.
Rose, P.W., and P.C. Rosendahl. 1983. Classification of Landsat data for hydrologic
application, Everglades National Park. Photo. Eng. Rem. Sen. 49:505-511.
SCS. 1976. National Range Handbook. U.S. Department of Agriculture, Soil Conser-
vation Service.
SCS Engineers. 1979. Analysis of selected functional characteristics of wetlands.
Contract No. DACW73-78-R-0017, Reston, VA.
Simon, B.D., L.J. Stoerzer, and R.W. Watson. 1987. Evaluating the functional
significance of wetlands for flood storage. Wisconsin Dept. Natur. Resour., Madison,
Wisconsin.
Shears, J.R. 1989. The application of airborne remote sensing to the monitoring of
coastal saltmarshes. In: Remote Sensing for Operational Applications. Technical
Contents of the 15th Annual Conference of the Remote Sensing Society, University of
Bristol, September 13-15,1989. Barrett, B.C., and K.A. Brown, eds. pp. 371-379.
-------
B-109
Shudliner, P.W., Cope, C.F., and Newton, R.B. 1979. Ecological effects on highway
fills of wetlands. Research Report. National Cooperative.
Solomon, R.D., Colbert, B.K., Hanses, W.J., Richardson, S.E., Ganter, L.W. and B.C.
Vlachos. 1977. Water Resources Assessment Methodology (WRAM) - Impact
Assessment and Alternative Evaluation. Technical Report Y-77-1, Environmental.
Effects Laboratory, U.S. Army Engineer Waterways Experiment Station, CE,
Vicksburg, MS.
State of Maryland Department of Natural Resources. Undated. Environmental
Evaluation of Coastal Wetlands (Draft). Tidal Wetlands Study, pp. 181-208.
State University of New York at Syracuse (SUNY). 1987. Wetlands evaluation
system for Onondaga County, New York State. Draft. 93 pp.
Stevenson, J.C. and N. Confer. 1978. Summary of Available Information on Chesa-
peake Bay Submerged Vegetation. FWS/OBS 78/66. U.S. Fish and Wildlife service,
co-sponsored by Maryland Dept. of Natural resources and U.S. Environmental Protec-
tion Agency. 335 pp.
Sweet, B.C., J.E. Poppleton, A.G. Shuey, and T.O. Peeples. 1980. Vegetation of
central Florida's east coast use the distribution of 6 vegetational complexes of Merritt
Island and Cape Canaveral peninsula. Rem. Sen. Env. 9:93-108.
Ulanowicz, R.E., and D. Baird. 1986. A Network Analysis of the Chesapeake Bay
Ecosystem. Maryland department of Natural Resources Tidewater Administration.
Ref. No. [UMCEES] CBL 86-79.
U.S. Army Engineer Division, Lower Mississippi Valley. 1980. A habitat evaluation
system for water resources planning. US Army Corps of Engineers, Lower Missis-
sippi Valley Division, Vicksburg, MS.
U.S. Army Engineer Division, New England. 1972. Charles River: Main Report and
Attachments. Waltham, Massachusetts.
U.S. Department of Agriculture. 1978. Wetlands evaluations criteria-water and
related land resources of the Coastal Regions, Massachusetts. Soil Conservation
Service, Amherst, Massachusetts. U.S. Environmental Protection Agency, Office of
Wetlands Protection 1988. Use of Geographic Information Systems for Wetlands
Protection. Washington, D.C.: Office of Wetlands Protection, US EPA.
-------
B-1101
U.S. EPA. 1987. Report of the workshop on habitat requirements for the Chesapeake
Bay living resources. Prepared by Eastern Research Group, Inc.
U.S. Fish and Wildlife Service (U.S. FWS). 1980. Habitat Evaluation Procedures
(HEP) Manual (102ESM). U.S. Department of the Interior, Washington, D.C.
U.S. Fish and Wildlife Service. 1982. Standards for the development of habitat
suitability index models. 103 ESM. U.S. Fish Wildl. Serv. n.p.
Virginia Institute of Marine Science. Undated. Evaluations of Virginia Wetland
(Mimeographed). Glouchester Point, VA.
-------
B-111
B7.0 Benthic Infauna Community Structure
B7.1 Rationale The objective of monitoring the benthic infauna is to detect and describe spatial and
temporal changes in the structure and function of benthic communities. These results
can be used to assess the condition - the "health" - of benthic habitats, to monitor rates
of recovery following environmental interventions, and, since the benthic infauna
represent a significant food source for many estuarine organisms, to provide an early
warning of potential impacts to the estuarine ecosystem.
The benthic habitat may broadly be divided into hard- and soft-bottom habitats. Since
the predominant habitat in estuaries is soft-bottom, assessment of soft-bottom commu-
nity structure will be the emphasis of this section.
Monitoring the abundance of individual organisms in the benthic infauna is a widely-
accepted means of measuring the condition of the benthic habitat (Bilyard, 1987).
Benthic infaunal organisms are exceptional indicators of benthic conditions since:
they are generally sedentary - observed effects are in response to local
environmental conditions
they are sensitive to habitat disturbance - communities undergo dramatic
changes in species composition and abundance in response to environmen-
tal perturbations
they often mediate the transfer of nutrients and toxic substances in the
ecosystem - via bioturbation and as important prey organisms
Furthermore, monitoring benthic infauna is one of a handful of methods which pro-
vides in situ measures of biotic health and habitat quality. The assessment of benthic
infaunal community structure is a powerful tool in the evaluation of spatial and
temporal effects of anthropogenic and natural disturbance.
B7.2 Monitoring The level of effort required to assess benthic community structure is relatively high. A
Design field survey is required to collect organisms; sorting, identifying, and enumerating
Considerations specimens require labor-intensive and generally expensive processes. Evaluations of
benthic community structure and function may also be costly and time-consuming.
The results of benthic monitoring programs can vary substantially depending on the
objectives and corresponding design specifications. The characteristics that are
primarily responsible for the variability in the results are:
-------
B-112
type of sampling gear
sample sorting protocols
level of taxonomy
location and timing of sample collection
It is essential to understand the effect of these design characteristics on the results and
to standardize them as much as possible to ensure the comparability of monitoring
effects throughout the estuary.
Analyses of power-cost efficiencies are essential in selecting the appropriate sampling
gear and sample processing protocols. Ferraro et al. (1989) provide an example of
power-cost analyses.
Sampling Devices
Sample collection protocols influence all subsequent laboratory and data analysis; it is
key that benthic samples be collected using acceptable and standardized techniques.
Several types of devices can be used to collect benthic macroinvertebrate samples:
trawls, dredges, grabs, and box corers (Mclntyre et al., 1984). Many of these devices
sample the benthos in a unique manner. Accordingly, conducting comparisons among
data collected using different devices is inadvisable. It is recommended that the same
area and volume of the sediment be sampled since different species of benthic
macroinvertebrates exhibit different horizontal and vertical distributions (Elliot, 1971).
Furthermore, collection of sediments and benthic organisms should be done concur-
rently in order to mitigate the costs of field sampling and to permit sound correlation
and multivariate analyses (see Sediment Pollutants). Therefore, it is recommended
that the sampling device also be suitable for sampling the sediment.
Collection of an acceptable sediment sample for infaunal analyses requires that the
sampler:
create a minimal bow wake when descending
form a leakproof seal when the sediment sample is taken
prevent winnowing and excessive sample disturbance when ascending
-------
B-113
allow easy access to the sample surface in order that undisturbed
subsamples may be taken
Penetration well below the desired sampling depth is preferred to prevent sample
disturbance as the device closes. It is optimal to use a sampler that has the means of
weight adjustment in order that penetration depths may be modified.
Trawls and Dredges - Trawls and dredges generally collect organisms over a variable
and relatively large area (Fredette et al., 1989; U.S. EPA, 1987a). Their design and
use precludes the collection of infaunal organisms at sediment depths greater than a
few centimeters.
The data collected using these devices are qualitative since consistent and reproducible
sampling of a consistent benthic area is not possible. However, these devices are
suitable for reconnaissance of potential sampling sites. Data collected using trawls or
dredges may be used to generate a list of organisms present in the area. A list of
species may be used to select sampling locations for a monitoring program.
Grab Samplers - Grab samplers are capable of consistent sampling of bottom habi-
tats; depending on the size of the device, areas 0.02 to 0.5 m2 and depths ranging from
5 to 15 cm may be sampled. Limitations of grab samplers include:
variability among samples in penetration depth depending on sediment
properties
oblique angles of penetration which result in varying penetration depths
within a sample
the sample is inevitably folded resulting in the loss of information concern-
ing the vertical structure of benthic communities in the sediments
However, the careful use of these devices will provide quantitative data. Grab sam-
plers are the tools of choice for a number of estuarine and marine monitoring programs
due to their ability to provide quantitative data at a relatively low cost (Fredette et al.,
1989; U.S. EPA, 1987a).
Core Samplers - Core samplers utilize a surrounding frame to ensure vertical entry;
vertical sectioning of the sample is possible (U.S. EPA, 1987a). These devices are
capable of maximum penetration depths of 15 cm and may collect volumes 5 to 10
times that of grab samplers. Limitations of box corers include:
-------
B-114
its large size and weight require the use of cranes or winches and a large
vessel for deployment
higher construction expenses
lack of calibration studies to permit comparisons to grab samples
The Hessler-Sandia box corer uses dividers to section the core into subsections,
facilitating subsampling of the core. Box corers are recognized as the tools of choice
for maximum accuracy and precision when sampling soft bottom habitats.
Suction Samplers - In situ suction devices use the flow of water to collect sediments
and benthic organisms (Eleftheriou and Holme, 1984). An open-ended suction tube
draws sediments and organisms into the flow of water, organisms are trapped in a
sieve or mesh bag at the end of the tube. Limitations of suction devices include:
little control of the volume of sediment collected
collection of animals from surrounding sediments, inflating infauna
abundances
organisms are typically turbulently abraded during collection, making
identifications even more difficult
Some samplers can be operated remotely; more often they are operated by divers. Any
diver-operated collection procedures will be restricted to SCUBA depths and condi-
tions permitting safe diving (e.g., relatively calm waters, reasonable visibility) thereby
limiting their use in monitoring programs.
Sediment Profiling Camera - The sediment profiling camera is a unique tool which
allows vertical, in situ imaging of the water-sediment interface. Photographs of
biological activity may be made at a maximum depth of 18 cm, however, penetration
depth of the viewing prism may be a limited due to the physical characteristics of the
sediment (i.e. penetration depths are greater in silt than sand).
The sediment profiling camera system can give measures of organism tube density,
thickness of the pelletal layer, and infauna present in the image area (Rhoads and
Germano, 1982). Although this tool provides qualitative information concerning the
benthos activity, the sediment profiling camera system cannot provide quantitative
information on species diversity, abundance, and biomass. Information concerning
-------
B-115
species composition cannot be gained using this tool. Consequently, characterizations
of the benthic community structure cannot be made from the data taken from these
images.
The use of the sediment profiling camera is not recommended for the collection of
benthic community structure data. However, the sediment profile camera proves to be
effective as a reconnaissance tool. Delineation of physically similar sampling sites
may be determined through the use of this tool, aiding in the selection of sampling
stations. Ground-truthing of these images by means of laboratory analyses of collected
material is highly recommended.
Recommendations - If quantitative analyses of benthic community structure are
required, a grab or core sampler is recommended. Either of these sampling devices
permit adequate sampling of both sediment and benthic infaunal communities with the
one sampling device.
Area Sampled
Different species of benthic macroinvertebrates have different scales of horizontal and
vertical spatial distribution (Elliot, 1971; Livingston et al, 1976; Downing, 1979).
Costs of laboratory analyses of the sample increase with increased sample unit size.
Analyses of spatial and temporal scale, statistical power, as well as costs, will assist in
determining optimal sample unit size. It is highly recommended that a standard
sample unit size (same surface area and depth) be analyzed in order to ensure data
comparability.
Sorting Protocols
There are several options which may be considered when designing sorting protocols.
In order to assure comparability among studies, a consistent set of sorting procedures
is highly recommended.
Sieving: Mesh Size and Location - The use of different sieve mesh sizes limits the
comparability of results among estuarine monitoring studies (Reish, 1959; Rees,
1984). The major advantage of using a smaller mesh size is the retention of both
juveniles and adults organisms as well as large- and small-bodied taxa. The major
disadvantage is the concomitant increased cost of sample processing. For example,
using a 0.5 mm mesh over a 1.0 mm mesh could increase retention of total
macrofaunal organisms by 130-180 percent; however, increased costs for processing
the samples may be as much as 200 percent (U.S. EPA, 1987a).
-------
B-1161
It is recommended that a standard mesh size is selected for all monitoring studies
within the estuary. In Puget Sound, the 1.0 mm mesh size is commonly used in impact
assessments and the 0.5 mm mesh size is sometimes used in recruitment studies. The
issue of data comparability was resolved by recommending sequential screening and
analysis procedures (Becker and Armstrong, 1988). By screening samples through
both the 1.0 and 0.5 mm sieves, the 1.0 mm fraction could be analyzed separately and
the data compared with other impact studies throughout the estuary. The data for the
0.5 mm fraction of the samples could then be combined to yield information on the
complete sample for comparison with other recruitment data.
Sieving can be conducted either aboard the survey vessel or onshore after the cruise is
completed. The major difference is that aboard the vessel, sieving occurs prior to
fixation; onshore, sieving occurs after fixation. Fixation of organisms in the sediment
before being sieved has the problem of the timely distribution of adequate concentra-
tions of the fixative to all the organisms in the sample. Sieving characteristics of
organisms can be greatly modified by decomposition and deterioration of body parts.
When a large number of samples are to be collected during a cruise, field sieving
results in a significant reduction in the volume of material stored and later transported
to the laboratory.
Use of Relaxants - The use of relaxants prior to sieving and fixation reduces the
tendency of organisms to distort their shape or autotomize, and consequently facilitates
taxonomic identifications and morphometric measurements (U.S. EPA, 1987a).
Complete organisms, having a "natural" appearance, are easier to identify to lower
taxonomic groups than are distorted or fragmented specimens.
The magnitude to which relaxants can influence taxonomic identification, thereby
limiting comparisons among monitoring studies, is unknown. It is not recommended
that comparisons be conducted between studies in which relaxants were used and those
in which they were not used.
Use of Stains - Vital stains primarily rose bengal are added to samples to
facilitate sorting. The stain colors most infauna, enhancing their contrast with back-
ground debris. Not all taxa stain adequately and some taxonomists have found that
staining interferes with the identification of certain taxa. Although staining may aid
the sorting process, a proper quality control program should ensure that sorting effi-
ciency is maintained whether or not a stain is used.
-------
B-117
Level of Taxonomy
The necessity to identify organisms to the level of species has recently been questioned
(Warwick, 1986). The liabilities of identifying organisms to the species level include:
inaccurate and imprecise identifications (Ellis, 1985)
identifications are time-consuming and costly; costs can be reduced as
much as 30-50 percent by limiting identification of samples to higher
taxonomic levels (U.S. EPA, 1987a)
Identifications to higher taxonomic groups can provide gross characterizations of
benthic infaunal assemblages and may be sufficient to meet program objectives
concerning detection of community responses to anthropogenic disturbance. Warwick
(1986) contends that detection of anthropogenic disturbances would be facilitated
since these perturbations impact community structure at taxonomic levels higher than
the species level. Natural disturbances, which generally affect infauna community
structure by species replacement, would have little influence on analyses at higher
taxonomic levels.
However, identification to higher taxonomic levels may sacrifice the potential wealth
of information available using species-level identification. Furthermore, it is gener-
ally regarded that species is the taxon most susceptible to environmental stress (U.S.
EPA, 1987a).
The level of taxonomic identification will depend on monitoring program objectives,
sample size, study sites, and analytical measures. Data based on lower taxonomic
levels can be grouped for future comparisons with higher level taxa. It is strongly
recommended that all samples identified only to higher taxonomic levels be properly
archived, since comparisons of lower taxonomic data may be required later.
Selection of Sampling Period
Benthic infauna assemblages are dynamic; the most common temporal patterns
observed in benthic communities are those associated with seasonal changes. Seasonal
variation in benthic assemblages may be due to changes in physical, chemical, and/or
biological parameters: i.e. temperature, light transmissivity, dissolved oxygen, preda-
tion, recruitment, etc.
-------
B-118
Given the seasonal variation characteristic of benthic assemblages in general, it is
recommended that direct comparisons between samples collected during different
seasons be avoided. Studies investigating interannual variation in the characteristics of
benthic communities should be conducted during the same season (preferably the same
month) each year.
There are a variety of approaches to assess the effects of anthropogenic perturbance on
benthic communities of the estuary. Many assessment approaches are derived from
modifications of Connell's (1978) Intermediate Disturbance Hypothesis. Some
researchers have proposed that selected community structure measures e.g. species
composition, abundance, etc. are predictable along a gradient of environmental
disturbance (Pearson and Rosenberg, 1978; Gray and Mirza, 1979; Warwick, 1986).
A corollary to this supposition is that indicator species may be identified, whose
responses would epitomize community responses to habitat perturbations.
These assessment approaches may be grouped into three categories:
biological indices
indicator species
multivariate analyses
However, there has been little consensus among biologists regarding the suitability of
various techniques for describing community characteristics and/or for assessing
estuarine impacts. A critical evaluation of the use of biological indices to detect
environmental change is presented in an EPA Technical Support Document (U.S.
EPA, 1985; Table B7-1). In addition to measures of changes in the abundances of
pollution sensitive, pollution tolerant, and opportunistic species, the indices shown in
Table B7-1 are evaluated on the basis of the following criteria:
biological meaning
ease of interpretation
sensitivity to community changes due to anthropogenic sources
The results of these evaluations and additional information on other analytical methods
are summarized below.
B7.3 Existing
Analytical
Methods
-------
1 -8',
B-119
TABLE B7-1. BIOLOGICAL INDICES
Index/Method
Biological
Characteristic Measured
Recommended
forMonitoringa
Biological integrity
Bray-Curtis
Dominance1*
Infaunal index
No. individuals
No, species
Community structure
Dissimilarity
Community structure
Community structure
Total abundance
Total laxa
Opportunistic and Community structure
pollution tolerant species
Pollution-sensitive
species
Biomass
Margalefs SR
Pielou's J
Shannon-Wiener H'
Community structure
Standing crop
Diversity
Evenness
Diversity
B
P,B,F
Bc
P,BJF
P,B
P,B,F
P.B.F'
P,B,Fd
ap plankton), B (benthos), andF
-------
B-1201
Biological Indices
The numbers of individuals and the numbers of species have been found to be good
indicators of anthropogenic disturbance, as well as of other environmental stresses
(Table B7-1; Pearson and Rosenberg, 1978; Warwick, 1986). Furthermore, these
simple biological indices are less ambiguous and are often as informative as diversity
indices (U.S. EPA, 1985; Green, 1979; Hurlbert, 1971). Measures of biomass have
inherent problems in the collection of the data e.g. loss or gain of weight due to
preservative medium, drying times, evaporative weight loss, etc. However, biomass,
used in conjunction with species composition and abundance data, may demonstrate the
relative effects of enrichment on benthic communities (Pearson and Rosenberg, 1978).
More complicated indices - e.g., biological integrity, species diversity, species richness,
dominance, evenness - have found varying degrees of acceptance. Indices of diversity
and biological integrity often have the following limitations (Green, 1979):
often lack biological meaning
are not robust empirical indicators of any important correlates of "estuarine
health"
do not incorporate information of form and function of resident species
are susceptible to biases associated with well-described taxa
However, these indices are a widely used measure of community structure.
The dominance index is a measure of the degree to which one or a few species domi-
nate the community. The dominance index, herein defined as the minimum number of
species required to account for 75 percent of the total number of individuals, has been
useful in describing community structure (Swartz et al., 1985). It is easily calculated,
does not assume an underlying distribution of individuals among species, and is statisti-
cally testable.
Several graphical analyses of these simple biological indices have been developed;
most require further study before their effectiveness is known. For example, the use of
Gray and Mirza's (1979) log-normal distribution of individuals per species and
Warwick's (1985) abundance-biomass comparisons (ABC) methods, though useful for
some habitats, have been found to be unpredictable for other habitats and other locales
(U.S. EPA, 1989; Beukema, 1988).
-------
B-121
Pearson and Rosenberg's (1979) species-abundance-biomass (SAB) model appears to
be applicable to several forms of environmental disturbance and may be used in
conjunction with other analyses to assess spatial and temporal impacts due to distur-
bance (Figure B7-1). The SAB model provides a means to assess the condition of the
habitat based on species numbers, total abundance, and total biomass (Figure B5-1).
Severely polluted areas are identified by the lack of organisms; low species numbers,
total abundance, and biomass. Highly polluted areas are identified by a few, highly
abundant, populations of small-bodied, pollution-tolerant, opportunistic species (e.g.,
Capitella capitata); total abundance is high, although biomass and species numbers
remain low. Less polluted areas (e.g., further from a point source, or over time follow-
ing the cessation of organic input) may be identified by higher species numbers and
biomass, as more competitive larger-bodied species are able to tolerate sediment
conditions. More pristine areas are characterized by a number of small populations of
Figure B7-1.
Generalized SAB diagram
of changes along a
gradient of organic
enrichment (from Pearson
andRosenburg, 1978).
Increasing organic input
PO = peak of opportunists
E = ecotone point
TR = transition zone
S = species numbers
B = total biomass
A = total abundance
-------
B-122
competitively dominant species; species numbers remain elevated, although biomass
and total abundance are low.
Indicator Species
Examination of abundances of individual indicator species are generally informative
and may reduce the cost of the analysis. The absence of pollution sensitive species
and the enhancement of opportunistic and pollution tolerant species may assist in
defining the spatial and temporal extent and magnitude of impacts. However, indica-
tor variables must possess the following characteristics (Green, 1984):
provide a sufficiently precise and accurate appraisals of:
- species of concern
- anthropogenic disturbances to benthic communities
- presence/absence or the magnitude of anthropogenic perturbance
to the estuarine ecosystem
cost-effective and a statistically reliable alternative to monitoring all
critical benthic community measures of habitat perturbance
appropriate for the spatial and temporal scale demanded by the study
objectives
Pearson and Rosenberg (1978) demonstrated that along a gradient of organic enrich-
ment, a predictable community of benthic infauna could be observed (Figure B7-2);
communities consist of opportunistic, tolerant species in areas of severe pollution,
giving way to less tolerant and more competitively dominant species further from
severely polluted areas (Figure B7-2). This model of indicator species composition
and distribution been found to be useful in assessing both natural and anthropogenic
disturbances.
Several indices, derived from the distributions of pollution sensitive species and
opportunistic disturbance-tolerant species, have been developed in order to evaluate
impacts to infaunal benthic community structures: nematode/copepod ratio (Raffaelli
and Mason, 1981), infaunal index (Word, 1978), organism-sediment index (Rhoads
and Germano, 1986). Polluted or disturbed conditions may be indicated by:
high values of the nematode/copepod ratio
low values (< 65) of the infaunal index
-------
B-123
Figure B7-2. Diagram of
changes in fauna and
sediment structure along a
gradient of organic
enrichment (from Pearson
and Rosenberg, 1978).
low values (< +7) of the organism-sediment index
However, further studies of the response patterns of these indicator indices to anthro-
pogenic perturbations are required in order to select appropriate indicators of benthic
community impact.
It is recommended that multiple pollution-tolerant and pollution-sensitive indicator
species are selected by clearly defined criteria. Multiple indicator species often
provide a more complete representation of environmental conditions.
Multivariate Analyses
Numerical classification encompasses a wide variety of techniques that have been used
in the analysis of benthic infauna data to distinguish groups of entities (e.g., sample
locations) according to similarity of attributes (e.g., species). These techniques differ
from most multivariate methods in that no assumptions are made concerning the
underlying distributions of the variables. Detailed descriptions of numerical classifica-
tion analysis can be found in Romesburg (1984), Clifford and Stephenson (1975),
Boesch (1977), Sneath and Sokal (1973), and Anderberg (1973). Boesch (1977) is
-------
B-124
..: ,; .:;
particularly valuable as an introduction and guide to the use of numerical classification
analysis in marine environmental studies. Guidance on the interpretation of classifica-
tion results is provided in an EPA Technical Support Document (Table B7-1; U.S.
EPA, 1988).
Ordination analyses have also been used to reduce the dimensionality (i.e., the number
of variables) of the data set while maintaining the relationship among similar and
dissimilar entities. At present no single ordination technique has been shown to be
clearly superior for the analysis of biological data (U.S. EPA, 1985).
Multivariate analyses are effective heuristic tools. They generate visual representa-
tions which often indicate where further analyses ought to be conducted.
Analytical Approach Recommendations
Some of the most informative measures of community structure are the simplest
(Table B7-1):
number of individuals
number of species
dominance
infaunal index
abundances of pollution sensitive species
abundances of opportunistic and pollution tolerant species
These indices have proved to be useful over various habitats and regions in assessing
changes to benthic community structures (U.S. EPA, 1985). Values for these indices
may be determined from the list of species abundances generated during the taxonomic
identifications of collected specimens. Furthermore, the values of these six variables
may be easily tested statistically using parametric or nonparametric techniques. It is
recommended that no single index or analytical method be used to assess impacts;
rather the assessment of impacts should incorporate information that each variable and
method contributes concerning benthic community structure.
-------
B-125
11 ^ f "-.<>.' t, ,*> v .. ' - , - o f, ซ ปป S^X:,-!
* ' " * s . ^ "^ 5
f
* N ' - . ^ -
. ^
Selection of Reference Sites
Selection of reference sites is key to the evaluation of environmental impact assess-
ment. Monitoring of reference sites must provide measures of natural variability
against which measures of anthropogenic impacts are detected. Thus, reference sites
must be similar in every way with sampling sites of the monitoring program except
that it is unaffected by anthropogenic disturbances of concern. Results of analyses
using reference measures provide the means of discerning variability due to anthropo-
genic sources from natural variability of the ecosystem. Characteristics which should
be controlled for include:
sediment characteristics (i.e. grain size)
water depths
flow characteristics
compared to monitoring program sampling sites. Several reference sites may be
required in order to meet these criteria.
B7A QA/QC Sample Collection
Considerations
Sampling device should be inspected for wear and tear leading to sample leakage upon
ascent. It is prudent to have a backup sampler onboard the survey vessel in case the
primary sampler is found to be unsuitable during the cruise.
The following sample acceptability criteria should be satisfied (U.S. EPA, 1987a):
sediment is not extruded from the upper face of the sampler such that
organisms by have been lost
overlying water is present, indicating minimal leakage
the sediment surface is relatively flat indicating minimal disturbance or
winnowing (Figure B7-3)
the entire surface of the sample is included in the sampler
the desired penetration depth is achieved
-------
B-1261
Acceptable if Minimum
Penetration Requirement Met
and Overlying Water is Present
Unacceptable
(Washed, Rock Caught in Jaws)
Unacceptable
(Canted with Partial Sample)
Unacceptable
(Washed)
Figure B7-3.
Examples of acceptable
and unacceptable samples
(U.S. EPA, 1987a).
If the sample does not meet all the criteria, the sample should be rejected.
Taxonomic Identification
A key QA/QC issue is taxonomic standardization. Consistent taxonomic identifica-
tions are achieved through interaction among taxonomists working on each major
group. Participation of the laboratory staff in regional taxonomic standardization
programs is recommended to ensure regional consistency and accuracy of identifica-
tions.
Five percent of all samples identified by one taxonomist should be re-identified by
another taxonomist who is also qualified to identify organisms in that major group. It
is advisable that at least three individuals of each taxon should be sent for verification
to recognized experts. These verified specimens should then be placed in a permanent
reference collection. All specimens in the reference collection should be stored in
labeled vials which are segregated by species and sample. Reference specimens
should be archived alphabetically within major taxonomic groups.
-------
B-127
^m
*
/
It is also recommended that at least 20 percent of each sample be re-sorted for QA/QC
purposes. Re-sorting is the examination of a sample or subsample that has been sorted
once and is considered free of organisms. Re-sorting should be done by someone
other than the one who sorted the original sample. If a sample is found that does not
meet the recommended 95 percent removal criterion, the entire sample should be re-
sorted.
B7.5 Statistical Consideration of statistical strategies will mitigate the high costs of collecting and
Design processing samples. See also Statistical Design Considerations: Power Analysis and
Considerations Power Cost Analysis (Appendix B Introduction; Section B.3).
Temporal Stratification of the Data
The time of the year should be controlled or stratified in the design the use of
annual averages is seldom good practice. Temporal stratification of the data should
not be attempted until sufficient knowledge of long-term natural cycles is attained.
Initially, simple regression analyses may be conducted on seasonally stratified data in
order to identify monotonic temporal trends. Further examinations of whether condi-
tions are improving or degrading over time may be examined using various statistical
time series analyses.
B7.6 Monitoring of benthic community structure provides in situ measures of the benthic
Use Of Data habitat and remains a powerful tool in the evaluation of spatial and temporal effects of
anthropogenic and natural disturbance. The presence or absence of certain infaunal
organisms are useful in indicating the previous condition of the environment. Moni-
toring of benthic infaunal communities also provides data required in the design and
validation of benthic community dynamics models (Pearson and Rosenberg, 1978) and
the selection of biological indicators (Word, 1978).
In addition, monitoring of benthic communities directly provides accurate information
essential in assessing the effectiveness of pollution abatement programs (Bilyard,
1987). For example, benthic infauna monitoring provided information used to assess
the effectiveness of pollution abatement plans in the recovery of Southern California
waters (Reish, 1986; SCCWRP, 1988). These studies indicate that analyses of benthic
infaunal communities may be effectively used to monitor long-term recovery of the
receiving environment (Reish, 1986).
-------
B-1281
Currently, benthic communities have not proved useful for identifying specific chemi-
cals or classes of chemicals present in the environment. Further information concern-
ing specific responses to specific contaminants is required before infaunal community
structure becomes useful in identifying sources of contaminants (U.S. EPA, 1989). In
addition, the use of benthic community structure to predict specific effects on potential
predators is cautioned. Information on trophic relationships, competition, and preda-
tion is often not available. The capability to predict the effects of altered prey commu-
nities on predators may improve with research on these topics. Factors such as prey
quality, distribution of prey, and interactions among species will be important compo-
nents of this research.
However, benthic invertebrates serve as effective indicators of environmental condi-
tion; delineating the magnitude, spatial extent, and temporal trends of anthropogenic
and natural perturbations to the ecosystem (Reish, 1986; Bilyard, 1987). Monitoring
of benthic infauna will provide relevant accurate data fundamental to achieving the
objectives of most estuarine monitoring programs.
Rationale B7.7
Summary and
The objective is to detect and describe spatial and temporal changes in the Recommendations
structure and function of benthic communities
Provides in situ measures of habitat quality and is a powerful tool in
assessing environmental impact
Monitoring Design Considerations
It is recommended that consistent types of sampling gear, sample sorting
protocols, level of taxonomy, and location and timing of sample collection
be implemented to allow for comparisons among studies
- Collection of undisturbed sediment requires that the sampler:
- create a minimal bow wake when descending
- form a leakproof seal when the sediment sample is taken
- prevent winnowing and excessive sample disturbance when
ascending
- allow easy access to the sample surface in order that undisturbed
subsamples may be taken
-------
B-129
Penetration well below the desired sampling depth is preferred to prevent
sample disturbance as the device closes.
Grab samplers and box corers are recognized as the tools of choice for
maximum accuracy and precision when sampling soft bottom habitats
Sorting through a standard sieve mesh size is recommended - further
sorting through other mesh sizes may be conducted in addition to sorting
through this standard mesh size
Relaxants facilitate identification and morphometric measurements,
however standard procedures must be implemented to ensure comparisons
among studies
Vital stains may facilitate sorting, however a proper QA program should
ensure that sorting efficiency is maintained
Identifications to higher taxonomic levels may be sufficient to meet
program objectives, however it is recommended that all samples be
archived if comparisons to lower taxonomic levels are required at a later
date
In order to reduce the variation due to seasonal differences, sampling
should be conducted during the same season - preferably the same month -
each year
Existing Analytical Methods
It is recommended that simple measures of community structure be used to
assess the condition of the estuarine benthos: number of individuals,
number of species, dominance, infaunal index, abundance of pollution
sensitive species, and abundance of pollution tolerant species (Table B3-1)
Selected biological indices should retain biological meaning, be robust
indicators of estuarine "health", and incorporate species form and function
Indicator species should possess the following characteristics:
- sensitive to benthic perturbances of concern,
- cost-effective and statistically reliable alternative to measuring all
species in a monitoring program
-------
B-130
- statistically reliable indicative measures of habitat perturbance
- appropriate for the spatial and temporal scale demanded by the
study objectives
Selection of reference sites is key to the evaluation of environmental
impact due to anthropogenic perturbances; several reference sites may be
required to proper control for sampling sites
QA/QC Considerations
Taxonomic standardization is key to the analysis of community structure -
recommended protocols include consistent interactions among taxono-
mists, re-identification of selected samples, use of a reference collection,
and re-sorting of selected samples or subsamples
Statistical Design Considerations
Power analyses may be applied to determine the appropriate number of
sample replicates required to detect a specified difference, thereby optimiz*
ing the high costs of collecting and processing samples
Use of Data
Provide essential information in order to assess impacts due to anthropo-
genic perturbance, monitor recovery of the receiving environment, and
validate community and population models
Amjad, S., and J.S. Gray. 1983. Useofthenematode/copepodratioas an index of
organic pollution. Mar. Poll. Bull. 14:178-181.
Anderberg, M.R. (1973). Ouster Analysis for Applications. New York, NY: Aca-
demic Press. 359 pp.
Becker, D.S. and J.W. Armstrong. 1988. Development of regionally standardized
protocols for marine environmental studies. Marine Pollution Bulletin. 19(7)310-313.
Bernstein, B.B., and R. W. Smith, 1986. Community approaches to monitoring.
IEEE Conference Proceedings, Oceans '86 pp 934-939.
B7.8 Literature
Cited and
References
-------
B-131
Beukema, J.J. 1988. An evaluation of the ABC method (abundance-biomass compari-
son) as applied to macrozoobenthic communities living on tidal flats in the Dutch
Wadden Sea. Mar. Biol. 99:425-433.
Bilyard, G.R. 1987. The value of benthic infauna in marine pollution monitoring
studies. Mar. Poll. Bull. 18:581-585.
Boesch, D.F. 1977. Application of numerical classification in ecological investiga-
tions of water pollution. EPA 600/3-77-033. Office of Research and Development,
US Environmental Protection Agency. Corvallis, OR. 115pp.
Clifford, H.T. and W. Stephenson. 1975. An introduction to numerical classification.
New York, NY: Academic Press. 229 pp.
Connell, J.H. 1978. Diversity in tropical rainforests and coral reefs. Science
199:1302-1310.
Downing, J. A. 1979. Aggregation, transformation, and the design of benthos sam-
pling programs. J. Fish. Res. Board Can. 36:1454-1463.
Eleftheriou, A. and N.A. Holme. 1984. Macrofauna techniques. In: Methods for the
Study of Marine Benthos. IBP Handbook No. 16. London, England: Blackwell
Scientific, pp. 140-216.
Elliot, J.M. 1971. Some methods for the statistical analysis of samples of benthic
invertebrates. Scientific Publication No. 25, Freshwater Biological Assn., Ferry
House, U.K. 148pp.
Ellis, D. 1985. Taxonomic sufficiency in pollution assessment. Mar. Poll. Bull.
16:459.
Elmgren, R., S. Hansson, U. Larsson, B. Sundelin, and P.D. Boehm. 1983. The
"Tsesis" oil spill: Acute and long-term impact on the benthos. Mar. Poll. Bull.
15:249-253.
Ferraro, S.P., F.A. Cole, W.A. DeBen, and R.C. Swartz. 1989. Power-cost efficiency
of eight macrobenthic sampling schemes in Puget Sound, Washington, USA. Caa J.
Fish. Aquat. Sci. 46:2157-2165.
-------
B-132
Fredette, T.J., D.A. Nelson, T. Miller-Way, J.A. Adair, V.A. Sotler, J.E. Clausner,
E.B. Hands, and F.J. Anders. 1989. Selected tools and techniques for physical and
biological monitoring of aquatic dredged material disposal sites. Final Report. U.S.
Army Engineer Waterways Experiment Station, Vicksburg, MS.
Gauch, H.G. 1982. Multivariate Analysis in Community Ecology. Cambridge, UK:
Cambridge University Press. 298 pp.
Gray, J.S., and F.B. Mirza. 1979. A possible method for the detection of pollution
induced disturbance on marine benthic communities. Mar. Poll. Bull. 10:142-146.
Grizzle, R.E. 1984. Pollution indicator species of macrobenthos in a coastal lagoon.
Mar. Ecol. Prog. Ser. 18:191-200.
Green, R.H. 1979. Sampling Design and Statistical Methods for Environmental
Biologists. John Wiley & Sons, NY. 257pp.
Green, R.H. 1984. Statistical and nonstatistical considerations for environmental
monitoring studies. Environ. Monitoring and Assessment 4:293-301.
Hurlbert, S.H. 1971. The nonconcept of species diversity: A critique and alternative
parameters. Ecol. 52:577-586.
Hurlbert, S.H. 1984. Pseudoreplication and the design of ecological field experi-
ments. Ecol. Monogr. 54:187-211.
Jackson, J.B.C., J.D. Cubit, B.D. Keller, V. Batista, K. Bums, H.M. Caffey, R.L.
Caldwell, S.D. Garrity, C.D. Getter, C. Gonzales, H.M. Guzman, K.W. Daufman, A.H.
Knapp, S.C. Levings, M.J. Marshall, R. Steger, R.C. Thompson, and E. Weil. 1989.
Ecological effects of a major oil spill on Panamanian coastal marine communities.
Sci. 243:37-44.
Lambshead, P.J. 1984. The nematode/copepod ratio, some anomalous results from
the Firth of Clyde. Mar. Poll. Bull. 15:256-259.
Lambshead, P.J. and H.M. Platt. 1985. Structural patterns of marine benthic assem-
blages and their relationship with empirical statistical models. Proc. 19th Eur. Mar.
Biol. Symp. pp. 371-380.
-------
B-133
Livingston, R.J., R.S. Lloyd, and M.S. Zimmerman. 1976. Determination of sampling
strategy for benthic macrophytes in poUuted and unpolluted coastal areas. Bull. Mar.
Sci. 26:569-575.
Lunz, J.D., and D.R. Kendall. 1982. Benthic resource analysis technique, a method
for quantifying the effects of benthic community changes on fish resources. In:
Conference Proceedings on Marine Pollution. Oceans 1982. National Oceanic and
Atmospheric Administration, Office of Marine Pollution Assessment. Rockville, MD.
pp. 1021-1027.
Mclntyre, A.D., J.M. Elliot, and D.V. Ellis. 1984. Introduction: design of sampling
programs. IBP Handbook No. 16. In: Methods for the Study of Marine Benthos.
(N.A. Holme and A.D. Mclntyre, eds.). Oxford: Blackwell Scientific Publications.
pp. 1-26.
NRC. 1990. Managing Troubled Waters: The role of Marine Environmental Moni-
toring. National Research Council. Washington, D.C: National Academy Press.
125 pp.
Pearson, T.H., and R. Rosenberg. 1978. Macrobenthic succession in relation to
organic enrichment and pollution of the marine environment. Oceangr. Mar. Biol.
Ann. Rev. 16:229-311.
Raffaelli, D. 1987. The behavior of the nematode/copepod ratio in organic pollution
studies. Mar. Environ. Res. 23:135-152.
Raffaelli, D. and C.F. Mason. 1981. Pollution monitoring with meiofauna, using the
ratio of nematodes to copepods. Mar. Poll. Bull. 12:158-163.
Rees, H.L. 1984. A note on mesh selection and sampling efficiency i benthic studies.
Mar. Pollut. Bull. 15:225-229.
Reish, D.J. 1959. A discussion of the importance of screen size in washing quantita-
tive marine bottom samples. Ecology 40:307-309.
Reish, D.J. 1986. Benthic invertebrates as indicators of marine pollution: 35 years of
study. IEEE Conference Proceedings, Oceans '86 pp 885-888.
Rhoads, D.C., and D.K. Young. 1970. The influence of deposit-feeding organisms on
sediment stability and community trophic structure. J. Mar. Res. 28:150-178.
-------
B-1341
Rhoads, D.C., and J.D. Germane. 1982. Interpreting long-term changes in benthic
community structure: A new protocol. Hydrobiologia 142:291-308.
Rhoads, D.C., and J.D. Germane. 1986. Characterization of organism-sediment
relations using sediment profile imaging: An efficient method of remote ecological
monitoring of the sea floor (REMOTS system). Mar. Ecol. prog. Ser. 8:115-128.
Romesburg, H.C. 1984. Cluster Analysis for Researchers. Belmont, CA: Lifetime
Learning Publications. 334 pp.
Rygg, B. 1985. Distribution of species along pollution-induced diversity gradients in
benthic communities in Norwegian fjords. Mar. Poll. Bull. 12:469-474.
SCCWRP. 1988. Recovery of Santa Monica Bay after termination of sludge dis-
charge. 1988-1989 Annual Report, Southern California Coastal Water Research
Project, pp 46-53.
Self, S.G., and R.H. Mauritsen. 1988. Power/sample size calculations for generalized
linear models. Biometrics. 44:79-86.
Sheills, G.M., and KJ. Anderson. 1985. Pollution monitoring using the nematode/
copepod ratio, a practical application. Mar. Poll. Bull. 16:62-68.
Sneath, P.H.A. and R.R. Sokal. 1973. Numerical Taxonomy: The Principles and
Practices of Numerical Classification. San Francisco, CA: Freeman. 573 pp.
Swartz, R.C., D.W. Schultz, G.R. Ditsworth, W.A. DeBen, and F.A. Cole. 1985.
Sediment toxicity, contamination, and macrobenthic communities near a large sewage
outfall. In: Validation and Predictability of Laboratory Methods for Assessing the Fate
and Effects of Contaminants in Aquatic Ecosystems. (T.T. Boyle, ed.). Philadelphia,
PA: American Society for Testing and Materials (ASTM). pp. 152-175.
U.S. EPA. 1985. Recommended biological indices for 301(h) monitoring programs.
EPA 430/9-86-002. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1987a. Puget Sound protocols. Final Report. 31pp. Prepared for
Region X, Office of Puget Sound.
-------
B-135
U.S. EPA. 1987b. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1988. ODES data brief: Use of numerical classification. Prepared for the
Office of Marine and Estuarine Protection. 18 pp.
U.S. EPA. 1989. Sediment classification methods compendium. Prepared for the
Office of Water Regulations and Standards.
U.S. EPA. 1990. Environmental Monitoring and Assessment Program: Ecological
Indicators. EPA 600/3-90-060. Office of Research and Development, Washington,
DC.
Warwick, R.M. 1985. A new method for detecting pollution effects on marine
macrobenthic communities. Mar. Biol. 92:557-562.
Warwick, R.M. 1986. The level of taxonomic discrimination required to detect
pollution effects on marine benthic communities. Mar. Poll. Bull. 19:259-268.
Word.J.Q. 1978. The infaunal trophic index. 1978 Annual Report, Southern Califor-
nia Coastal Water Research Project, pp 19-39.
Word, J.Q., B.L. Myers, and AJ. Mearns. 1977. Animals that are indicators of
marine pollution. 1977 Annual Report, Southern California Coastal Water Research
Project, pp 199-207.
-------
B-1361
f >^*^-. ,......<... \\ xx-sV^w-s v.X*,'-*'- v,\ .s-MKv.w. w*'* *. sw.sv. .frv^s,
s ฐ ff f f * s* v. s " s v v. s\ss.sV ss ss s s -.
". f "% %X% V-fe V. ift SV-'k'" "S * ^ป\ VS ^N*- '',' s " '>"'" " "
* sซ.s.s jf> * tl! . "''X'*'"' ^?#-js*iS ซ^ ;, %\ \ %?s v AS*. .N''. xซs ป'ป > i * ^-t ** ^-is *-^^*-t^ ss !{ O **\ s s^ ^ s " "
ฃ> V 6- .Mฃo4' N s-. *v V% vN^^^SS>^V&sV^&^\^^^^'NSป^ Xv%%d^.*K*
-------
B-137
B8.0 Fish Community Structure
B8.1 Rationale Fish are important economic, recreational, and aesthetic components of the estuarine
ecosystem. In order to protect and preserve healthy fish stocks, good estimates offish
population abundances, detailed knowledge of fish life histories, and information on
the effects of natural and anthropogenic disturbance are required.
The objective of monitoring fish populations is to detect and describe spatial and
temporal changes in the structure and function offish communities. These results can
be used to assess the condition - the "health" - of estuarine habitats, to monitor rates
of recovery following environmental interventions, and to provide an early warning of
potential impacts to recreational and commercial fisheries. In addition, demersal fish
are good indicators of benthic conditions since:
they are generally sedentary - observed effects are in response to local
environmental conditions
they are relatively sensitive to habitat disturbance - i.e. communities
undergo changes in abundance in response to environmental perturbations
they often mediate the transfer of nutrients and toxic substances in the
ecosystem as important predator and prey organisms
Monitoring fish population abundances is one of a handful of methods which provides
in situ measures of biotic health and habitat quality. The assessment of fish popula-
tions remains a powerful tool in the evaluation of spatial and temporal effects of
anthropogenic and natural disturbance.
B8.2 Monitoring Collection, analyses, and evaluations of fish community structure and function are
Design typically time-consuming, labor-intensive, and expensive tasks. A survey vessel
Considerations manned by an experienced crew and specially equipped with gear to collect organisms
is required. Expert taxonomists are needed to identify and enumerate collected fish
specimens.
The results offish monitoring programs can vary substantially depending on the
objectives and corresponding design specifications. The characteristics that are
primarily responsible for the variability in the results are:
type of sampling gear
volume sampled
-------
B-138H
' ~ ~.^~~, ~^~^
"V ->ซ!
location and timing of sample collection
It is essential to understand the effects of these monitoring design characteristics on the
results and to standardize them as much as possible to ensure the comparability of
sampled data throughout the estuary.
Analyses of power-cost efficiencies are useful in selecting the appropriate sampling
gear and sample processing protocols. Ferraro et al. (1989) provide an example of
power-cost analyses.
Sampling Devices
Sample collection protocols influence all subsequent laboratory and data analysis; it is
key that fish population samples be collected using acceptable and standardized
techniques. Several types of devices can be used to collect fish samples: traps and
cages, passive nets, trawls (active nets), and photographic surveys (Fredette et al.,
1989). Many of these devices selectively sample specific types offish. Accordingly,
conducting comparisons among data collected using different devices is inadvisable.
Traps and Cages - Traps and cages are usually designed to attract and capture
specific organisms. They are useful in studies examining the activities of a particular
target organism in a given area. Traps and cages provide only qualitative measures of
organisms in a particular area.
Passive Nets and Trawls - Nets are highly selective in the species that are captured
and in the efficiency of retaining captured specimens. The size of the net, its configu-
ration and orientation, and the avoidance behavior of the target species should be
considered when using any net. The mesh size effects the speed at which the net can
be towed. The slower the towing speed, the more likely that some organisms will
either avoid or escape the net. It is highly recommended that the duration, direction,
and speed of towing be set in order to compare trawls.
Passive nets are deployed at a fixed position; organisms become entangled or trapped
within the netted area. Passive nets are used to collect selected target species and
provide a qualitative means of sampling fish populations. Limitations associated with
passive nets include:
nets, ordinarily, must remain in-place for an extended period of time
deployment and recovery of nets are typically time-consuming processes
-------
B-139
Trawls (active nets) are drawn through the water and results are more immediate
compared to passive nets. Trawls are typically used to collect large quantities of fish
at various depths (Fredette et al., 1989). Otter trawls are less selective than passive
nets, capturing a wider variety offish and epibenthic macroinvertebrates.
Photographic Surveys - Photographic surveys are effective when the bottom topogra-
phy is uneven or where trawling is not possible. However, the utility of photographic
surveys is limited by water clarity, difficulties identifying species, and fish avoidance
to the camera system. In addition, further studies comparing photographic surveys to
trawls are required before comparisons may be made. This method is also limited by
the qualitative nature of the data.
Volume Sampled
Different species of fish have different scales of horizontal and vertical spatial distri-
bution (Gushing, 1975; Bond, 1979). Costs of laboratory analyses of the sample
increase with increased volume sampled. Analyses of spatial and temporal scale,
statistical power, as well as costs, will assist in determining optimal sample volume. It
is highly recommended that a standard sample volume (same tow duration, tow speed,
and net opening area) be analyzed in order to ensure data comparability.
Selection of Sampling Period
Fish assemblages are dynamic; the most common temporal patterns observed in fish
communities are those associated with seasonal changes. Seasonal variation in fish
assemblages may be due to changes in physical, chemical, and/or biological param-
eters: i.e. temperature, light transmissivity, dissolved oxygen, predation, recruitment,
etc.
Given the seasonal variation characteristic of fish assemblages in general, it is recom-
mended that direct comparisons between samples collected during different seasons be
avoided. Studies investigating interannual variation in the characteristics of fish
communities should be conducted during the same season (preferably the same month)
each year.
B8.3 Analytical There are a variety of approaches to assess the effects of anthropogenic perturbance on
Methods fish communities of the estuary. These assessment approaches may be grouped into
Considerations three categories:
-------
B-1401
biological indices
indicator species
multivariate analyses
However, there has been little consensus among biologists regarding the suitability of
various techniques for describing community characteristics and/or for assessing
estuarine impacts. A critical evaluation of the use of biological indices to detect
environmental change is presented in an EPA Technical Support Document (EPA,
1985; Table B8-1). In addition to measures of changes in the abundances of pollution
sensitive, pollution tolerant, and opportunistic species, the indices shown in
Table B8-1 are evaluated on the basis of the following criteria:
biological meaning
ซ ease of interpretation
sensitivity to community changes due to anthropogenic sources
The results of these evaluations and additional information on other analytical methods
are summarized below.
Biological Indices
The numbers of individuals and the numbers of species have been found to be good
indicators of anthropogenic disturbance, as well as of other environmental stresses
(U.S. EPA, 1985). Furthermore, these simple biological indices are less ambiguous
and are often as informative as diversity indices (U.S. EPA, 1985; Green, 1979;
Hurlbert, 1971). Measures of biomass have inherent problems in the collection of the
data e.g., loss or gain of weight due to preservative medium, drying times, evapora-
tive weight loss.
More complicated indices - e.g., species diversity, species richness, dominance,
evenness have found varying degrees of acceptance. Diversity indices, measures of
the distribution of individuals among species, have the following limitations (Green,
1984):
often lack biological meaning
-------
B-141
fli'PiMi^f^ |||
i|;
-------
B-142
are not robust empirical indicators of any important correlates of "estuarine
health"
do not incorporate information of form and function of resident species
are susceptible to biases associated with well-described taxa
However, species diversity indices are a widely used measure of community structure.
The dominance index is a measure of the degree to which one or a few species domi-
nate the community. The dominance index, herein defined as the minimum number of
species required to account for 75 percent of the total number of individuals, has been
useful in describing community structure (Swartz et al., 1985). It is easily calculated,
does not assume an underlying distribution of individuals among species, and is
statistically testable.
Indicator Species
Examination of abundances of individual indicator species are generally informative
and may reduce the cost of the analysis. The absence of pollution sensitive species
and the enhancement of opportunistic and pollution tolerant species may assist in
defining the spatial and temporal extent and magnitude of impacts. However, indica-
tor variables must possess the following characteristics (Green, 1984):
provide a sufficiently precise and accurate appraisals of:
- specie(s) of concern
- anthropogenic disturbances to benthic communities
- presence/absence or the magnitude of anthropogenic perturbance
to the estuarine ecosystem
cost-effective and a statistically reliable alternative to monitoring all
critical benthic community measures of habitat perturbance
appropriate for the spatial and temporal scale demanded by the study
objectives
However, further studies of the response patterns of fish species subjected to anthropo-
genic perturbations are required in order to select appropriate indicators of environ-
mental impact.
-------
B-143
Multivariate Analyses
Numerical classification encompasses a wide variety of techniques that have been used
in the analysis offish data to distinguish groups of entities (e.g., sample locations)
according to similarity of attributes (e.g., species). These techniques differ from most
multivariate methods in that no assumptions are made concerning the underlying
distributions of the variables. Detailed descriptions of numerical classification analy-
sis can be found in Romesburg (1984), Clifford and Stephenson (1975), Boesch
(1977), Sneath and Sokal (1973), and Anderberg (1973). Boesch (1977) is particularly
valuable as an introduction and guide to the use of numerical classification analysis in
marine environmental studies. Guidance on the interpretation of classification results
is provided in an EPA Technical Support Document (Table B8-1; U.S. EPA, 1988).
Ordination analyses have also been used to reduce the dimensionality of the data set
while maintaining the relationship among similar and dissimilar entities. At present no
single ordination technique has been shown to be clearly superior for the analysis of
biological data (U.S. EPA, 1985).
Multivariate analyses are effective heuristic tools. They generate visual representa-
tions which often indicate where further analyses ought to be conducted.
Analytical Approach Recommendations
Some of the most informative measures of community structure are the simplest
(Table B8-1):
number of individuals
number of species
dominance
abundances of pollution sensitive species
abundances of opportunistic and pollution tolerant species
These indices have proved to be useful over various habitats and regions in assessing
changes to fish community structures (U.S. EPA, 1985). Values for these indices may
be determined from the list of species abundances generated during the taxonomic
-------
B-144
identifications of collected specimens. Furthermore, the values of these six variables
may be easily tested statistically using parametric or nonparametric techniques. It is
recommended that no single index or analytical method be used to assess impacts;
rather the assessment of impacts should incorporate information that each variable and
method contributes concerning benthic community structure.
Selection of reference sites is key to the evaluation of environmental impact assess-
ment.
Results of analyses using reference measures provide the means of comparison to
which anthropogenic impacts are detected. It is essential that selected reference sites
exhibit similar:
bottom characteristics (i.e., bottom topography, sediment grain size,
percent cover)
water depths
flow characteristics
compared to monitoring program sampling sites. Several reference sites may be
required in order to meet these criteria.
Sample Collection
The primary consideration for an adequate sampler is that it consistently collect
samples. Nets should be inspected for wear and tear leading to sample loss. It is
prudent to have backup gear on board the survey vessel in case the primary gear is
found to be unsuitable during the cruise.
Taxonomic Identification
A key QA/QC issue is taxonomic standardization. Consistent taxonomic identifica-
tions are achieved through interaction among taxonomists working on each major
group. Participation of the laboratory staff in regional taxonomic standardization
programs is recommended to ensure regional consistency and accuracy of identifica-
tions.
B8.4 QA/QC
Considerations
-------
B-145
The verified specimens should then be placed in a permanent reference collection. All
specimens in the reference collection should be stored in labeled vials which are
segregated by species and sample. Reference specimens should be archived alphabeti-
cally within major taxonomic groups.
B8.5 Statistical Consideration of statistical strategies will mitigate the high costs of collecting and
Design processing samples. See also Statistical Design Considerations: Power Analysis and
Considerations Power-cost Analysis (Appendix B Introduction; Section B.3).
Temporal Stratification of the Data
The time of the year should be controlled or stratified in the design the use of
annual averages is seldom good practice. Temporal stratification of the data should
not be attempted until sufficient knowledge of long-term natural cycles is attained.
Initially, simple regression analyses may be conducted on seasonally stratified data in
order to identify monotonic temporal trends. Further examinations of whether condi-
tions are improving or degrading over time may be examined using various statistical
time series analyses (e.g., temporal autocorrelation, spectral analyses).
B8.6 Monitoring of fish community structure provides in situ measures of the estuarine
Use Of Data habitat and remains a powerful tool in the evaluation of spatial and temporal effects of
anthropogenic and natural disturbance. The presence or absence of certain fish is
useful in indicating the condition of the environment. Monitoring offish communities
also provides data required in the design and validation of fish population dynamics
models and the selection of biological indicators (U.S. EPA, 1990).
In addition, monitoring of fish communities may directly provide accurate information
essential in assessing the effectiveness of pollution abatement programs. Analyses of
fish communities may be effectively used to monitor long-term recovery of the receiv-
ing environment (Gushing, 1975).
Fish serve as effective indicators of environmental condition; delineating the magni-
tude, spatial extent, and temporal trends of anthropogenic and natural perturbations to
the ecosystem. Monitoring of fish will provide relevant accurate data fundamental to
achieving the objectives of many estuarine monitoring programs.
-------
B-1461
Rationale
The objective is to detect and describe spatial and temporal changes in the
structure and function offish communities in order to protect the economic,
recreational, and aesthetic value of the estuarine fisheries
Provides in situ measures of habitat quality and is a powerful tool in assessing
environmental impact
Monitoring Design Considerations
It is recommended that consistent types of sampling gear, volume sampled,
and location and timing of sample collection be implemented to allow for
comparisons among studies
In order to reduce the variation due to seasonal differences, sampling should
be conducted during the same season - preferably the same month - each year
Existing Analytical Methods
It is recommended that simple measures of community structure be used to
assess the condition of the estuarine fish: number of individuals, number of
species, dominance, abundance of pollution sensitive species, and abundance
of pollution tolerant species (Table B8-1)
Selected biological indices should retain biological meaning, be robust
indicators of estuarine "health", and incorporate species form and function
Indicator species should possess the following characteristics:
- sensitive to perturbances of concern
- cost-effective and statistically reliable alternative to measuring all
species in a monitoring program
- statistically reliable indicative measures of habitat perturbance
- appropriate for the spatial and temporal scale demanded by the study
objectives
Selection of reference sites is key to the evaluation of environmental impact
due to anthropogenic perturbances; several reference sites may be required to
proper control for sampling sites
B8.7
Summary and
Recommendations
-------
B-147
QA/QC Considerations
Taxonomic standardization is key to the analysis of community structure -
recommended protocols include consistent interactions among
taxonomists, re-identification of selected samples, and use of a reference
collection
Statistical Design Considerations
Power analyses may be applied to determine the appropriate number of
sample replicates required to detect a specified difference, thereby optimiz-
ing the high costs of collecting and processing samples
Use of Data
Provide essential information in order to assess impacts due to anthropo-
genic perturbance, monitor recovery of the receiving environment, and
validate community and population models
B8.8 Literature Anderberg, M.R. 1973. Cluster Analysis for Applications. New York, NY:Aca-
Cited and demic Press. 359 pp.
References
Boesch, D.F. 1977. Application of numerical classification in ecological investiga-
tions of water pollution. EPA 600/3-77-033. Office of Research and Development,
US Environmental Protection Agency. Corvallis, OR. 115pp.
Bond, C.E. 1979. Rinlngv of Fishes. Philadelphia, PA: Sanders College Publishing.
514 pp.
Clifford, H.T. and W. Stephenson. 1975. An introduction to numerical classification.
New York, NY: Academic Press. 229pp.
Cushing, DJ. 1975. Marine Ecology and Fisheries. Cambridge, U.K.: Cambridge
University Press. 278pp.
Ferraro, S.P., F.A. Cole, W.A. DeBen, and R.C. Swartz. 1989. Power-cost efficiency
of eight macrobenthic sampling schemes in Puget Sound, Washington, USA. Caa J.
Fish. Aquat. Sci. 46:2157-2165.
-------
B-148
Fredette, T.J., D.A. Nelson, T. Miller-Way, J.A. Adair, V.A. Sotler, J.E. Clausner,
E.B. Hands, and F.J. Anders. 1989. Selected tools and techniques for physical and
biological monitoring of aquatic dredged material disposal sites. Final Report. U.S.
Army Engineer Waterways Experiment Station, Vicksburg, MS.
Gauch, H.G. 1982. Multivariate Analysis in Community Ecology. Cambridge, UK:
Cambridge University Press. 298 pp.
Green, R.H. 1979. Sampling Design and Statistical Methods for Environmental
Biologists. John Wiley & Sons, NY. 257pp.
Green, R.H. 1984. Statistical and nonstatistical considerations for environmental
monitoring studies. Environ. Monitoring and Assessment 4:293-301.
Hurlbert, S.H. 1971. The nonconcept of species diversity: A critique and alternative
parameters. Ecol. 52:577-586.
Hurlbert, S.H. 1984. Pseudoreplication and the design of ecological field experi-
ments. Ecol. Monogr. 54:187-211.
Romesburg, H.C. 1984. Cluster Analysis for Researchers. Belmont, CA: Lifetime
Learning Publications. 334 pp.
Self, S.G., and R.H. Mauritsen. 1988. Power/sample size calculations for generalized
linear models. Biometrics. 44: 79-86.
Sneath, P.H.A. and R.R. Sokal. 1973. Numerical Taxonomy: The Principles and
Practices of Numerical Classification. San Francisco, CA: Freeman. 573 pp.
Swartz, R.C., D.W. Schultz, G.R. Ditsworth, W.A. DeBen, and F.A. Cole. 1985.
Sediment toxicity, contamination, and macrobenthic communities near a large sewage
outfall. In: Validation and Predictability of Laboratory Methods for Assessing the Fate
and Effects of Contaminants in Aquatic Ecosystems. (T.T. Boyle, ed.). Philadelphia,
PA: American Society for Testing and Materials (ASTM). pp. 152-175.
U.S. EPA. 1978. Use of small otter trawls in coastal biological surveys. EPA
600/3-78-083. Office of Research and Development, CorvaUis, OR. 35pp.
-------
B-149
U.S. EPA. 1985. Recommended biological indices for 301(h) monitoring programs.
EPA 430/9-86-002. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1987a. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1987b. Puget Sound protocols. Final Report. 31pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1988. ODES data brief: Use of numerical classification. Prepared for the
Office of Marine and Estuarine Protection. 18 pp.
U.S. EPA. 1990. Environmental Monitoring and Assessment Program: Ecological
Indicators. EPA 600/3-90-060. Office of Research and Development, Washington,
DC.
-------
B-1501
-------
B-151
B9.0 Fish and Shellfish Pathobiology
Pathobiological methods provide information concerning damage to organ systems of
fish and shellfish through an evaluation of their altered structure, activity, and func-
tion. Anatomic pathology methods can give an indication of the nature of the altered
state, for example, by identifying the specific type of tumor present in the animal.
Reproduction/development studies examining the reproductive capacity of animals
can provide information to aid in estimating and predicting population abundance and
recruitment. Biochemical/enzymology studies seek to detect differences in enzymatic
activity as a mechanism that modulates biologic activity. Immunological methods
can demonstrate altered immune response as a suggestion of changed bodily defense
mechanisms and altered susceptibility to disease.
Pathobiological methods should be used in concert with one another to investigate
cause and effect relationships in contaminant exposures. Anatomic pathology can
serve as a vital link between the adverse end effects on populations and communities
observed in an estuary and the changes in activity and function observed by other
methods.
B9.1 Rationale Pathobiological methods can be used to examine adverse effects of pollutants on fish
and shellfish. The presence of toxics in water and sediments may not immediately
result in visible changes in these organisms. Biomarkers offer a more sensitive and
reliable assessment of exposure risks than ambient water or sediment quality monitor-
ing. Monitoring of pathobiological effects provide information necessary to make
determinations of the existence of adverse effects in animals (e.g., tumors), population
productivity and stability (affected by reproduction and disease states) and the loss of
organisms deemed valuable for ecological, aesthetic, recreational, scientific or eco-
nomic reasons. Table B9-1 outlines some of the terms used in this narrative to de-
scribe pathobiological methods.
Although the value of these methods for establishing cause and effect links due to
pollutants has been proven during laboratory toxicity studies, some questions remain
regarding their ability to establish such links for field-collected organisms that are
exposed to a variety of natural environmental stresses and combinations of contami-
nants (Hinton and Couch, 1984; Couch and Harshbarger, 1985; Mix, 1986;
Sindermann, 1990). However, properly conducted multidisciplinary monitoring
studies using these methods can provide regulatory agencies with evidence of impaired
health status in animals exposed to contaminants in estuarine ecosystems. This
information can then be used to direct laboratory confirmation of the cause, if neces-
sary (see for example, Buckley et al., 1985; Gardner et al., 1990). Continued monitor-
ing by these methods can assess changes in the organisms' health status during and
-------
B-1521
, histpchernisti^;:;
mmunoassay
immunology
inclusion b6ซii!;
p'athobiology
smooth i
reticuium (SER)
IfiM^
ar
|l|i|ij^^
:m:::;:-'^':'-::1-1^-'^^
i$|l ;pr:p|sigal ;j[ฃ0jr|^^
:ii= bodies present in the jrucleiiS- or e|yjt|)pasm i6f i^rtajii^cfeii||
:;:;| viruses or as the result of degenerative: diseases ore:
(Stedman's Medical Dictionary, 1981;
, and.idleM(i^^^^ilx|^^
'VaiiSi*'!*'.^w^inn t1ซA :'/!*ซ*MMA''1^:'-:' ': -':': ^'';:::"::"::': ;'->;:::::;"::x-:":;:::'
tha|resultf^n^:ihe:djs^e:|i^e|i?^^
:;a cpnitipig:^
^
-------
B-153
following reduction or elimination of contaminant sources to meet management goals
to improve the biological status of the ecosystem.
B9.2 A field survey is required to collect target organisms and numerous tissue samples may
Monitoring Design be required for pathobiological methods (Hargis et al., 1990; U.S. EPA, 1987a). In
Considerations certain instances, a large sample size may be needed to establish statistical significance
because of normal variation from animal to animal, species and genera differences,
and migratory habits of fish. For these reasons, and the labor-intensive nature of
pathobiological methods, the availability and allocation of funds, time, equipment, and
trained personnel must be considered when planning to include pathobiological
methods in monitoring programs.
Selection of Target (Indicator) Species
A key component of any pathobiological monitoring program is the selection of target
species. The fundamental criterion is the ability to use the selected species to make
comparisons between sampling locations and sampling periods.
It is recommended that the target species possess the following characteristics :
abundant, temporally and spatially, to allow for adequate sampling
large enough to provide adequate amounts of tissue for analysis
sedentary (non-migratory) in nature to assure that pathobiological abnor-
malities are representative of the study area
easily collected
existing database of exposures and sensitivities
U.S. EPA has provided a list of suggested fish target species for different regions in
the United States (Table B9-2; U.S. EPA, 1987a). Fish provide sufficient tissue
biomass for analyses and may indicate potential threats to human populations. How-
ever, fish are often motile and pathobiological abnormalities detected may not be
representative of the study area.
Bivalve molluscs, either attached to the substratum or burrowing in sediments, have
been sampled extensively to examine the condition of these commercially important
estuarine species (i.e., Mussel Watch, National Status and Trends program). The most
-------
B-154
TABLE B9-2. HIGHEST RANKING CANDIDATE FISHES FOR USE
AS PATHOBIOLOGY MONITORING SPECIES
State
Locality
MASSACHUSETTS Swampscou
Lynn
South Essex
Boston
Fall River
New Bedford
RHODE ISLAND Newport
NEW YORK Upper East River
Lower East River
Lower Hudson River
NEW JERSEY Cape May
Secondary Selection Criteria
Species
Economic
Importance
VIRGINIA
Portsmouth
Virginia Beach
Winter flounder
Yellowtail flounder
Ocean pout
Windowpane
Winter flounder
Yellowtail flounder
Ocean pout
Winter flounder
Yellowtail flounder
Windowpane
American eel
Ocean pout
Winter flounder
Yellowtail flounder
Ocean pout
Windowpane
Winter flounder
Windowpane
Winter flounder
Scup
Summer flounder
Winter flounder
Scup
Weakfish
Winter flounder
Windowpane
Weakfish
Spot
Scup
American eel
Hogchoker
Spot
Red hake
Windowpane
Summer flounder
Spot
Summer flounder
Atlantic croaker
Hogchoker
Spot
Red hake
Summer flounder
Yes
Yes
No
No
Yes
Yes
No
Yes
Yes
No
No
No
Yes
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
No
Yes
No
No
Yes
Yes
Yes
Yes
No
Yes
No
Yes
Bioassay
Species
Yes
Yes
No
No
Yes
No
No
Yes
No
No
No
No
Yes
No
No
No
Yes
No
Yes
No
Yes
Yes
No
No
Yes
No
No
Yes
No
No
No
Yes
No
No
Yes
Yes
Yes
No
No
Yes
No
Yes
-------
B-155
TABLE B9-2.
(continued)
Secondary Selection Criteria
State
CALIFORNIA
(NORTHERN)
CALIFORNIA
(SOUTHERN)
WASHINGTON
Locality
San Francisco
Oakland
Monterey
Santa Cruz
Watson ville
Goleta
Santa Barbara
L.A. County
Orange County
Hyperion
Oceanside
Escondido
San Elijo
San Diego
Central Puget Sound
Species
English sole
Pacific sanddab
Big skate
English sole
Starry flounder
Pacific staghom sculpin
English sole
Curlfin sole
English sole
English sole
Curlfin sole
Dover sole
Pacific danddab
Longspine combfish
Spotted cusk-eel
English sole
Pacific sanddab
Dover sole
Curlfin sole
English sole
Dover sole
Pacific sanddab
English sole
Dover sole
Longspine combfish
Big skate
California skate
Dover sole
Blackbelly eelpout
Pacific sanddab
English sole
Dover sole
English sole
Pacific sanddab
Longspine combfish
English sole
Dover sole
Rock sole
Spotted raffish
Rex sole
C-O sole
Economic
Importance
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
Yes
Yes
Bioassay
Species
Yes
No
No
Yes
No
No
Yes
No
Yes
Yes
No
No
No
No
No
Yes
No
No
No
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
Yes
No
Yes
No
No
Yes
No
No
No
No
No
-------
B-1561
common target species have been oysters (Crassostrea virginica) and Mussels
(Mytilus spp.), although pathobiological and bioassay studies have also been per-
formed on other species, as well as crabs and penaeid shrimp (see also Table B10-3).
Secondary considerations, based on economic importance and status as a bioassay
organism, may be applied to further winnow the list of candidate target species to a
practical number of species to be analyzed.
Sampling Location
Appropriate locations of sampling stations depend on the objectives of the study. To
evaluate whether there is a statistically significant increase in lesions due to pollution,
stations should be located to collect specimens from contaminated and uncontaminated
(background or control) areas for statistical comparison. It is also important to demon-
strate a dose-response relationship between the pollutant concentration and tumor
incidence by locating sample stations along a contamination gradient (i.e., from highly
contaminated to moderately contaminated to uncontaminated). It is also recommended
that stations be located in areas where the geographic area of contamination is large
enough that the sampled fish could reasonably be expected to have spent a consider-
able amount of time within the influence of the pollutant (U.S. EPA, 1987a). U.S.
EPA (1989) may be consulted for further fish sampling methodologies, and informa-
tion on sampling bivalves and other invertebrates is contained in Couch (1978),
Yevich and Barszcz (1983), Turgeon etal. (1991), and other sources.
Laboratory Techniques
Light Microscopy Methods - LM (histologic) methods use the light microscope to
view cells and tissues. These methods can detect changes such as inflammatory
responses, alterations in the appearance of cells, cancerous/precancerous lesions, and
damage due to parasites (U.S. EPA, 1986). The advantages of LM methods are that
they are organ specific (i.e., the lesion can be localized to a specific organ) and can
detect microscopic and subtle cellular alterations that are not evident on visual inspec-
tion. However, LM methods are more expensive (approximately $30/sample) and
slower than gross methods (May, 1990).
Electron Microscopy Methods - EM methods use the electron microscope to detect
changes in tissue at cellular and subcellular levels, such as the identification of the
nature of inclusion bodies or changes in the amount of smooth endoplasmic reticulum.
B9.3 Existing
Analytical
Methods
-------
B-157
An advantage of EM methods is that they can be highly specific because the pollutant
can be localized within certain parts of the cell. These inclusions may contain the
causative chemical agent; lead, gold, iron, bismuth, uranium, beryllium, mercury,
copper and arsenic are a few of the metals that can be deposited intracellularly
(Sorenson and Smith, 1981). Additionally, EM methods can be used to investigate
subcellular mechanisms of pollutant action. The major disadvantage of EM methods is
that they are very expensive (minimum $400/fish) and very slow, requiring highly
skilled technical expertise.
Histochemical Methods - Histochemical methods use the microscope in conjunction
with special chemical tests and stains to localize certain enzyme systems or reaction
products in the cell. For instance, histochemical assays have been used on liver tissue
during and after tumor formation to yield important information on the biochemistry of
specific lesions (Hinton et al., 1988; Luna, 1968; Sumner, 1988). Histochemical
methods are reliable and can be highly specific as to the class of organic compound
and exact metal species, if they are properly developed. However, highly skilled
technical expertise is required to carry out the methods in the laboratory.
In vitro Tests - In vitro tests are generally more sensitive than whole animal systems,
less expensive to carry out and of shorter duration. However, using in vitro tests,
defense mechanisms found in the intact animal are missing. In vitro systems offer the
greatest flexibility for the testing and study of environmental contaminants. They can
be designed to be relevant to the species of interest in a given area, and multiple types
of measurements can be taken from a single test system (e.g., metabolic products,
mitotic activity, cytotoxicity and genetic damage). However, comparative in vitro and
in vivo studies are needed to correlate and relate changes that occur in each system
(Landolt and Kocan, 1983).
Anatomic Pathology Methods
Anatomic pathology methods examine tissues with the naked eye (gross anatomic
pathology), the aid of a light microscope (LM methods), or the electron microscope
(EM methods). Gross anatomic methods are concerned with obvious adverse changes
in tissue which can be observed in the field (Hunn, 1988; Hargis et al., 1990). The
advantage of gross methods is that large numbers of specimens can be examined
rapidly. However, the methods are generally nonspecific (i.e., it is not possible to
determine that a specific pollutant led to a specific disease). An exception to this is
cataracts in fish, which have been linked with polynuclear aromatic hydrocarbon
pollution in the field (Hargis and Zwemer, 1988).
-------
B-158
The disadvantage of the application of anatomic pathology methods to fish and shell-
fish is that these methods require specialized personnel and laboratories. These
methods are, however, generally standardized, routine, and operational in existing
Federal, State, university, veterinary, and private diagnostic laboratories that specialize
in aquatic animal pathology. Current activities are aimed at developing a field-to-
laboratory response-diagnostic scenario where adverse effects are found in the field,
diagnosis is made in the laboratory, and experimental studies are initiated to verify
results (May, 1990). See Yevich and Barszcz (1980), Howard and Smith (1983),
Johnson and Bergman (1984), Klontz (1985), Meyers and Hendricks (1985), and U.S.
EPA (1987a) for more information on these methods and techniques.
Reproduction/Development Methods
Reproduction studies are designed to examine a number of parameters in the reproduc-
tive process: gross examination of the egg and numbers of eggs, embryo viability, the
proportion of fertilized eggs (i.e., fertilization success), and larval development and
viability (U.S. EPA, 1986). Studies that examine the egg itself incorporate usage of
microscopic methods which are time-consuming and expensive (West, 1990), but have
the advantage of serving as a direct measure of reproductive success from adverse
effects of environmental pollutants. Some methods deal with the egg at the molecular
level and analyze the action of chemical and physical agents whose toxicity is directed
toward genetic (DNA) components of the egg.
Cytogenotoxic tests measure a diverse array of effects that, depending upon the
method, may indicate gene mutation, chromosome damage (micro- and macrolesions),
primary DNA damage, or oncogenesis (i.e., tumor formation and development)
(Brusick, 1980; Landolt and Kocan, 1983; Klingerman, 1982; Shugart, 1990). An-
other area that is currently being investigated as a suitable, sublethal assay for toxicity
of certain wastes in the marine environment is the fin regeneration test (Weis et al.,
1990; Cross, 1990). The information collected from these types of studies yields a
general impression about the reproductive capacity of the animals which will, in turn,
be reflected in population recruitment and abundance.
Reproductive hormones have been applied to fish to monitor environmental pollutants.
Current studies use gonadotropic and steroidogenic hormone levels, which dictate the
reproductive capacity of the animal, to assess reproductive capacity in fish (Veranasi,
1990). The analysis is a sensitive indicator of exposure affecting major biological
processes that impact the whole population. However, additional information about
-------
B-159
the normal reproductive cycle of the animals is necessary to apply these methods in the
field (Veranasi, 1990).
Biochemical Methods
Biochemical methods have been investigated for use as biomarkers of environmental
contamination in field studies since they are inherently sensitive and may provide
basic information about the early changes experienced by a cell in response to environ-
mental contamination. The development of a suite of indicators having both specific
and nonspecific responses can provide information on the type of stressors, mecha-
nisms of action, the extent of physiological dysfunction, and potential long-term
population consequences (Thomas, 1990).
Fish can respond to generalized stress, contaminants being one type of stress, through
induction of stress proteins (Pickering, 1981; Sanders, 1990). Stress proteins are
currently being investigated as a generalized biochemical indicator of stress in fish, as
a chemical class pollutant indicator, or a mode of action indicator. The methods for
detecting stress proteins in fish involve radioisotopes or immunologic methods to
measure the specific amount after a stress occurs. cDNA probes are now being used
experimentally to correlate protein synthesis with induction to detect and quantify
various enzyme systems. These methods afford a high degree of sensitivity.
It has been suggested that the induction (increased synthesis) of fish hepatic microso-
mal mono-oxygenase (MO) enzyme could serve as a sensitive biological indicator for
detection of certain classes of chemicals in water (Payne et al., 1987; Kleinow et al.,
1987; Lech et al., 1982; Jimenez et al., 1990; Haux and FOrlin, 1988).
Metallothioneins (MT) have been under consideration for use as a monitoring tool for
trace environmental metal pollution due to their induction as a result of exposure to
certain metals (Engel and Roesijadi, 1987; Garvey, 1990; Haux and Forlin, 1988).
Additional scientific research is required to understand the basic biology offish before
the exact significance of field studies with pollutants can be ascertained.
A concern of monitoring biochemical variables in fishes to detect environmental
pollutants is that their exact biological significance to the functional integrity of the
organism is not known. In addition, for most of the molecular indicators studied, the
normal range of values for a particular fish population and the factors influencing these
values are often not known (Neff, 1985). Applied research is needed to develop
simplified field methodologies. Even with these limitations, biochemical methods
-------
B-160
hold considerable promise as sensitive early indices of exposure to environmental
stressors (Thomas, 1990).
Immunologic Methods
Immunologic biomarkers are simple, sensitive, reproducible, workable in the field, and
very promising (Weeks et al., 1990; D. Anderson, 1990a; R.S. Anderson, 1990). The
advantage of immune indicators is that they provide supportive evidence for linkage
between a stressor (toxicant, etc.) and disease outbreaks in fish and shellfish. The
immune response can be a monitor for identification of a specific antigen or microor-
ganism responsible for pathological conditions in fish. Biologists can perform quick
and sensitive assays in the field or in their own diagnostic laboratories because many
immune assays are becoming available in kits (Rowley, 1990; Matthews et al., 1990).
Many immunological assays do not require sacrifice of the animal; blood samples can
be taken periodically to follow the kinetics of the effects of stress in a single animal,
although the effects of handling stress in aquatic species must also be evaluated. The
immune response is physiologically similar among most vertebrates and similar
equipment and materials can be used to test all species of fish as well as shellfish (see
Anderson, 1987). There is a rapidly growing literature on immunotoxicology from
veterinary and aquatic animal sciences.
The selection of immune system indicators for the study of stress effects on the
immune response depends on many factors, including specific study objectives,
available equipment, training, personnel, and length and number of assays. The most
sophisticated and sensitive assays are costly and require highly trained personnel,
whereas simple assays can be performed by field biologists with only basic laboratory
supplies.
The limitation of immune indicators is that the response is sometimes too broad to
provide conclusive evidence that the specific antibody-antigen reaction evident in an
immunological assay is actually the specific complex desired. Cross-reactions of
antibody and heightened responses to nonspecific factors may prevent the interpreta-
tion of assays with absolute certainty. It is difficult to know what immune indicator is
most affected and what immunological assay to apply because the immune response in
fish or shellfish is distinctive for each antigen or disease-causing agent. Expensive
materials and laboratory equipment are needed for sophisticated immunological
assays. Confirmatory assays are advisable to be sure that a particular stressor is the
only cause of a particular effect. Many animals should be sampled because of natural
variability from animal to animal.
-------
B-161
IT"
Physiologic Methods
Hematologic methods have been used for many years by biologists to assess the
general health offish in hatcheries and research laboratories. The procedures are well-
standardized (Blaxhall and Daisley, 1973; Wedemeyer and Yasutake, 1977) and are
easy to carry out, even in the field. Hematological measures are not immune from
stress of capture but are influenced far less than some other measurements such as
blood glucose (Larson et al., 1985). Hematologic methods have been successfully
implemented in the field and have been rated as the best physiologic method for
evaluation of pollutants (U.S. EPA, 1986). For the measurement of hematological
factors in fish, techniques similar to those used in human and veterinary clinical
laboratories are generally used with minor modifications (Heath, 1987; Bouck 1984).
Blaxhall and Daisley (1973), Wedemeyer and Yasutake (1977), and Ellis (1977)
provide practical guides to the adaptation of these methods for use on fish blood.
The hematocrit/erythrocyte picture may not be as sensitive to pollution as is the
leucocyte count, at least as far as its response to metals is concerned (Larson et al.,
1985). It remains to be determined how sensitive the procedure is to subtle environ-
mental changes. In general, chemical and physical stressors cause a decrease in the
leucocrit, whereas infections produce the opposite response. However, it is also
possible to get elevations in granulocytes concomitant with a decrease in lymphocytes,
thereby yielding an unchanged leucocrit (Peters et al., 1980). Thus, the method has its
limitations for the detection of chronic stress (Wedemeyer et al., 1983).
There are several obstacles, such as capture stress, that limit the potential usefulness of
physiological tests in field work. Capture stress precludes the use of sensitive, early
indicators of environmental stress. The use of physiological responses of wild fish to
assess environmental quality is difficult because responses due to toxicants often
cannot be distinguished from those induced during handling of the wild fish (Bouck,
1984).
B9.4 Q A/QC
Considerations
Pathobiological methods have a wide range of sensitivities and response time (i.e.,
when a response can be detected). For instance, certain immunologic, electron micro-
scopic, and biochemical methods can detect early changes in cells and are very sensi-
tive. Gross anatomic pathologic methods, on the other hand, can only detect cellular
changes after sufficient time passes so that the lesions can be seen with the naked eye,
and are therefore less sensitive.
-------
B-1621
General considerations for QA/QC have been covered earlier in this document. With
regard to the pathobiological methods, it should be noted that careful and consistent
handling of aquatic specimens is required to minimize trauma and confounding effects,
such as exposure to air. Organisms should be held in the laboratory under conditions
as near to those found at the site of collection as possible, for as short a time as practi-
cable, before performing assays. Fish and shellfish collected for histopathological
examination must be properly fixed (e.g., immersed in a formaldehyde or glutaralde-
hyde containing solution) to stop metabolic activity. This may require opening or
sectioning the organisms to allow the fixative to rapidly penetrate all tissues and
preserve the cellular structure in its existing condition. The organisms must still be
active or moribund, but not dead. Failure to follow proper fixation procedures will
interfere with the interpretation of lesions in anatomic pathologic studies.
As with other monitoring methods, samples must be accurately labeled at the time of
collection and routine QA/QC procedures should be instituted, including tracking of
samples, careful recording of methods used, using fresh solutions, and treating both
control and exposed samples equally for comparability. Whenever possible, organ-
isms for each group to be tested should be of the same species, age, and sex. Sections
of tissues and organs, or for small organisms, the whole animal, should also be pre-
pared as uniformly as possible with respect to homogeneity and orientation so that
microscopic observations can be made on the same organs and areas for each speci-
men. Subsamples of sections should be also be examined (blind) by another aquatic
animal pathologist to confirm the diagnoses. See U.S. EPA (1987a) for additional
information on QA/QC procedures for histopathological examinations.
Statistical strategies may mitigate the high costs of pathobiological monitoring meth- B9.5
ods. As discussed earlier, power analyses sconsidering the strategy of compositing Statistical Design
samples can often lead to a cost-effective monitoring design strategy. Analyses of Considerations
power-costs are necessary in selecting the appropriate sample/replicate number,
sample location, and sampling frequency. If the primary objective of a monitoring
program is to determine pathobiological differences among sampling locations,
composite sampling may be an appropriate strategy. However, a limitation of compos-
ite sampling is the inability to directly estimate the range and variance of the popula-
tion of individual samples. Also, the use of space- or time-bulking strategies should
also be carefully considered because significant information concerning spatial and
temporal heterogeneity may be lost. See Statistical Design Considerations: Compos-
ite Sampling, Power Analysis, and Power-Cost Analysis (Appendix B Introduction;
Section B.3).
-------
B-163
Given that the monitoring program must accommodate a fixed level of sampling effort,
the best option for pathobiological methods is to collect more replicates at fewer
locations. For histopathology, from 25 to 100 animals per species may need to be
surveyed to detect low incidences of disease. The selection of appropriate statistical
analyses must be made with these limitations in mind, as well as the specific types of
statistical analyses that can be performed on different types of data. For example,
quantitative information from measurements of enzyme levels or computerized image
analysis/morphometric programs may be effectively analyzed by parametric statistics
(if the data fit the conditions for normal distributions, limitations of assumptions, etc.),
but histopathological observations may need to be rated on presence/absence or
categorized qualitative scales that will require non-parametric techniques. U.S. EPA
(1987a) provides additional information on the types of sampling designs and statisti-
cal tests that may be appropriate for these data.
B9.6
Use of Data
Data Interpretation
Data interpretation for pathobiological methods may be limited because there are
relatively few trained personnel, facilities and equipment may be expensive, and
references for techniques that are newly emerging may be scarce. However, because
of the recent interest in developing biomarkers for monitoring the effects of natural
and anthropogenic environmental stresses on aquatic organisms, many research
programs are underway at federal, state, and university facilities to develop, standard-
ize, and validate the most appropriate biomarkers for sentinel species that will estab-
lish cause and effect links for pollutant exposure. Basic laboratory research and
experimental studies must be conducted in conjunction with field work to identify the
relationships between contaminant levels, structural, biochemical, or functional
pathologies, and population health in the field (Johnson and Bergman, 1984;
McCarthy, 1990). A critical concern is the coordination of efforts and creation of a
multidisciplinary approach.
It is important to establish baseline data on selected species as well as to demonstrate
the alterations in the health of those species' due to contaminant exposures. Methods
and techniques, vocabulary, and interpretation of lesions and effects must be standard-
ized. In addition to a rapidly expanding reference base of literature on baseline
measures of health and histological atlases for several species offish and shellfish,
courses, workshops, meeting, special symposia (e.g., Responses of Marine Organisms
to Pollutants/Woods Hole, MA and Plymouth, England; Annual Aquatic Toxicology
Workshop, Canada), and professional societies (e.g., Society for Invertebrate Pathol-
ogy, American Fisheries Society/Fish Health Section) are facilitating training in
-------
B-164
techniques and communication among different investigators and laboratories. Dis-
eases of fish and shellfish have been reviewed by several authors (i.e., Sparks, 1985;
Ferguson, 1989; Roberts, 1989; Sindermann, 1990). For histopathological studies of
tumors in fish, standardization of nomenclature will be available in Dawe et al. (in
press).
As for any monitoring program, long-term studies of health and disease in aquatic
organisms will aid in identifying and interpreting the observed pathobiological effects,
and it will be important to archive data so that it will be available for future compari-
sons. The Registry of Tumors in Lower Animals at the National Museum of Natural
History, Smithsonian Institution, Washington, D.C. houses over 3500 cases of neoplas-
tic and nonneoplastic lesions representing a wide variety of host aquatic species,
pathogens, and environmental insults from field and laboratory studies conducted
around the world. The Registry also contains the Registry of Marine Pathology,
originally developed by the National Marine Fisheries Service, and the collection of
crustacean histopathological material researched by Dr. Phyllis T. Johnson, NMFS.
Although primarily serving as a clearinghouse for information on neoplastic diseases
and research, the Registry is also able to direct inquiries for information on
nonneoplastic diseases, and the materials archived there are available for study by
qualified investigators. Data collected from the long-term monitoring efforts of
NOAA's National Status and Trends Program (NS&T) and the EPA's Environmental
Monitoring and Assessment Program (EMAP) will also provide useful information for
the interpretation of the various pathobiological (biomarker) methods that are being
employed in these studies.
Data Integration
Measuring or evaluating the effects of stressors on fish and shellfish by
pathobiological indicators is difficult primarily because of the large number of vari-
ables that can influence biological response. Variables including water temperature,
nutritional status, species, sex, reproductive and developmental stages, and physiologi-
cal functions can render tests difficult to compare and evaluate.
Anderson (1990a) outlined a plan including four levels of study in a tiered approach to
investigate the effects of stress on the immune system and to quantify the possible
contribution of these environmental and physiological variables on the stress response.
The plan, generally speaking, appears to be applicable to all pathobiological methods
and may be adapted to them using the following four levels: (1) observations of fish
and shellfish populations in the field; (2) studies of caged fish or shellfish in the field;
(3) in vivo exposures in the laboratory; and (4) in vitro assays in the laboratory.
-------
B-165
The first level of study includes the collection of information from field observations
on relative abundance, reduction in sport-fish catches, or decline in commercial
harvests. Reduced yields reported by anglers in previously productive fishing grounds
may be correlated with environmental stressors or pollutants occurring at these sites.
If sick or moribund fish or shellfish are called to the attention of local biologists by
anglers, gross morphological descriptions might be made and isolation of the disease
agent attempted.
The second level of investigation involves the use of caged fish or shellfish placed in
the field to test for the presence or action of environmental stressors. Groups of fish or
shellfish can be placed in pens or cages at suspected sites and their responses (disease
signs, acute mortality) can be compared with those of control fish or shellfish caged in
unstressful areas. Fish from the cages can be sampled at various intervals and the
intensity of various immune responses quantified as a function of the time the fish or
shellfish were exposed to water having defined characteristics. Levels of immune
responses can be compared by injecting specific disease agents into both control and
test fish or shellfish, and recording disease and death rates.
The third level of assay is in vivo laboratory tests where immune response can be
evaluated in experiments with calibrated dilutions of specific contaminants and kinetic
measurements of the immune response at each dilution. Challenges with disease
agents can be more easily controlled in the laboratory to provide information on how
the stressor makes the fish or shellfish more susceptible to specific pathogens.
A fourth level of investigation is the recently developed method for testing the effects
of chemicals, drugs, and other stressors in vitro. Spleens and other immunopoietic
organs can be removed from fish and placed in tissue culture media to test their
reactions to pollutants (Anderson et al, 1986) using the passive hemolytic plaque
assay. Use of this method to measure the effects of stressors allows maximal control
of the levels of pollutants. Because the immune response is monitored under defined
laboratory and environmental conditions, important information is obtained about how
specific stressors affect the immune response. In vitro methods require fewer fish than
in vivo methods because tissue and organ samples can be divided into many sections,
which also reduces the variability of responses.
McCarthy (1990) presented a research strategy to validate biomarkers and provide the
scientific understanding necessary to interpret biomarker responses of environmental
species. An evolving monitoring program was proposed that focused broadly on
evaluation of contamination in an array of ecosystem types. The challenges and
obstacles to be addressed in such a program include the following:
-------
B-1661
The quantitative and qualitative relationships between chemical exposure,
biomarker responses, and adverse effects must be established
Responses due to chemical exposure must be distinguishable from natural
sources of variability (ecological and physiological variables, species-
specific differences, and individual variability) if biomarkers are to be
useful in evaluating contamination
The validity of extrapolating between biomarker responses measured in
individual organisms and some higher-level effect at a population or
community level must be established
The use of exposure biomarkers in animal surrogates to evaluate the
potential for human exposure should be explored
The plan is an ambitious multi-year research and development program involving the
formulation of a long-term, interagency, interdisciplinary activity such as the Environ-
mental Monitoring and Assessment Program (EMAP) by EPA in cooperation with
other federal agencies. The research questions are presented herewith; the reader is
referred to McCarthy (1990) for the research description, objective and approach.
Rationale B9.7
Summary and
Pathobiological methods provide information concerning damage to Recommendations
biological organ systems through evaluation of their altered structure,
activity and function and can be used to determine adverse effects of
pollutants in the environment.
Pathobiological methods should be used in concert with one another to
correlate cause and effect type relationships, rather than in isolation
because it is difficult to interpret results from separate studies.
Monitoring Design Considerations
It is important to demonstrate a dose-response relationship between the
pollutant concentration and tumor incidence by locating sample stations
along a contamination gradient (i.e., from highly contaminated to moder-
ately contaminated to uncontaminated).
-------
B-167
Stations should be located in areas where the geographic area of contami-
nation is large enough that the sampled fish could reasonably be expected
to have spent a considerable amount of time within the influence of the
pollutant.
Large sample size or numbers of samples may be needed to establish
statistical significance because of normal variation from animal to animal,
species and genera differences, and migratory habits of the fish.
Existing Analytical Methods
Anatomic Pathology Methods-
Tissues are examined with the naked eye or the aid of a microscope
Light microscopic (LM) methods
- detect changes at the cellular level
- organ and time-series specific
Electron microscopic (EM) methods
- detect subcellular changes
- specific as to area of cell or site of pollutant action
Histochemical methods
- detect enzyme systems or reaction products in the cell
- specific to chemical class
Reproduction/Development Methods-
Reproductive capacity is reflected in population recruitment and abundance
Methods that examine the egg itself provide a direct measure of reproduc-
tive effects from pollutants
Cytogenetic (DNA) tests
- in vitro tests offer greatest flexibility in terms of sensitivity,
expense, and time required, and are frequently conducted for
convenience and availability
- sister-chromatid exchange (SCE) assays are DNA damage tests
used as a screening tool, and are dose-responsive and sensitive to
low concentrations of pollutants
-------
B-1681
- the aneuploidy technique is simple and easy to use, but is inaccu-
rate for sublethal effects
- chromosomal alteration analysis
Biochemical Methods-
Several methods are specific for a certain compound class
Use a suite of indicators for environmental monitoring
Microsomal mono-oxygenase (MO) assays could serve as sensitive bio-
logical indicators for certain chemical classes
Metallothionein (MT) assays are nonspecific for metal exposure, but can
provide information on likelihood of a particular metal pool to produce a
pathological effect
Stress proteins are not pollutant-specific, but could result from generalized
stress
Immunologic Methods
Immunologic assays can provide information on pollutant-induced stress
effects, however they may require confirmatory assays
Many animals should be sampled because of natural variability from
animal to animal
Physiologic Methods-
Serious "interferences" can be caused by stress induced during fish or
shellfish collection that limit the potential usefulness of physiological tests
because effects of toxicants cannot be distinguished from those induced
during handling of the wild organisms
Hematologic methods for fish, such as measurements of hematocrit
(packed cell volume), hemoglobin concentration, erythrocyte count and
leucocyte count (or volume) can be successfully used, although the fish are
not immune from stress of capture, hematologic methods are influenced far
less than some other measurements such as blood glucose
The leucocrit can be a more sensitive measure of metals pollution than the
hematocrit
-------
B-169
QA/QC Considerations
Pathobiological methods have a wide range of sensitivities
Organisms must be carefully handled and properly prepared for each
method
The main areas of concern with regard to analytical QA/QC are precision,
accuracy, representativeness, completeness, and comparability
Statistical Design Considerations
Compositing tissue sampling consists of mixing tissue samples from two or
more individual organisms collected at a particular location and time
period
Space- (combining composites from several locations) and/or time- (com-
bining several composites over time from one location) bulking strategies
should be used judiciously since information concerning spatial and
temporal heterogeneity may be lost
Power analyses have shown that for a fixed level of sampling effort, a
monitoring program's power is generally increased by collecting more
replicates at fewer locations
The appropriate statistical tests must be performed and can vary depending
on the type of data generated
Use of Data
Data use and interpretation may be limited by relatively few trained
personnel and expensive equipment, but many research and training
programs are now underway at federal, state, and university facilities to
improve biomarker methods in sentinel species for effective monitoring of
contaminant exposures
Basic laboratory research must be conducted and biological methods must
be tested in the field
-------
B-170'
Information communication
- need to coordinate efforts, communicate and create a
multidisciplinary approach to relate lab and field studies and
establish cause and effect type relationships
Data integration
- long-range research strategies should be followed to validate
biomarkers and provide scientific understanding necessary to
interpret biomarker responses of environmental species.
Adams, S.M. (ed.) 1990. Biological Indicators of Stress in Fish. American Fisheries B9.8
Society Special Symposium No. 8. Literature Cited
and References
Adams, S.M., L.R. Shugart, and G.R. Southworth. 1990. Application of bioindicators
in assessing the health of fish populations experiencing contaminant stress. In: Bio-
markers of Environmental Contamination (McCarthy, J. and L. Shugart, eds.), pages
333-353. Boca Raton, FL: CRC Press.
Anderson, D., B. Roberson and O. Dixon. 1979. Cellular immune response in Rain-
bow Trout, Salmo gairdneri Richardson to Yersinia Ruckeri O-antigen monitored by
the passive haemolytic plaque assay test. J. Fish Dis. 2:169-178.
Anderson, D., O. Dixon and E. Lizzio. 1986. Immunization and culture of Rainbow
Trout organ sections in vitro. Veterinary Immunology and Immunopathology
12:203-211.
Anderson, D. 1990a. Immunological indicators: Effects of environmental stress on
immune protection and disease outbreaks. American Fisheries Society Symposium
8:38-50, Washington, DC.
Anderson, D. 1990b. Passive hemolytic plaque assay for detecting antibody-produc-
ing cells in fish. In: Techniques in Fish Immunology. (Stolen, J., J. Fletcher, D.
Anderson, B. Roberson, W. van Muiswinkel, eds). Fair Haven, NJ: SOS Publications.
Anderson, D..O. Dixon and W. van Muiswinkel. 1990. Reduction in the numbers of
antibody-producing cells in Rainbow Trout, Oncorhynchus mykiss, exposed to suble-
thal doses of phenol before bath immunization. In: Aquatic Toxicology. (Nriagu, J.,
ed.). A Wiley-Interscience Publication, John Wiley and Sons.
-------
B-171
Anderson, R.S. 1987. Immunocompetence in invertebrates. In: Pollutant Studies in
Marine Animak (Giam, C.S. and L.E. Ray, eds.). Boca Raton, FL: CRC Press.
Anderson, R.S. 1990. Effects of pollutant exposure on bactericidal activity of
Mercenaria mercenaria hemolymph. In: Biological Markers of Environmental
Contaminants (McCarthy, J.F. and L.R. Shugart, organizers), American Chemical
Society, Los Angeles, CA, 25-30 September 1990.
Blaxhall, P. and K. Daisley. 1973. Routine haematological methods for use with fish
blood. J. Fish Biol. 5:771.
Bouck, G.R. 1984. Physiological responses of fish: problems and progress toward
use in environmental monitoring. In: Contaminant Effects on Fisheries (Cairns, V.W.,
P.V. Hodson and J.O. Nriagu, eds.). New York: John Wiley & Sons.
Brown, D.A., C.A. Bowden, K. Chatel, and T.R. Parsons. 1977. The wildlife commu-
nity of lona Island Jetty, Vancouver, B.C., and heavy metal pollution effects. Environ.
Conserv. 4:213-216.
Brusick, D. 1980. Principles of Genetic Toxicology. New York, New York: Plenum
Press.
Buckley, L.J. et al. 1985. Comparative swimming stamina, biochemical composition,
backbone mechanical properties, and histopathology of juvenile striped bass form
rivers and hatcheries of the eastern United States. Trans. Amer. Fish. Soc. 114:114-
124.
Couch, J. A. 1978. Diseases, parasites, and toxic responses of commercial penaeid
shrimps of the Gulf of Mexico and South Atlantic coasts of North America. Fish.
Bull. 76:1-44.
Couch, J.A. and J.C. Harshbarger. 1985. Effects of carcinogenic agents on aquatic
animals: an environmental and experimental overview. Environ. Carcinogenesis
Revs. 3(1):63-105.
Cross, J. and J. E. Hose. 1988. Evidence for impaired reproduction in White Croaker
(Genyonemus lineatus) from contaminated areas off southern California. Marine
Environmental Research 24:185-188.
-------
B-172
Dawe, C.J., J.C. Harshbarger, R. Wellings, and J.D. Strandberg. In Press. The
Palhobiology of Spontaneous and Induced Neoplasms in Fishes: Comparative Charac-
terization. Nomenclature, and Literature. Academic Press, New York.
Dixon. 1982. Marine Biological Letters 3:155-161.
Dorigan, J.V. and F.L. Harrison. 1987. Physiological Responses of Marine Organ-
isms to Environmental Stresses. Washington, D.C.: U.S. Department of Energy.
Elskus, A. and J. Stegeman. 1989. Induced cytochrome P-450 in Fundu/ws
heteroclitus associated with environmental contamination by polychlorinated biphe-
nyls and polynuclear aromatic hydrocarbons. Marine Environmental Research
27:31-50.
Engel, D. 1987. Metal regulation and molting in the Blue Crab, Callinectes sapidus:
Copper, zinc and metallothionein. Biol. Bull. 172:69-82.
Engel, D. and G. Roesijadi. 1987. Metallothioneins: A monitoring tool. In: Pollu-
tion Physiology of Estuarine Organisms. (Vernberg, W., A. Calabrese, F. Thurberg
and F. J. Vemberg, eds.). pp 421-438. University of South Carolina Press.
Engel, D. 1988. The effect of biological variability on monitoring strategies:
Metallothioneins as an example. Water Resources Bulletin 24(5):981-987.
Ellis, A. 1977. The leucocytes of fish: A review. J. Fish Biol. 11:453.
Ferguson, H.W. 1989. Systemic Pathology of Fish: A Text and Atlas of Comparative
Tissue Responses in Diseases of Teleosts. Ames: Iowa State University Press.
Fisher, W.S. 1988. Disease Processes in Marine Bivalve Molluscs. American Fisher-
ies Society Special Publication 18.
Gardner, G.R., P.P. Yevich, J.C. Harshbarger, and A.R. Malcolm. 1991. Carcinoge-
nicity of Black Rock harbor sediment to the eastern oyster and trophic transfer of
Black Rock Harbor carcinogens from the blue mussel to the winter flounder. Environ.
Health Perspect. 90:53-66.
Garvey, J. 1990. Metallothionein: A potential biomarker of exposure to environmen-
tal toxins. In: Biomarkers of Environmental Contamination. (McCarthy, J. and L.
Shugart, eds). Boca Raton, FL:CRC Press
-------
B-173
Giam, C.S. and L.E. Ray. 1987. Pollutant Studies in Marine Animals. Boca Raton,
FL: CRC Press.
Haasch, M., P. Wejksnora, J. Stegeman and J. Lech. 1989. Cloned Rainbow Trout
liver P ^50 complementary DNA as a potential environmental monitor. Toxicology
and Applied Pharmacology 98:362-368.
Hargis, W., M. Roberts and D. Zwerner. 1984. Effects of contaminated sediments and
sediment-exposed effluent water on estuarine fish: Acute toxicity. Marine Environ-
mental Research 14:337-354.
Hargis, W. and D. Zwerner. 1988. Effects of certain contaminants on eyes of several
estuarine fishes. Marine Environmental Research 24:265-270.
Hargis, W. 1990. Virginia Institute of Marine Science, Gloucester Point, VA, per-
sonal communication.
Harrington, F.W. and J.O. Corliss. 1991. Microscopic Anatomy of Invertebrates.
volumes 1-15 (some still in press). New York: Wiley-Liss.
Haux, C. and L. Fo'rlin. 1988. Biochemical methods for detecting effects of contami-
nants on fish.
Heath, A. 1987. Water Pollution and Fish Physiology. Boca Raton, FL: CRC Press.
Hernberg, S. 1976. Biochemical, subclinical, and clinical responses to lead and their
relation to different exposure levels as indicated by concentration of lead in blood. In:
Effects and Dnse-Response Relationships of Toxic Metals. (Norberg, G., ed.) Elsevier,
Amsterdam, 404.
Hinton, D.E. and J.A. Couch. 1984. Pathobiological measures of marine pollution
effects. In: Concepts in Marine Pollution Measurements (White, H.H., ed.), pp. 7-32.
College Park, MD: Maryland Sea Grant College.
Hinton, D., J. Couch, S. Teh and L. Courtney. 1988. Cytological changes during
progression of neoplasia in selected fish species. In: Aquatic Life Toxicology. Toxic
Chemicals and Aquatic Life: Research and Management. (Malins, D., A. Jensen and
M. Moore, ed.). Elsevier Science Publishers.
-------
B-174
\ *
Holeton, G. 1972. Gas exchange in fish with and without hemoglobin. Respir.
Physicol. 14:142.
Hose, I.E., J.N. Cross, S.G. Smith and D. Diehl. 1989. Reproductive impairment in a
fish inhabiting a contaminated coastal environment off southern California. Environ-
mental Pollution 57:139-148.
Howard, D.W. and C.S. Smith. 1983. Histological Techniques for marine Bivalve
Molluscs. NOAA Technical Memorandum. MNFS-F/NEC-25. U.S. Department of
Commerce/NOAA, Woods Hole, MA.
Hunn, J. 1988. Fish and Wildlife Service. US Dept. of the Interior. Field Assessment
of the Effects of Contaminants on Fishes. Report. Columbia, MO.: Biological Report
88 (19).
Jeme, N. and A. Nordin. 1963. Plaque formation in agar by single antibody-produc-
ing cells. Science 140:405.
Jimenez, B., A. Oikari, S. Adams, D. Hinton and J. McCarthy. 1990. Hepatic en-
zymes as biomarkers of environmental, physiological and toxicological variables. In:
Biomarkers of Environmental Contamination. (McCarthy, J. and L. Shugart, eds).
Boca Raton, FL: CRC Press.
Johnson, R. and H. Bergman. 1984. Use of histopathology in aquatic toxicology: A
critique. In: Contaminant Effects on Fisheries (Cairns, V., P. Hodson, J. Nriagu,
eds.). John Wiley and Sons.
Kleinow, K., M. MeLancon and J. Lech. 1987. Biotransformation and induction:
Implications for toxicity, bioaccumulation and monitoring of environmental
xenobiotics in fish. Environ. Health Perspect. 71:105-119.
Klingerman, A.D. 1982. Fishes as biological detectors of the effects of genotoxic
agents. In: Mutaeenicitv. New Horizons in Genetic Toxicology. (Meddle, J., ed).
Academic Press.
Klontz, G. 1985. Diagnostic methods in fish diseases: Present status and needs. In:
Fish and Shellfish Pathology (Ellis, A., ed.). Academic Press Inc.
-------
B-175
Landolt, M. L. and R.M. Kocan. 1983. Fish cell cytogenetics: A measure of the
genotoxic effects of environmental pollutants. In: Aquatic Toxicology (Nriagu, j. ed).
A Wiley-Interscience Publication, John Wiley and Sons.
Larson, A., C. Haux and M. Sjobeck. 1985. Fish physiology and metal pollution:
Results and experiences from laboratory and field studies. Ecotoxicol. Environ. Saf.
9:250.
Lech, J., M. Vodicnik and C. Elcombe. 1982. Induction of mono-oxygenase activity
in fish. In: Aquatic Toxicology (Weber, L., eds.) pp. 107-148. Raven Press.
Luna, L. 1968. Manual of Histologic Staining Methods of the Armed Forces Institute
of Pathology. New York: McGraw-Hill Book Company, The Blakiston Division.
Malins, D.C. and A. Jensen. 1988. Aquatic Toxicology. Amsterdam: Elsevier
Science Publishers.
Matthews, E., J. Warinner and B. Weeks. 1990. Assays of immune function in fish
macrophages. In: Techniques in Fish Immunology. (Stolen, J., T. Fletcher, D. Ander-
son, B. Roberson, W. van Muiswinkel, eds). Fair Haven, NJ: SOS Publications.
May, E.B. 1990. Maryland Department of natural Resources, Fisheries Division,
Annapolis, MD, personal communication.
May, E.B., M.J. Garreis, and M.M. Lipsky. 1985. Histological markers of environ-
mental effect. Fourth Symposium on Coastal and Ocean Managements.
McCarthy, J.F. 1990. Concluding remarks: Implementation of a biomarker-based
environmental monitoring program. In: Biomarkers of Environmental Contamination.
(McCarthy, J., and L. Shugart, eds). Boca Raton, FL: CRC Press.
McCarthy, J.F. and L.R. Shugart. 1990. Biomarkers of Environmental Contamina-
tion. Boca Raton, FL: Lewis Publishers.
McLea, D. and M. Gordon. 1977. Leucocrit: A simple haematological technique for
measuring acute stress in Salmonid fish, including stressful concentrations of pulpmill
effluent. J. Fish Res. Bd. Can. 34:2164.
-------
B-1761
Meyers, T.R. and J.D. Hendricks. 1985. Histopathology. In: Fundamentals of
Aquatic Toxicology: Methods and Application (Rand, G.M. and S.R. Petrocelli, eds.),
pp. 283-331. New York: Hemisphere Publishing Co.
Mix, M.C. 1986. Cancerous diseases in aquatic animals and their association with
environmental pollutants: a critical literature review. Mar. Environ. Res. 20(1 and
Myers, M.S., J.T. Landahl, M.M. Krah, and B.B. McCain. 1991. Relationships
between hepatic neoplasms and related lesions and exposure to toxic chemicals in
marine fish from the U.S. West coast. Environ. Health Perspect. 90:7-16.
Neff, J.M. 1985. Use of biochemical measurements to detect pollutant-mediated
damage to fish. American Society for Testing and Materials. Special Technical
Publication. 854:155-181.
Overstreet, R.M. 1988. Aquatic pollution problems, southeastern U.S. coasts:
histopathological indicators. Aquatic Toxicology 11:213-239.
Passino, D.R. 1984. Biochemical indicators of stress in fishes: An overview. In:
Aquatic Toxicology (Nriagu, J., ed.). A Wiley-Interscience Publication, John Wiley
and Sons.
Patton, J. and J. Couch. 1984. Can tissue anomalies that occur in marine fish impli-
cate specific pollutant chemicals? In: Concepts in Marine Pollution Measurements
(White, H., ed). Maryland Sea Grant College, University of Maryland, Publisher.
Payne, J., L. Fancey, A. Rahimtula and E. Porter. 1987. Review and perspective on
the use of mixed-function oxygenase enzymes in biological monitoring. Comp.
Biochem. Physiol. C86:233-245.
Peters, G., H. Delventhal, and H. Klinger. 1980. Physiological and morphological
effects of social stress in the eel, Anguilla anguilla. L. Art. Fisch. Wiss. 30: 157.
Pickering, A. 1981. Stress and Fish. Academic Press.
Roberts, RJ. 1989. Fish Pathology London, England: Ballmore Tindall.
-------
B-177
Rowley, A. 1990. Collection, separation and identification of fish leucocytes. In:
Techniques in Fish Immunology. (Stolen, J., T. Fletcher, D. Anderson, B. Roberson,
W. van Muiswinkel, eds). Fair Haven, NJ: SOS Publications.
Sanders, B. 1990. Stress proteins. In: Biomarkers of Environmental Contamination.
(McCarthy, J. and L. Shugart, eds). Boca Raton, FL: CRC Press.
Schmid. 1982. Chapter 36 In: Chemical Mutagens: Principles and Methods for their
Detection. (Hollaender, A., ed). Plenum Press.
Sindermann, C.J. 1983. An examination of some relationships between pollution and
disease. Rapp. P.-V. Reun. Cons. Int. Explor. Men 182:37-43.
Sindermann, C.J. 1990. Principal Diseases of Marine Fish and Shellfish. Volumes 1
and 2. San Diego, CA: Academic Press.
Snieszko, S. 1974. The effects of environmental stress on outbreaks of infectious
diseases offish. J. Fish Biol. 6:197-208.
Sorenson, E.M.B. and N.K.R. Smith. 1981. Hemosiderin granules: Cytotoxic re-
sponse to arsenic exposure in channel catfish, Ictaluruspunctatus. Bull. Environ.
Contarn. Toxicol. 27:645-653.
Sorenson, E.M. 1991. Metal Poisoning in fish. Boca Raton, FL: CRC Press.
Sparks, A.K. 1985. Synopsis of Invertebrate Pathology Exclusive of Insects.
Amsterdam: Elsevier Science Publishers.
Stedman's Medical Dictionary. 1982. 24th Edition. Baltimore, MD: Williams &
Wilkins.
Stegeman, J., B. Woodin, A. Goksoyr. 1988. Apparent cytochrome P-450 induction
as an indication of exposure to environmental chemicals in the flounder, Platichthys
flesus. Marine Ecol. Prog. Serv. 46:55-60.
Stegeman, J., K. Renton, B. Woodin, Y. Zhang and R. Addison. 1990. Experimental
and environmental induction of cytochrome P450E in fish from Bermuda waters. J.
Exp. Mar. Biol. Ecol. 138:49-67.
-------
B-178
Stegeman, J.J. and J.J. Lech. 1991. Cytochrome P-450 systems in aquatic species:
carcinogen metabolism and biomarkers for carcinogen and pollutant exposure.
Environ. Health Perspect. 90:101-116.
Sumner, B.E.H. 1988. Basic Histochemistrv. New York, NY: John Wiley & Sons.
Taber's Cyclopedic Medical Dictionary. 1985. Philadelphia, PA: F.A. Davis Com-
pany.
Thomas, P. 1990. Molecular and biochemical responses of fish to stressors and their
potential use in environmental monitoring. American Fisheries Society Symposium
8:9-28, Washington, DC.
Turgeon, D.D., S.B. Bricker and T.P. O'Connor. 1991 (in press). National Status and
Trends Program: Chemical and biological monitoring of U.S. coastal waters. In:
Ecological Indicators. Vol. 1 (McKenzie, D.H. and D.E. Hyatt, eds.). Essex, England:
Elsevier Applied Science.
U.S. EPA. 1985. Methods for measuring the acute toxicity of effluents to freshwater
and marine organisms. 3rd Edition. EPA/600/4-85/013. Environmental Monitoring
and Support Laboratory, U.S. Environmental Protection Agency. Cincinnati, OH.
U.S. EPA. 1986. Proceedings and summary of the workshop on finfish as indicators
of toxic contamination. Office of Marine and Estuarine Protection. Washington, DC.
July 27-28, 1986, Arlie, VA.
U.S. EPA. 1987a. Guidance for conducting fish liver histopathology studies during
30l(h) Monitoring. EPA 430/09-87-004. Washington, DC.
U.S. EPA. 1987b. Bioaccumulation monitoring guidance: Strategies for sample
replication and compositing, vol. 5. EPA 430/9-87-003. Office of Marine and Estua-
rine Protection, Washington, DC. 51 pp.
U.S. EPA. 1987c. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
U.S. EPA. 1988. Short-term methods for estimating the chronic toxicity of effluents
and receiving waters to marine and estuarine organisms. EPA/600/4-87/028. Environ-
-------
B-179
mental Monitoring and Support Laboratory, US Environmental Protection Agency.
Cincinnati, OH.
U.S. EPA. 1989. Rapid bioassessment protocols for use in streams and rivers. EPA
444/4-89-001. Washington, DC.
U.S. EPA. 1990. Three year assessment of reproductive success in Winter Flounder,
Pseudopleuronectes Americanos (Walbaum), in Long Island Sound, with comparisons
to Boston Harbor: 1986-1988. I. Reproductive cycle: Vitellogenin. II. Comparative
reproductive success: Biology, biochemistry, chemistry. III. Comparative embryo
development and mortality. Final Report. Regions I, II. Long Island Sound Project.
Veranasi, U. 1990. National Oceanic and Atmospheric Administration, National
Marine Fisheries Service, Northwest Fisheries Science Center, Seattle, WA, personal
communications.
Vogelbein, W., 3. Fournie, P. Van Veld, and R. Huggett. 1990. Hepatic neoplasms in
the Mummichog, Fundulus heteroclitus, from a creosote-contaminated site. Cancer
Research 50:5978-5986.
Warinner, J., E. Mathews and B. Weeks. 1988. Preliminary investigations of the
chemiluminescent response in normal and pollutant-exposed fish. Marine Environ-
mental Research 24:281-284.
Wedemeyer, G. and W. Yasutake. 1977. Clinical methods for the assessment of the
effects of environmental stress on fish health. Technical Paper 89, US Fish and
Wildlife Service, US Dept. of the Interior, Washington, DC.
Wedemeyer G., R. Gould and W. Yasutake. 1983. Some potentials and limits of the
leucocrit test as a fish health assessment method. J. Fish Biol. 23:711.
Weeks, B. and J. Warinner. 1984. Effects of toxic chemicals on macrophage phago-
cytosis in two estuarine fishes. Marine Environmental Research 14:327-335.
Weeks, B. and J. Warinner. 1986. Functional evaluation of macrophages in fish from
a polluted estuary. Veterinary Immunology and Immunopathology 12:313-320.
Weeks, B., J. Warinner, P. Mason and D. McGinnis. 1986. Influence of toxic chemi-
cals on the chemotactic response of fish macrophages. J. Fish Biol. 28:653-658.
-------
B-180
Weeks, B., A. Keisler, J. Warinner and E. Mathews. 1987. Preliminary evaluation of
macrophage pinocytosis as a technique to monitor fish health. Marine Environmental
Research 22:205-213.
Weeks, B., R. Huggett, J. Warinner and E. Mathews. 1990. Macrophage responses of
estuarine fish as bioindicators of toxic contamination. In: Biomarkers of Environmen-
tal Contamination. (McCarthy, J. and L. Shugart, ed.). Boca Raton, FL: CRC Press.
Weis, J.S., P. Weis and Zimmerer. 1990. Potential utility of fin regeneration in testing
sublethal effects of wastes. In: Oceanic Processes in Marine Pollution Vol.6.. Physi-
cal and Chemical Processes: Transport and Transportation. (Baumgortner, D. and I.
Duedall, eds). Krieger Pub. Co.
West, G. 1990. Methods of assessing ovarian development in fishes: A review. Aust.
J. Mar. Freshwater Res. 41:199-222.
Wolke, R., C. George and V. Blazer. 1984. Pigmented macrophage accumulations
(MMC;PMB): Possible monitor of fish health. In: Parasitology and Pathology of
Marine Organisms in the World Ocean. (Hargis, W., ed.). NOAA Technical Report
NMFS 25, p.93.
Yevich, P.P. and C.A. Barszcz. 1980. Preparation of aquatic animals for
histopathological examination. In: International Mussel Watch. Appendix 6-13, pages
212-220. Washington, D.C.: U.S. National Academy of Sciences.
Yevich, P.P. and C.A. Barszcz. 1983. Histopathology as a monitoring for marine
pollution: Results of histopathological examinations of the animals collected for the
1976 Mussel Watch Program. Rapp. P.-v. Reun. Cons. Int. Explor. Mer 182:96-102.
-------
B-181
B10.0 Bioaccumulation
Bioaccumulation is the overall process of biological uptake and retention of chemical
contaminants obtained from foods, water, sediments, or any combination of exposure
pathways - a list of terms may be found in Table B10-1. It is a consequence of an
organism's physiological limitations to transform and excrete the invading chemical
substances.
^j^^i^^j^^^^^Si^S^^^ii^ji^i^
pi^i^l^^
pjj||i|^^
B10.1 The presence of toxics in the waters and sediments of estuaries can have adverse
Rationale ecological and human health effects. Potential consequences of bioaccumulation of
chemical contaminants in estuarine organisms include, but are not limited to:
significant mortality to susceptible estuarine organisms
lethal or sublethal chronic toxic responses at later stages of the life cycle or
under conditions of stress for susceptible estuarine organisms
the transfer of increasingly greater concentrations of toxic pollutants to
organisms at higher trophic levels - including humans
It is difficult to calculate the uptake and bioaccumulation of contaminants from
measured values in the water and sediments. Contaminants often occur in the water
-------
B-182
column at very low concentrations which are difficult to detect, yet can be accumu-
lated in detectable amounts in fish and shellfish tissue. In those cases where toxics
which can be measured in the water and sediments, there is a large degree of uncer-
tainty associated with even the most sophisticated models for predicting uptake and
bioaccumulation. Direct monitoring of the concentration of contaminants of concern
in the tissue of selected estuarine organisms can provide a spatial and temporal record
of contaminant concentrations and bioaccumulation in the estuary.
Monitoring of bioaccumulation in the estuary will provide the data which links expo-
sure and effects, and thus can generate important insights into ecological effects,
human health risks, and routes and extent of pollutant exposure. This information can
be used to both evaluate the effectiveness of the CCMP in mitigating the potential
effects due to bioaccumulation and to provide an early warning system for the poten-
tial risks associated with the bioaccumulation of xenobiotics.
Bioaccumulation monitoring studies have varied substantially, depending upon study- B10.2
specific objectives. Generally, bioaccumulation monitoring programs whose objective Monitoring Design
was to assess human health effects were significantly different in their selection of Considerations
target species and tissues when compared to those which were intended to assess
impacts to estuarine communities. Human health risk assessments usually examine the
muscle tissue of commercially important fish and shellfish, whereas environmental
impact assessments often survey organisms at lower trophic levels (e.g. infaunal
macroinvertebrates). Unfortunately, formulations which would allow comparisons of
bioaccumulation data between different species and/or different tissues types have not
yet been developed. Thus, it is essential that monitoring design elements be standard-
ized to allow for comparisons among estuarine studies.
A common deficiency in many programs, is the inability to collect sufficient tissue
biomass of appropriate species across sampling locations throughout the study. The
selection of appropriate species and tissues must account for natural fluctuations in
populations as well as changes due to anthropogenic perturbations. Indigenous species
initially present may not be available later, limiting temporal and spatial comparisons.
Selection of Target Species
A key component of any bioaccumulation monitoring program is the selection of
target species. Concentrations of chemical residues in tissues of target species serve as
indicators of contamination throughout the biological system. The fundamental
-------
B-183
criterion is the ability to use the selected species to make comparisons between sam-
pling locations and sampling periods.
Target Species Characteristics - It is recommended that the target species possess
the following characteristics (U.S. EPA, 1985a):
high bioaccumulation potential for selected contaminants of concern
metabolic regulation of selected contaminants should be weak or absent in
order to assess a "worst case scenario"
abundant, temporally and spatially, to allow for adequate sampling
large enough to provide adequate amounts of tissue for analysis
sessile or sedentary in nature to assure bioaccumulation is representative of
the study area
easily collected
existing database of exposures and sensitivity
U.S. EPA has provided a list of suggested fish and macroinvertebrate target species
(Tables B10-2 and B10-3; U.S. EPA, 1985a) for different regions in the United States.
Secondary considerations, based on economic importance and status as a bioassay
organism, may be applied to further winnow the list of candidate target species to a
practical number of species to be analyzed.
Candidate target species should be objectively ranked based on ecological characteris-
tics which would enhance their potential for bioaccumulation, and facilitate sampling
and/or analytical procedures (U.S. EPA, 1985a). Macroinfauna have several advanta-
geous characteristics as target species, including:
are sedentary; therefore, bioaccumulation will be representative of the
study area
represent a significant food source for higher trophic levels and therefore
may indicate contaminants available in the food chain
-------
B-184!
TABLE B10-2. HIGHEST RANKING CANDIDATE FISHES FOR USE
AS BIO ACCUMULATION MONITORING SPECIES
Secondary Selection Criteria
State Locality
MASSACHUSETTS Swampscolt
Lynn
South Essex
Boston
Fall River
New Bedford
RHODE ISLAND Newport
NEW YORK Upper East River
Lower East River
Lower Hudson River
NEW JERSEY Cape May
VIRGINIA Portsmouth
Virginia Beach
Species
Winter flounder
Yellowtail flounder
Ocean pout
Windowpane
Winter flounder
Yellowtail flounder
Ocean pout
Winter flounder
Yellowtail flounder
Windowpane
American eel
Ocean pout
Winter flounder
Yellowtail flounder
Ocean pout
Windowpane
Winter flounder
Windowpane
Winter flounder
Scup
Summer flounder
Winter flounder
Scup
Weakfish
Winter flounder
Windowpane
Weakfish
Spot
Scup
American eel
Hogchoker
Spot
Red hake
Windowpane
Summer flounder
Spot
Summer flounder
Atlantic croaker
Hogchoker
Spot
Red hake
Summer flounder
Economic
Importance
Yes
Yes
No
No
Yes
Yes
No
Yes
Yes
No
No
No
Yes
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
No
Yes
No
No
Yes
Yes
Yes
Yes
No
Yes
No
Yes
Bioassay
Species
Yes
Yes
No
No
Yes
No
No
Yes
No
No
No
No
Yes
No
No
No
Yes
No
Yes
No
Yes
Yes
No
No
Yes
No
No
Yes
No
No
No
Yes
No
No
Yes
Yes
Yes
No
No
Yes
No
Yes
-------
B-185
TABLE B10-2.
(continued)
Secondary Selection Criteria
State Locality
CALIFORNIA San Francisco
(NORTHERN)
Oakland
Monterey
Santa Cruz
Watson ville
CALIFORNIA Goleta
(SOUTHERN)
Santa Barbara
L.A. County
Orange County
Hyperion
Oceanside
Escondido
San Elijo
San Diego
WASHINGTON Central Puget Sound
-
Species
English sole
Pacific sanddab
Big skate
English sole
Starry flounder
Pacific staghom sculpin
English sole
Curlfin sole
English sole
English sole
Curlfin sole
Dover sole
Pacific danddab
Longspine combfish
Spotted cusk-eel
English sole
Pacific sanddab
Dover sole
Curlfin sole
English sole
Dover sole
Pacific sanddab
English sole
Dover sole
Longspine combfish
Big skate
California skate
Dover sole
Blackbelly eelpout
Pacific sanddab
English sole
Dover sole
English sole
Pacific sanddab
Longspine combfish
English sole
Dover sole
Rock sole
Spotted ratfish
Rex sole
C-Osole
Economic
Importance
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
No
Yes
Yes
Bioassay
Species
Yes
No
No
Yes
No
No
Yes
No
Yes
Yes
No
No
No
No
No
Yes
No
No
No
Yes
Yes
Yes
Yes
No
No
No
No
No
No
No
Yes
No
Yes
No
No
Yes
No
No
No
No
No
-------
B-1861
TABLE'
jia.-;:
si&yL clam' |Aty:a"
isldrldicaj-
: Alaska to
^!j^?M^^^M^^^^iiS:
Florida,
Hawaii
(^^
However, macroinvertebrates may not provide sufficient tissue biomass for analyses.
Adult fish of the species listed here provide sufficient tissue biomass for analyses and
arc direct measures of contaminants available to human populations. However, fish
are often motile and bioaccumulation measured may not be representative of the study
area.
-------
B-187
It is recommended that bioaccumulation studies be conducted for a number of species.
Species have different bioaccumulation potentials for various contaminants; monitor-
ing multiple species will ensure that bioaccumulation for a number of contaminants are
sufficiently evaluated. Monitoring of species along a food chain might provide
information concerning the transfer of contaminants to higher trophic levels. Species
with extensive historical bioaccumulation data is also preferred. Where possible,
species being collected for other aspects of the monitoring program (e.g., fisheries
data, pathobiology, pathogens) should be included in the bioaccumulation analyses, to
enhance the coordinated evaluation of data.
Caged Indicator vs. Indigenous Species - The California Mussel Watch Program, the
US Mussel Watch Program, and the National Oceanic and Atmospheric Administra-
tion (NO A A) Status and Trends Program have employed the use of both resident and
caged transplant mussels to monitor bioaccumulation of toxic chemicals over space
and time (Goldberg et al., 1978; Boehm, 1984; Ladd et ai, 1984). Caged indicator
target species offer several advantages over indigenous species:
the biology and ecology of these indicator species are usually well de-
scribed
descriptions of culture and/or maintenance of the organisms under labora-
tory conditions are often available
the method provides control of initial temporal and spatial variation of
individuals and/or biomass
the method allows the use of a specific age, size and/or genetic stocks
the method assures on-site bioaccumulation
By controlling for initial conditions, the rate of change of tissue contamination may be
calculated. In addition, the Long Island Sound Estuary Program has shown that
chemical residue analyses on caged indicator species appears to be a promising
approach for identifying sources of pollution (U.S. EPA, 1982). However, transfer
through the water is the only exposure pathway assessed; adequate methods for caging
sediment ingesting organisms are not yet available. Because bioaccumulation of
sediment-sorbed contaminants is probably mediated by transfer through interstitial
waters (Knezovich and Harrison, 1987) and results obtained often do not relate to
species found on site, caged organisms may not provide an accurate estimate of the
bioavailability of certain contaminants occurring at the site.
-------
B-188
The advantages of indigenous species are:
results obtained will relate directly to those species which may be impacted
and provide a direct measure of potential risks to human health
no cost of transplanting organisms
no possibility of loss of cage/buoy system
no possibility of introduction of a "nuisance" species
However, common indigenous species may not meet the criteria for use as bioaccumu-
lation target species, negating the advantage of using a native species. The most
common problem is collecting sufficient tissue biomass of appropriate indigenous
species across sampling locations throughout the study.
Selection of Tissues
The type and location of tissue analyzed will depend upon the objectives of individual
monitoring programs. It is essential that all analyses be conducted with the same
tissue type from the same species in order that scientifically and statistically valid
comparisons may be made.
Target Tissues - In fish, liver tissue is closely associated with regulation and storage
of many toxic substances (Fowler, 1982). Its high fatty tissue content tends to accu-
mulate hydrophobic contaminants. Thus, contaminant levels in the liver can be used to
estimate the range of contaminants being assimilated. For macroinvertebrates, hepato-
pancreas or digestive gland tissue performs functions analogous to fish liver tissue.
Contaminants in edible muscle tissue represent those contaminants that are retained in
a form that allows transfer to humans. Sampling of muscle tissue is appropriate for
human exposure assessments and quantitative health risk determinations. Within a
fillet, contaminant concentrations may vary, therefore it is recommended that a consis-
tent location within the muscle tissue be analyzed (U.S. EPA, 1989).
Whole body analyses ought to be conducted when predators consume the whole body
of the target organism. If organisms are not cleansed of materials contained in the
digestive tract, contaminants in the gut contents will be included in the analyses. To
provide the most accurate estimate of the total amount of contaminants available to
most macroinvertebrate predators, this type of depuration may not be required.
-------
B-189
Sediment Total Organic Carbon and Acid Volatile Sulfides, and Tissue Lipid
Normalization - Toxic sediment concentrations of hydrophobic contaminants have
been found to be related to the total organic carbon content (TOC) of the sediment
(Karickhoff et al., 1979). The bioavailability of metal contaminants has been found to
be related to the acid volatile sulfide (AVS) concentrations of the sediment (DiToro
et al., in press). TOC and AVS normalization have been conducted to estimate the
concentrations of sediment contaminants which are bioavailable over different sam-
pling locations. Sediment TOC and AVS normalizations are recommended to account
for the variability in bioavailable sediment contaminant concentrations between
locations.
Lipids appear to be a storage site of organochlorines, hydrocarbons, and other hydro-
phobic contaminants in a variety of organisms (de Boer, 1988). In fact, just as TOC
and AVS are considered by many to be the most important parameters in defining
organic and metal concentrations in sediments, lipid content is considered the salient
parameter in defining hydrophobic residue concentrations found in tissues (Phillips
and Segar, 1986). Likewise, tissue lipid normalizations have been suggested to
account for the variability in tissue contaminant concentrations between individuals.
Normalization of sediment contaminant concentrations to TOC, AVS, and tissue
contaminant concentrations to the tissue lipid concentrations have been made to permit
comparisons of toxic chemical residues in tissues between studies, locations, individu-
als, and tissue types. These sediment and tissue normalizations assume the following:
contaminants partition predominantly to sediment TOC, sediment AVS,
and organismal lipids
contaminants partition reversibly between the sediment particles and
organism
rapid steady-state kinetics of contaminants
sediment is the only source for the bioaccumulation
Lipid normalization has been performed in order that individuals with differing body
fat levels may be compared. Again, it is essential that all analyses be conducted with
the same tissue type from the same species in order to ensure that scientifically and
statistically valid comparisons may be made. TOC/lipid normalized accumulation
factors (AF) have also been used to predict tissue residue concentrations (Ferraro
et al., 1990; Lake et al., 1987).
-------
B-190
Three-fold differences in lipid concentration may result as a consequence of various
lipid analysis techniques. Development of protocol standardization or intercalibration
between lipid techniques are required before chemical residues in tissues can be
compared among studies. Microtechniques have been developed for analyses of lipids
from single individuals (Gardner et al., 1986).
Finally, the decision to normalize for TOC, AVS, and percent lipid will depend upon
monitoring program objectives. For example, if the monitoring objective is to identify
"hot spots" - i.e. those areas where high body burdens present a risk to human and
ecological health - normalization may not be appropriate. Whereas, if the objective is
to identify temporal trends in bioaccumulation rates, normalization may be justified.
Selection of Sampling Period
The timing of sampling should be based on biological cycles which influence an
organism's susceptibility to bioaccumulation. The frequency of sampling should be
related to the expected rate of change in tissue concentrations of contaminants. A
consistent sampling period is recommended in order that spatial and temporal compari-
sons may be conducted.
For crustaceans, just after molting, before hardening of the integument occurs restrict-
ing its permeability, there is a significant increase in potential for bioaccumulation of
toxic contaminants.
For many aquatic vertebrate and invertebrate organisms, the reproductive cycle exerts
a major influence on tissue concentrations of many contaminants - especially hydro-
phobic compounds (Phillips, 1980). For many species, lipophilic (hydrophobic)
contaminants are transferred from the muscle and liver to the eggs as lipids are mobi-
lized and transported to the egg during oogenesis (Spies et al., 1988; Gardner et al.,
1985). There is evidence that lipophilic contaminants have deleterious effects on the
developing egg and/or the larvae (Spies et al., 1988; Hansen et al., 1985; Niimi, 1983).
Transfer of contaminants from adult to egg, and its potential impacts to future fisheries
recruitment and fisheries production are under investigation.
It is recommended that target species be sampled when tissue contaminant concentra-
tions are expected to be at their highest level in order to evaluate worst-case scenarios.
For those organisms where the body is consumed whole, contaminant levels are
usually at their highest at or just before spawning. For organisms where the muscle
tissue is consumed, contaminants in muscle tissue usually reach a peak well before
spawning.
-------
B-191
B10.3 Questions to be considered during the choice of an appropriate analytical method
Exist! ng include the parameters of interest, desired detection limits, sample size requirements or
Analytical restrictions, methods of preservation, technical and practical holding times, and matrix
Methods interferences. Several U.S. EPA documents (1986a, 1987a, 1990b) discuss the com-
mon analytical problems encountered during monitoring analyses of tissue samples.
Chemical Residue Analyses
Several factors determine achievable detection limits for a specific contaminant,
regardless of analytical procedure - a list of analytical procedures and U.S. EPA
method numbers is given in Table B10-5. These factors include:
size of the tissue available for analysis - in general, the ability to detect
low contaminant levels is improved with greater amounts of tissue avail-
able for analysis; a minimum of 30 g (wet weight) is usually considered
adequate (U.S. EPA, 1987a)
presence of interfering substances
range of pollutants to be analyzed - an optimal method may be developed
without regard to potential effects on other parameters
level of confirmation - qualitative (e.g. presence or absence) or quantita-
tive (e.g. residue concentrations) analyses
level of pollutant found in the field and in analytical blanks
Selection of appropriate methods ought to be based on a trade-off between full-scan
analyses, which are economical but cannot provide optimal sensitivity for some
compounds, and alternate methods that are more sensitive for specific compounds but
can result in higher analytical costs.
Metals and Metalloids- Nitric acid/perchloric acid digestions should be considered
for analysis of tissue samples (U.S. EPA, 1987a). Perchloric acid is especially useful
for the dissolution of fats.
Trace element analyses by ICP (U.S. EPA method 6010) allows for several elements to
be measured simultaneously. Detection limits of ICP for most metals are generally
comparable to those achieved by GFAAS; however, detection limits for several metals
are significantly lower using AAS: arsenic, selenium, and mercury.
-------
B-1921
GFAAS is the recommended analytical method for detection of metals and metalloids
in tissues (U.S. EPA, 1987a). Cold vapor AAS (EPA method 7470) analysis is the
only recommended technique for mercury (US EPA, 1986a).
Graphite furnace AAS (GFAAS) is more sensitive (i.e. lower detection limits) than
flame AAS, but is more subject to matrix and spectral influences - GFAAS requires
particular caution with regard to laboratory contamination. GFAAS requires more
skilled laboratory technicians. Both AAS methods require that the concentration of
each element be determined by a separate analysis, making the analysis of a large
number of contaminant metals both labor-intensive and relatively expensive compared
to ICP.
Semi-Volatile Organic Compounds - Analysis of semi-volatile organic compounds
involves a solvent extraction of the sample, cleanup of the characteristically complex
extract, GC analysis, and quantification (U.S. EPA, 1987a; U.S. EPA, 1986a). There
are two GC/MS options for detecting extractable organic compounds: internal stan-
dard technique and isotope dilution. The isotope dilution technique is recommended
because reliable recovery corrections can be made for each analyte with a labeled
analog or a chemically similar analog (U.S. EPA, 1986a). The isotope dilution method
is more expensive and less widely employed than the internal standard technique,
which is the current method of choice in the Contract Laboratory Program (CLP).
Mass spectrometry provides positive compound identification by comparison of both
retention time and spectral patterns with standard compounds.
The identification of pesticides and PCB congeners can be made by GC/ECD analysis
(U.S. EPA method 8080). GC/ECD provides greater sensitivity relative to GC/MS,
however GC/ECD does not provide positive compound identification. Confirmation
of pesticides and PCBs on an alternative GC/ECD or preferably by GC/MS, when
sufficient concentrations occur, is recommended for reliable results (U.S. EPA,
1985b). All other organic compound groups are recommended for analysis by GC/
MS.
Volatile Organic Compounds - Analysis of volatile organic compounds also in-
volves a solvent extraction of the sample, cleanup of the characteristically complex
extract, GC analysis, and quantification (U.S. EPA, 1987a; U.S. EPA, 1986a). The
purge and trap GC/MS technique is employed for detecting volatile organic com-
pounds in water.
A successful variation for detection of volatile organic residues in tissues involves a
device that vaporizes volatile organic compounds from the tissue sample under
-------
B-193
TABLE B10-4. LIST OF EXISTING ANALYTICAL TECHNIQUES
(U,S. EPA,
METALS/METALLOIDS
Atomic Absorption Spectrophotometty
(AAS)
-flame
- graphite furnace (GFAAS)
- cold vapor
-gaseous hydride (HYDAAS)
Inductively Coupled Plasma Emission
Spectrometry (ICP)
ORGANICS
* Gas Chromatography (GQ
-with electron capture detection (GC/ECD)
-with mass spectrometry (GC/MS)
US EPA method
7000 series
US EPA method 7470
US EPA methods
7060 and 7740
US EPA method
6010
US EPA method 8080
US EPA methods 8240
and 8270
vacuum and then condenses the volatiles in a super-cooled trap (Hiatt, 1981). The trap
is then transferred to a purge and trap device where it is treated as a water sample. The
isotope dilution option is recommended as it provides reliable recovery data for each
analyte (U.S. EPA, 1986a). However, this technique creates additional analytical time
and expense.
B10.4 QA/QC
Considerations
Sample Collection
In the field, sources of contamination include sampling gear, lubricants and oils,
engine exhaust, airborne dust, and ice used for cooling samples. Samples designated
for chemical analyses require all sampling equipment - e.g., siphon hoses, scoops,
containers - be made of noncontaminating material and be cleaned appropriately prior
-------
B-194
to use. Efforts should be made to minimize handling and to avoid sources of contami-
nation including potential airborne contamination (e.g., stack gases, cigarette smoke).
To avoid contamination from ice, whole samples should be wrapped in aluminum foil,
placed in watertight plastic bags, and immediately cooled in a covered ice chest.
Furthermore, samples and sampling containers should not be touched with ungloved
fingers.
Sample Handling and Storage
For analyses of metals, samples should be frozen and kept at -20ฐC (Table B10-5).
Although specific tissue sampling holding times have not been recommended by U.S.
EPA, a maximum of 6 months (28 days for mercury) would be consistent with that for
water samples. For analyses of volatile compounds, samples should be stored in the
iiiiiii^
a p ji
c Suggested.
|w|^:l||^ii(iง^;:gi:^||feP^
|:;::R|g:;i^plingtiiiie:28-daysป
-------
B-195
dark at 4ฐC (Table B10-6; U.S. EPA, 1987a). Analyses of volatile compounds should
be performed within 14 days of collection as recommended by U.S. EPA (1984). If
analyses of semivolatile compounds will not be performed within the recommended 7-
day holding time, freezing of the samples at -20ฐC is advised. Holding times for
frozen samples has not been established by the U.S. EPA. A general guideline of a
maximum of 6 months would be consistent with that for water samples (U.S. EPA,
1987a).
Resectioning tissue samples for chemical analyses should only be conducted with
noncontaminating tools under "clean room" conditions. It is recommended that whole
tissue samples not be frozen prior to resection if analyses will be conducted only on
selected tissues because freezing may causes internal organs to rupture and contami-
nate other tissues. A separate set of utensils for removing outer tissue and for
resectioning tissue for analysis is highly recommended. Incision troughs are subject to
contamination and should not be included in the analysis.
iMliiiBll
il
llll
Ifiaditicylis!
:*
m
^ipifliijii
isili
::.::ฅ::::::;::::::|:>|:|v|:;:: ::i:v::X::::;::';:-:-'::;::\X:X:::::::'::::;^
-------
B-196
Laboratory Analyses
Laboratory performance as measured by method QA/QC protocols should be used to
evaluate and select appropriate analytical methods. Changes to laboratory protocols
should only be considered if new protocols meet established performance criteria.
Suggested QA/QC protocols and analytical performance criteria are discussed below.
QA/QC reports should describe the results of quantitative QA/QC analyses, as well as
other elements critical to the laboratory analyses of water, sediments, and tissues to
ensure proper interpretation of the results. It is recommended that these reports be
recorded and stored in a database for future reference.
Field QA/QC Checks - Transfer blanks will whether any contamination was intro-
duced of reagents in the field or introduced during shipping of samples. Cross-
contamination blank is designed to verify the absence of contamination carried over
from one sample to another due to inadequate cleaning of field equipment.
Blind replicates, splits treated and identified as separate samples, may be sent to the
same laboratory for analysis or have one sample sent to a "reference" laboratory for
comparison. In addition, standard reference material may also be placed in a sample
container at the time of collection and sent "blind" to the laboratory.
Instrument QA/QC Checks - Calibration standards should be analyzed at the begin-
ning of sample analysis, and should be verified at the end of each 12-hr shift during
which analyses are performed (U.S. EPA, 1987b). The concentrations of calibration
standards should bracket the expected sample concentrations, or sample dilutions or
sample handling modifications (i.e., reduced sample size) will be required.
Method QA/QC Checks - Analysis of preparation blanks should be conducted to
demonstrate the absence of contamination from sampling or sample handling in the
laboratory. At least one method blank must be included with each batch of samples
and should constitute at least 5% of all samples analyzed.
Spike recovery analyses are required to assess method performance for the particular
sample matrix. Spike recoveries serve as an indication of analytical accuracy, whereas
analysis of standard reference materials measure extraction efficiency. Recommended
control limits include 75-125 percent recovery for spikes, and 80-120 percent recovery
for the analysis of standard reference materials (SRM).
-------
B-197
Replicates must be analyzed to monitor the precision of laboratory analyses. A
minimum of 5% of the analyses should be laboratory replicates. The control limits are
ฑ20 percent relative percent difference for duplicates. Triplicates should be analyzed
on one of every 20 samples or on one sample per batch if less than 20 samples are
analyzed.
Tables B10-7 and B10-8 provide a brief summary of QA/QC for laboratory analyses.
A discussion of detection limits may be found in QA/QC Considerations in the Water
Column Chemistry chapter (Section 2.4).
B10.5 Statistical Statistical strategies may mitigate the high costs of collecting sufficient tissue biomass.
Design See also Statistical Design Considerations: Composite Sampling, Power Analysis, and
Considerations Power-Cost Analysis (Appendix B Introduction; Section B.3).
B10.6
Use of Data
Results of the bioaccumulation analyses can be used to establish spatial and temporal
trends in the bioaccumulation of toxicants of selected estuarine fish and
macroinvertebrates. Geographic maps of these trends would provide a survey of
trends throughout the estuary.
In addition, monitoring studies might identify existing and potential problem areas for
fish and macroinvertebrate contamination. Human health risk, as well as environmen-
tal risk assessment calculations may also be determined from data collected from these
studies. The information may be used to develop management strategies for commer-
cial and recreational fishing.
B10.7
Summary and
Recommendations
Rationale
Presence of toxics in waters and sediments of estuaries can have adverse
ecological and human health effects
Monitoring the accumulation of chemical residues in tissues of estuarine
organisms will provide information essential in relating the presence of
selected contaminants in estuarine waters and sediment to its transfer and
accumulation in estuarine organisms
-------
B-198
TABLE BIO-?. SUMMARY OF QUALITY CONTROL
: Analysis Type
Friequency
||yil|b|ate spikes
x spikes .
^^^^^jji^^j^sM^
\o^l^jMf3^^&^MKjjfi.l
'^j^te^^j^&l^^^j^i^^j^j^^j^l.
|^|||ab^|rii|uii|i|
:i^i^i^''i)j^งmyijjim
s^;:io|'sairtples:'(i^ง:iw
il^aftefafix me^dwoti&MuWl
triplicates
iliniHi tijlil
Monitoring program may provide information necessary to assess environ-
mental and human health risks associated with measured levels of bioaccu-
mulation
-------
B-199
TABLE B10-8. SUMMARY OF WARNING
AND CONTROL LIMITS FOR QUALITY CONTROL SAMPLE
Analysis Type
Recommended
Warning Limit
Recommended
Control Limit
Surrogate Spikes
Method Blank
Phthalate,
Acetone
Other Organic
Compounds
Standard
Reference Materials
Matrix Spikes
Spiked Method Blanks
Analytical Replicates
Field Replicates
Ongoing Calibration
10 percent recovery
30 percent of the analyte
1 p.g total or 5 percent
of the analyte
95 percent
confidence interval
(50-65 percent recovery)
(50-65 percent recovery)
(50 percent recovery)
5 jig total or 50 percent
oftheanalyte
2.5 |lg total or 5 percent
oftheanalyte
95 percent confidence
interval for Certified
Reference Material
(50 percent recovery)
(50 percent recovery)
ฑ100 percent
coefficient of variation
25 percent of
initial calibration
Monitoring Design Considerations
It is recommended that consistent types of target species and tissues, data
normalization, and location and timing of sample collection be imple-
mented to allow for comparisons among studies
Target species should possess the following characteristics:
- high bioaccumulation potential for selected contaminants of
concern
-------
B-200 I
- metabolic regulation of selected contaminants should be weak or
absent
abundant, temporally and spatially, to allow for adequate
sampling
- large enough to provide adequate amounts of tissue for analysis
- sessile, or sedentary in nature to assure bioaccumulation is
representative of the study area
- easily collected
Analyses should be conducted on a number of different target species to
ensure sensitive measures of bioaccumulation for the contaminants of
concern
Caged indicator species
- method allows for control of initial temporal and spatial variation
of individuals and/or biomass
- method allows for the use of specific age, size and/or genetic
stocks
Indigenous species
- results obtained will relate directly to those species which may be
impacted
- no possibility of loss of holding apparatus and test organisms
Target tissues
- fish liver or macroinvertebrate hepatopancreas analyses may be
used to estimate the range of contaminants being assimilated
- muscle tissue analyses are appropriate for human exposure
assessments and quantitative health risk determinations
whole body analyses ought to be conducted when predators
consume the whole body of the target organism
TOC/AVS/Lipid Normalization
- recommend normalization to allow comparisons of chemical
residue concentrations in tissues between locations, individuals,
and tissue type - normalizing the data will depend upon the
objectives of the individual monitoring program
- differences in lipid analysis techniques will result in differences
in lipid detected - caution in comparing data between analytical
methods
-------
B-201
- ^ , * ~ป, * <, ; -ojv, vs*;j ^:-^l -^V^^ltlM
Time of Sampling
- hydrophobia contaminants are mobilized and transported to the
egg during oogenesis
- timing of sampling should be based on biological cycles which
influence an organism's susceptibility to bioaccumulate
- it is recommended that target species be sampled when tissue
contamination concentrations are expected to be at their highest
level
Existing Analytical Methods
It is recommended that consistent types of analytical protocols be imple-
mented to allow for comparisons among studies
Metals/Metalloids
- GFAAS is the recommended method for the detection of metals
and metalloids (U.S. EPA, 1987a)
- cold vapor AAS is the recommended protocol for mercury
detection
Organics
- GC/MS in conjunction with isotope dilution is recommended for
the detection of semi-volatile organic compounds
- vacuum super-cooled trap in conjunction with a purge and trap
device is recommended for the detection of volatile organics
(Hiatt, 1981)
- isotope dilution option is recommended as it provides reliable
recovery data for each analyte (U.S. EPA, 1986a)
QA/QC Considerations
Reports delineating the essential elements of the bioaccumulation compo-
nent of the program should be included with the quantitative QA/QC
analyses
Statistical Design Considerations
Six individuals comprising each of five replicate composites should be
adequate to detect a treatment difference equal to 100 percent of the overall
mean among treatments
-------
B-202
Use of Data
Establish spatial and temporal trends in the bioaccumulation of toxicants of
selected estuarine fish and macroinvertebrates
Identify existing and potential problem areas for fish and macroinverte-
brate contamination
Supply data which can be used to calculate the human health risk of
consuming estuarine fish and shellfish
Boehm, P.D. 1984. The Status and Trends Program: Recommendations for design
and implementation of the chemical measurement segment. Workshop Report.
Rockville, MD: NOAA.
deBoer, J. 1988. Chlorobiphenyls in bound and non-bound lipids of fishes: Compari-
son of different extraction methods. Chemosphere 17:1803-1810.
DeWitt, T. 1991. Hatfield Marine Science Center, Newport, Oregon. Personal
communication.
DiToro, D.M, J.D. Mahony, D.J. Hansen, K.J. Scott, A.R. Carlson, and G.T. Ankley.
In Press. Acid volatile sulfide predicts the acute toxicity of cadmium and nickel in
sediments.
Harrington, J.W., E.D. Goldberg, R.W. Risebrough, J.H. Martin, and V.T. Bowen.
1983. U.S. "Mussel Watch" 1976-1978: An overview of the trace-metal, DDE, PCB,
hydrocarbon, and artificial radionuclide data. Environ. Sci. Technol. 17:490-496.
Ferraro, S.P., H. Lee, R.J. Ozretich, and D.T. Specht. 1990. Predicting bioaccumula-
tion potential: A test of a fugacity-based model. Arch. Environ. Contain. Toxicol.
19:386-394.
B10.8 Literature
Cited and
References
Fowler, S.W. 1982. Biological transfer and transport processes. In: Pollutant Trans-
fer and Transport in the Sea, vol 2. (Kullenberg, G., ed). Boca Raton, FL: CRC Press
Gardner, W.S., T.F. Nalepa, W.A. Frez, E.A. Cichocki, and P.P. Landrum. 1985.
Seasonal patterns in lipid content of Lake Michigan macroinvertebrates. Can. J. Fish.
Aquat. Sci. 42:1827-1832.
-------
B"203
Gardner, W.S., W.A. Frez, E.A. Cichocki, and C.C.Parrish. 1986. Micromethod for
lipids in aquatic invertebrates. Limnol. Oceanogr. 30:1099-1105.
Goldberg, E.D., V.T. Bowen, G.H. Farrington, J.H. Martin, P.L. Parker, R.W.
Risebrough, W.Robertson, E.Schneider and E. Gamble. 1978. The mussel watch.
Environ. Conserv. 5:101-125.
Hansen, P.D..H. Von Westerhagen, andH. Rosenthal. 1985. Chlorinated hydrocar-
bons and hatching success in Baltic herring spring spawners. Mar. Environ. Res.
15:59-76.
Hiatt, M.H. 1981. Analysis of fish and sediment for volatile priority pollutants. Anal.
Chem. 53:1541-1543.
Karickhoff, S.W., D.S. Brown, and T.A. Scott. 1979. Sorption of hydrophobic
pollutants on natural sediments. Wat. Res. 13:241-248.
Knezovich, J.P. and F.L. Harrison. 1987. A new method for determining the concen-
tration of volatile organic compounds in sediment interstitial water. Bull. Environ.
Contam. Toxicol. 38:837-940.
Ladd, J.M., S.P. Hayes, M. Martin, M.D. Stephenson, S.L. Coale, 3. Linfield, and M.
Brown. 1984. California state mussel watch: 1981-1983. Trace metals and synthetic
organic compounds in mussels from California's coast, bays, and estuaries. Biennial
Report. Sacramento, CA: Water Quality Monitoring Report No. 83-6TS.
Lake, J.L., N.I. Rubinstein, and S. Pavignano. 1987. Predicting bioaccumulation:
Development of a partitioning model for use as a screen tool in regulating ocean
disposal of wastes. In: Fate and Effects of Sediment-bound Chemicals In Aquatic
Systems. (Dickson, K.L., A.W. Maki, and W.A. Brungs, eds). Florissant, CO: Sixth
Pellston Workshop.
Landrum, P.P. and J.A. Robbins. 1990. Bioavailability of sediment-associated
contaminants to benthic invertebrates. In: Sediments: Chemistry and Toxicitv of In-
Place Pollutants (Baudo, R., J.P. Giesy, and H. Muntau, eds). Lewis Publishers
Lee, H., II, B.L. Boese, J. Pelletier, M. Winsor, D.T. Specht, and R.C. Randall. 1989.
Guidance manual: Bedded sediment bioaccumulation tests. EPA ERLN Document
Nl 11, Pacific Ecosystems Branch, Newport, OR.
-------
B-204
Lee, H., II, M. Winsor, J. Pelletier, R.C. Randall, J. Bertling and B. Coleman. 1990.
Computerized risk and bioaccumulation system (Ver. 1.0). EPA ERLN Document
N137, Pacific Ecosystems Branch, Newport, OR.
Niimi, A.J. 1983. Biological and lexicological effects of environmental contaminants
in fish and their eggs. Can. J. Fish. Aquat. Sci. 40:306-312.
Pearson, T.H. and R. Rosenberg. 1978. Macrobenthic succession in relation to
organic enrichment and pollution of the marine environment. Oceanogr. Mar. Biol.
Ann. Rev. 16:229-311.
Phillips, D.J.H. 1980. Quantitative Aquatic Biological Indicators. London, England:
Applied Science Publ. Ltd.
Phillips, D.J.H. and D.S. Segar. 1986. Use of bio-indicators in monitoring conserva-
tive contaminants: Programme design imperatives. Mar. Poll. Bull. 17:10-17.
Plumb, R.H. 1981. Procedure for handling and chemical analysis of sediment and
water samples. Technical Report EPA/CE-81-1. U.S. EPA and Corps of Engineers,
U.S. Army Engineers Waterways Experimental Station, Vicksburg, MS.
Rubinstein, N.I., J.L. Lake, R.J. Pruell, H. Lee, B. Taplin, J. Heltshe, R. Bowen and S.
Pavignano. 1987. Predicting bioaccumulation of sediment-associated organic con-
taminants: Development of a regulatory tool for dredged material evaluation. Internal
Report. Narragansett, RI: US EPA.
Spies, R.B., D.W. Rice and J. Felton. 1988. Effects of organic contaminants on
reproduction of the starry flounder Platichthys stellatus in San Francisco Bay. Mar.
Biol. 98:181-189.
U.S. EPA. 1982. Method for use of caged mussels to monitor for bioaccumulation
and selected biological responses of toxic substances in municipal wastewater dis-
charges to marine waters. Draft. Environmental Monitoring Science Laboratory,
Cincinnati, OH.
U.S. EPA. 1984. U.S. EPA contract laboratory program statement of work for
organics analysis, multi-media, multi-concentration. IFB WA 85-T176, T177, T178.
Washington, DC.
-------
> v "
-'& v,\'\-."''. *" A'^ "^fv.v. frf'Sfw
B-205
U.S. EPA. 1985a. Bioaccumulation monitoring guidance: Selection of target species
and review of available bioaccumulation data. vol. 2. EPA 403/9-86-006. Office of
Marine and Estuarine Protection, Washington, DC. 52 pp.
U.S. EPA. 1985b. Bioaccumulation Monitoring Guidance: Recommended Analytical
Detection Limits, vol. 3. EPA 503/6-90-001. Office of Marine and Estuarine Protec-
tion, Washington, DC. 23 pp.
U.S. EPA. 1985c. Bioaccumulation monitoring guidance:Estimating the potential for
bioaccumulation of priority pollutants and 301(h) pesticides discharged into marine
and estuarine waters, vol 1. EPA 503/3-90-001. Office of Marine and Estuarine
Protection, Washington, DC. 56 pp.
U.S. EPA. 1985d. Contract laboratory program statement of work, inorganic analysis,
multi-media, multi-concentration. SOW No. 785. U.S. EPA, Washington, DC.
U.S. EPA. 1986a. Bioaccumulation Monitoring Guidance: Analytical methods for
U.S. EPA priority pollutants and 301 (h) pesticides in tissues from estuarine and marine
organisms. Vol. 4. EPA 503/6-90-002. Office of Marine and Estuarine Protection,
Washington, DC. 108pp.
U.S. EPA. 1986b. Test methods for evaluating solid wastes, physical/chemical
methods. SW-846, 3rd Edition. Environmental Protection Agency, Washington, DC.
U.S. EPA. 1987a. Quality Assurance/Quality Control (QA/QQ for 301(h) Monitor-
ing Programs: Guidance on Field and Laboratory Methods. EPA 430/9-86-004.
Office of Marine and Estuarine Protection, Washington, DC.
U.S. EPA. 1987b. Puget Sound protocols. Final Report. 31pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1987c. Bioaccumulation Monitoring Guidance: Strategies for Sample
Replication and Compositing, vol.5. EPA 430/9-87-003. Office of Marine and
Estuarine Protection, Washington, DC. 51 pp.
U.S. EPA. 1987d. Technical support document for ODES statistical power analysis.
EPA 430/9-87-005. Office of Marine and Estuarine Protection, Washington, DC.
34pp.
-------
B-206!
U.S. EPA. 1989. Assessing human health risks from chemically contaminated fish
and shellfish: A guidance manual. EPA 503/8-89-002. Office of Marine and Estuarine
Protection, Washington, DC. 136 pp.
U.S. EPA. 1990a. Assessment and Control of Bioconcentratable Contaminants in
Surface Waters, (draft). Office of Water Enforcement and Permits, Washington, D.C.
U.S. EPA. 1990b. Analytical procedures and quality assurance plan for the determi-
nation of xenobiotic chemical contaminants in fish. EPA 600/3-90-023. Office of
Research and Development, Environmental Research Laboratory, Duluth, MN. 23 pp.
Young, D.R., A.J. Mearns, and R.W. Gosset. 1990. Bioaccumulation and
biomagnification of DDT and PCB residues in a benthic and a pelagic food web of
Southern California.
-------
B-207
"S
B11.0 Bacterial and Viral Pathogens
Monitoring for pathogenic microorganisms is currently conducted by state environ-
mental and human health agencies in shellfish harvesting areas and bathing beaches.
National Pollution Discharge Elimination System (NPDES) pathogen monitoring
programs may be underway, at selected locations in the estuary, in order to assess both
the condition of water in the vicinity of discharges and surrounding areas and to assess
relative pathogen contributions from these permitted effluent discharges.
The National Shellfish Sanitation Program (NSSP) has been established as a joint
effort between the Food and Drug Administration (FDA), state agencies, and the
shellfish industry to set forth guidelines for the management of state shellfish pro-
grams. As part of the NSSP, FDA provides technical assistance to states in studying
specific pollution problems; provides data to establish closure levels for shellfish
harvesting; conducts applied research on various contaminants to assist in developing
standards and criteria; and evaluates the effectiveness of state shellfish sanitary control
programs. In addition, since 1966, data have been compiled periodically by FDA and
the National Oceanic and Atmospheric Administration (NOAA) on classification by
states of coastal and estuarine waters with regard to suitability for shellfishing activi-
ties. In addition to classifying their waters as to their suitability as shellfish harvesting
areas, states also issue beach closures. These closures are typically based on water
quality criteria developed by the Federal government.
B11.1 Human pathogens found in the estuarine environment include viruses and bacteria.
Rationale In the United States, viruses and bacteria are the most important human pathogens,
both in terms of the number of organisms released to the environment and in the
severity of the diseases they cause (OTA, 1987).
Humans can be exposed to pathogens by direct contact with contaminated waters (e.g.,
swimming, surfing, diving) and/or indirectly through ingestion of contaminated food
(e.g., molluscan shellfish). The preponderance of evidence indicates that the etiologic
agent for waterbome outbreaks of acute gastroenteritis (AGI) are the Norwalk-like
viruses (Kaplan et al., 1982). Hepatitis A has been linked to the consumption of raw
or partially cooked molluscan shellfish (Feingold, 1973; CDC, 1979; Ohara et al.,
1983). Bacteria responsible for typhoid and cholera are known to be water- and
seafood-borne. These pathogens can enter the estuarine environment through the
discharge of raw sewage, wastewater effluent from sewage treatment plants, failing
septic tanks, and the dumping of sewage sludge (OTA, 1987).
Monitoring of human pathogens provides information essential in relating the presence
of pathogens in marine waters and shellfish to outbreaks of disease. Monitoring data
-------
B-208
may be used to identify potential sources of pathogens. The assessment of pathogen
contamination should be a component of a monitoring program where pathogens may
present a risk to human health and economic vitality of an estuary.
Water Column Sampling
Bacteria are not uniformly distributed throughout the water column (Gameson, 1983);
bacterial abundances, several orders of magnitude greater than underlying waters, can
be found in a thin microlayer on the surface of the water (Hardy, 1982). If feasible, it
is recommended that samples of this microlayer be collected; separate samples of
underlying waters should also be sampled. However, standardized methods for
sampling this microlayer have not been established. If it is infeasible to sample both
the microlayer and underlying waters, the "scoop" method should be used to ensure
that the surface microlayer is sampled (U.S. EPA, 1978).
Water samples for analyses of bacterial analyses are frequently collected using steril-
ized plastic bags (e.g., Whirl-pak) or screw-cap wide-mouthed bottles. Several depths
may be sampled during one cast and/or replicate samples may be collected at a particu-
lar depth by using a Kemmerer or Niskin samplers (U.S. EPA, 1978). Any device
which collects water samples in unsterilized tubes should not be used for collecting
bacteriological samples without first obtaining data that support its use. Pumps may
be used to sample large volumes of the water column (U.S. EPA, 1978).
Sentinel Organisms
Analyses of sentinel shellfish tissue (e.g., mussels and oysters) offer several advan-
tages:
they concentrate pathogens; they may be useful for viral analyses since
viruses are often present in low numbers
they provide a means of temporally integrating water quality conditions
they may be direct measures of human exposure to pathogens
(i.e., consumption of contaminated food)
they may be deployed and maintained in a number of diverse locales
B11.2
Monitoring Design
Considerations
-------
B-209
However, the disadvantages of sentinel organisms include:
the cost of transplanting organisms
the possibility of loss of the cage/buoy system
species may not be tolerant of all test site conditions (e.g., low salinity,
high turbidity)
It is essential that monitoring design elements be standardized to allow for compari-
sons among estuarine studies. The selection of a standard sentinel species is recom-
mended; interspecies differences would not allow comparisons of body burdens.
B11.3 Many pathogens can be present in wastes, contaminated media, and infected organ-
Analytical isms. Human pathogens are of concern in estuarine waters because they are:
Methods
Considerations ซ associated with major debilitating diseases (e.g., hepatitis, cholera)
infectious at low doses
resistant to environmental stress
not readily enumerated due to low numbers
Table Bl 1-1 provides examples of pathogenic organisms known to cause adverse
human health effects. The alternative to directly identifying pathogens of concern has
been the enumeration of bacteria which are indicators of human waste contamination
or indicators of human illness.
Indicators of Human Health Risks
Due to the inability to enumerate pathogens of concern, indicators of human pathogen
densities have been used in order to assess human health risks. Indicators useful in
predicting infectious disease rates should have the following characteristics:
high abundances should be consistently found in human fecal wastes
should not have significant extra-human fecal sources
-------
B-2101
JปV\ >^ x., >
TABLE B1M,
MICROORGANISMS RESPONSIBLE FOB CAUSING ADVERSE
HUMAN HEALTH EFFECTS* (NOAA, 1988)
Disease
Pathogenic
Organism
Seafood
Source
Hepatitis Hepatitis A virus
Non-A and non-B hepatitis
Gastroenteritis Aeromonas hydrophitia
and Plesiomonas sMgelloides
Vibrio mimicus
Vibrio parahaemiolyticus
Vibrio vulnificus
Vibrio cholera 0 group
Vibrio cholera, Non-O group 1
Norwalk virus
Small round structured virus
Campy lobacter jejuni
Raw oysters
Steamed and raw clams
Cockles
Raw molluscan shellfish
Shellfish
Raw oysters
Clams and snails
Raw oysters
Crab
Shrimp
Lobster
Raw oysters
Raw oysters
boiled shrimp, boiled crab
Raw oysters
Raw oysters
Raw clams
Raw oysters
Raw clams
* Includes naturally occurring microorganisms as well as microorganisms associated
with pollution.
-------
fl-> rf%
B-211
% -v \ v> vwvix
,% ,-v . v ซ ^
s \*
must provide temporally and spatially reliable and accurate appraisals of
the pathogen of concern
Currently, there is no consensus on which indicator organism, if any, is specific for
human feces.
Coliform Bacteria - For decades, the concentration of coliform bacteria, either total or
fecal coliform, has been considered a reliable indicator of the presence and densities of
pathogens. The use of total coliform bacterial criteria to protect human health from
disease associated with contaminated water is widespread. Furthermore, total coliform
concentrations have the advantage of providing a basis for comparison with historical
data (U.S. EPA, 1988).
However, these indicators and their corresponding water quality standards have not
been related to incidences of disease through epidemiological studies. Controversy
exists on the efficacy of coliform bacterial to predict the presence and inactivation of
other types of pathogens (Pederson, 1980). In fact, recent studies indicate that fecal
coliform bacteria may not be a reliable indicator for predicting the risks associated
with direct exposures to pathogens in the marine environment (Cabelli et cd., 1979 and
1982). Fecal coliforms are not pathogenic, and are less resistant to environmental
stress compared to many pathogens (Borrego et a/., 1983). Furthermore, fecal colif-
orm bacteria are not specific to mammalian fecal pollution. The lack of this specificity
prompted the development of methods for enumerating fecal coliform bacteria specific
to mammalian fecal pollution (e.g., Escherichia coli). However, with respect to
recreational water quality criteria, the U.S. EPA has not recommended the use of E.
coli for marine and estuarine waters.
Enterococci - Enterococci are streptococcus bacteria indigenous to the intestines of
warm-blooded animals. Cabelli et al. (1983) found that the densities of enterococci
were highly correlated with the incidence of gastrointestinal symptoms (GI) among
swimmers; reported swimming-associated GI symptoms were poorly correlated with
fecal coliform densities. Enterococci also have the following advantageous character-
istics:
tolerant to high salinity and are of particular value in the analysis of
estuarine and marine waters
taxonomic identifications are relatively simple and can reveal the kinds of
mammalian pollution (e.g., humans, livestock)
-------
B-212
genetic fingerprinting techniques have been developed which can link
these bacteria in the environment to specific sources of contamination
State-of-the-art genetic fingerprinting techniques are very costly and require further
study in order to assess their reliability. The U.S. EPA has adopted enterococci as an
indicator of microbiological water quality for recreational marine waters.
Viruses - Viruses are being recognized as major etiologic agents for many outbreaks
of human illness as a result of swimming in contaminated water and consumption of
contaminated seafood. Viruses are excreted only by infected individuals; they are not
normal flora in the intestinal tract. Traditional indicators of microbial contaminants
appear to be inadequate for predicting human health risks associated with both con-
suming molluscan shellfish and swimming in waters that contain sewage associated
viruses. However, examination of water for enteric viruses is not recommended at this
time, except in special circumstances, due to limitations of methodologies. Even state-
of-the-art methods for concentrating viruses from water still are being researched and
continue to be modified and improved. None of the available virus detection methods
have been tested adequately with representatives from all of the virus groups of public
health importance. In addition, some of these methods require expensive equipment
and materials for sample processing and all virus assay and identification procedures
require expensive cell culture and related virology laboratory facilities.
Laboratory Techniques
It should also be noted that no single procedure is adequate to isolate all microorgan-
isms from water, and the presence of one microorganism does not signify the presence
or absence of any other (Table B11-2). A more detailed description of analytical
methods for a number of pathogens is presented in Standards Methods for the Exami-
nation of Water and Wastewater (Clesceri et al., 1989).
Fecal Bacteria - Two standard methods are presented here for the detection of fecal
bacteria: membrane filter procedure, and multiple-tube fermentation procedure.
The membrane filter (MF) technique involves sample filtration followed by direct
plating for detection and enumeration of coliform bacterial densities. The MF tech-
nique can be used to test relatively large volumes of samples, and yields numerical
results more rapidly than the multiple-tube procedure. The statistical reliability of the
MF technique is greater than the Most Probable Number (MPN) procedures (Clesceri
et ai, 1989). However, the MF technique has limitations, particularly in testing waters
with high turbidity and noncoliform (background) bacteria. The MF technique can be
-------
B-213
%-r
\ ' -4J.^\
TABLE Bli-2*
LABORATORY PROCEDURES FOR BACTERIAL INDICATORS
Laboratory Procedures
Tesf Organisms
Water
Sediment
Tissue
Fecal colifbrm
bacteria
Fecal coHfoim
bacteria/E. coli
Enterococci
C, perfringens*
MPN tubes using A-I
broth (Oescerieroi,
1989) (fecal coliform
bacteria/100 mL)
MPN tubes using A-I
broth (Descent a!.,
1989) (fecal conform
bacteria/100 mL)
MPN tubes using EC
broth (Oesceri era!,,
1989) (fecal coltforai
bacteria/lOOmL)
1981)(E.coli/100mL)
mE (Levin etui., 1975)
(enterococci/100 mL)
MPN tubes using iron
milk (St. John etal,r
MPN tubes using iron
milk (St. John et at.,
MPN tubes using ton
milk
-------
B-214
*" %s "" ,. "*..** -. O V "" ^5^ M^? . ,
^V. w
" v> ^ %^
^s *0 * " "V3"-
v. v^\% *. ^ 'v \^% %-WX"-%v
result in underestimations of actual bacterial density. This technique is commonly
used in assessing bacterial levels in shellfish.
The most probably number (MPN) index is an index of the number of coliform bacte-
ria that, most probably, would produce the results observed in the laboratory examina-
tion (Clesceri et al., 1989). Values for the MPN index can be obtained from standard
tables based on the results of the multiple-tube fermentation technique or from Tho-
mas' formula (Clesceri et al.y 1989).
Viruses - Detecting viruses in estuarine waters requires collecting a representative
water sample, concentrating viruses in the sample, and identifying and estimating
abundances of these concentrated viruses. Difficulties in detecting virus abundances
include:
vi ruses are very sm all (20-100 nm)
virus concentrations in waters are spatially and temporally variable, and
typically low
dissolved and suspended material in the sample interferes with accurate
virus detection
Because virus concentrations may be very low, significant volumes of water may be
required (e.g., on the order of tens to thousands of liters). Dose-response curves are
lacking for most pathogens; however, as few as 10 to 100 bacteria are capable of
inducing disease (OTA, 1987).
Three different techniques for concentrating and enumerating viruses are presented in
Standard Methods (Clesceri etal., 1989):
adsorption to and elution from microporous filters
aluminum hydroxide adsorption and precipitation
polyethylene glycol hydroextraction-dialysis
The adsorption and elution technique pressure-filters viruses on microporous filters
and then elutes them from the filter in a small liquid volume. Generally, two types of
filters are available: electronegative and electropositive filters. Currently, insufficient
-------
B-215
documentation on the efficacy of electropositive filters exists. Limitations to this
technique include:
clogging of adsorbent filter by suspended material
dissolved colloid material may interfere by competing with viruses for
adsorption sites on the filter
viruses adsorbed to suspended material may be removed during suspended
material clean-up procedures
Methods for recovering solids-associated viruses is found in Standard Methods (Cles-
ceri et al, 1989). In spite of these limitations, adsorption and elution technique remain
the most promising technique for detecting viruses.
The aluminum hydroxide adsorption and precipitation, and the polyethylene glycol
hydroextraction-dialysis techniques are used to reconcentrate viruses in proteinaceous
and organic buffer eluates. These methods may be used to concentrate viruses in
waters having high virus densities (e.g., wastewaters). They may also be used as a
second-step concentration procedure following processing of large fluid volumes
through microporous filters. However, these two techniques are impractical for
primary processing of large fluid volumes (desceri et al., 1989).
New Techniques - Traditional techniques lack the ability to detect apparently viable,
but non-culturable microorganisms. Recently developed monoclonal antibody and
gene probe techniques permit the detection and enumeration of both culturable and
non-culturable microorganisms. Since these methods have the ability to detect non-
culturable organisms, they may serve as more precise techniques for monitoring
microbiological water quality. Furthermore, these new techniques will allow direct
monitoring of pathogens of concern; the promise of monitoring specific pathogens,
such as Salmonella, Shigella, Giardia, or Legionella may soon be realized.
Sample Handling
Preservation and storage of water samples can be significant sources of error. Sample
bottles must be resistant to sterilization procedures.
Samples should be refrigerated (1-4ฐC) during transport to a laboratory and analyzed
within six hours of collection (U.S. EPA, 1978).
-------
B-2161
It is recommended that sterile distilled water be transported to the field, transferred to a
sample bottle, and processed routinely to ensure that samples were not contaminated
during collection and transport. It is also recommended that ten percent of the samples
be analyzed in duplicate. Furthermore, ten percent of the samples should be split and
analyzed by two or more laboratories.
Intralaboratory and interlaboratory quality control practices should be documented and
QA/QC reports should be available for inspection. Further recommended quality
assurance guidelines for a microbiology laboratory are available (Inhorn, 1977;
Bordner et ai, 1978) and are also discussed in Standard Methods (Clesceri et al,
1989).
B11.4 QA/QC
Considerations
Bacterial counts often are characterized as having a skewed distribution because of
many low values and a few high ones (Clesceri et al., 1989). Application of paramet-
ric statistical techniques requires the assumption of symmetrical distributions such as
the normal curve. An approximately normal distribution can be obtained from posi-
tively skewed data by converting numbers to their logarithms (Clesceri etal., 1989).
Accordingly, the preferred statistic for measuring central tendency of microbiological
data is the geometric mean.
Consideration of statistical strategies will mitigate the high costs of collecting and
processing samples. Analyses of power-costs are necessary in selecting appropriate
sample/replicate number, sample location and sampling frequency (Ferraro et al.,
1989). See also Statistical Design Considerations: Composite Sampling, Power
Analysis, and Power-Cost Analysis (Appendix B Introduction; Section B.3).
The assessment of pathogen contamination should be a component of a monitoring
program where pathogens may be present a risk to human health and the economic
vitality of an estuary. Monitoring of human pathogens provides information essential
in relating the temporal and spatial distribution of infectious agents in estuarine waters
and shellfish to the epidemiology of pathogens of concern (e.g., affected human
populations, locations, and timing of the outbreak).
Furthermore, monitoring data can be used to identify discharges which may be signifi-
cant sources of pathogens and to ensure that water quality standards are maintained.
Monitoring data can also be used to verify fate and transport, human health risk
assessment, and epidemiological modeling predictions.
B11.5
Statistical Design
Considerations
B11.6
Data Use
-------
B-217
B11.7
Summary and
Recommendations
Rationale
The objective is to detect and describe spatial and temporal changes in
abundances of indicators of human health pathogens
Monitoring indicators of human pathogens provides information essential
in relating the presence of infectious agents in estuarine waters and shell-
fish to the incidence of disease outbreaks and potential sources of these
agents.
Monitoring Design Considerations
If feasible, samples of the surface microlayer be collected separate from
samples of underlying waters. If not feasible, samples containing both
underlying and microlayer waters should be collected. It is highly recom-
mended that consistent types of sampling protocols be implemented to
allow for comparisons among studies
Analysis of tissues of sentinel organisms (e.g., mussels and oysters)
confers the following advantages:
- they concentrate pathogens
- they provide a means of temporally integrating water quality
conditions
- they may be deployed and maintained in a number of locales
Analytical Methods Considerations
Indicators of human pathogens should have the following characteristics:
- be consistently found in high abundance in human fecal wastes
- should not have significant extra-human fecal sources
- must provide temporally and spatially reliable and accurate
appraisals of the pathogen of concern
Fecal Coliform Bacteria Densities
- may not be a reliable indicator for predicting the risk associated
with direct exposures to pathogens
- this measure provides a means to compare to historical data
- Laboratory analyses include membrane filter technique and
multiple-tube fermentation technique
-------
B-218 I
Enterococci Densities
are highly correlated with the incidence of gastrointestinal
symptoms
- taxonomic identifications are relatively simple and can reveal the
kinds of mammalian pollution
- state-of-the-art genetic fingerprinting techniques can link bacteria
in the estuary to specific sources; however these techniques are
very costly and require further study in order to assess their
reliability
- Laboratory analyses include membrane filter technique
Virus Densities
- recognized as major etiologic agents for many outbreaks of
human illness
- examination of water samples for enteric viruses is not recom-
mended due to limitations of methodologies
- Laboratory analyses include absorption to an elution form
microporous filters, aluminum hydroxide absorption-precipita-
tion, and polyethylene glycol hydroextraction-dialysis
Monoclonal antibodies and gene probes show promise in detecting and
enumerating both culturable and non-culturable microorganisms
QAIQC Considerations
Sterile distilled water should be transported and transferred to sample
bottles in order to assess contamination during collection and transport
Ten percent of the samples should be analyzed in duplicate
Ten percent should be split and analyzed at two or more laboratories
Statistical Design Considerations
Preferred statistic for measuring central tendency of microbiological data is
the geometric mean
-------
B-219
Use of Data
Provide essential information in order to assess threats to human health
Establishes temporal and spatial trends in pathogen densities and identifies
potential relationships between the presence of the indicator and incidence
of human illness
Provides data which can be used to verify fate and transport, human health
risk, and epidemiology models
B11.8 Andrews, W.H. and M. W. Presnell. 1972. Rapid recovery of Escherichia coli from
Literature Cited estuarine water. Appl. Microbiol. 23:521.
and References
Bisson, J.W. and VJ. Cabelli. 1979. Membrane filtration enumeration method for
Clostridium perfringens. Appl. Environ. Microbiol. 37:55-66.
Bitton, G., B.N. Feldberg, and S.R. Farrah. 1979. Concentration of enteroviruses from
seawater and tap water by organic flocculation using non-fat dry milk and casein
water. Air Soil Pollut. 10:187.
Booze-Allen and Hamilton. 1983. A background document on pathogenic organisms
commonly found in municipal sludge. Prepared for ECAO, US EPA. Cincinnati, OH.
Bordner, R.H., J.A. Winter, and P.V. Scarpino, eds. 1978. Microbiological Methods
for Monitoring the Environment, Water and Waste. EPA 600/8-78-017. Environmen-
tal Monitoring and Support Lab., U.S. Environmental Protection Agency, Cincinnati,
Ohio.
Borrego, J.J., F. Arrabal, A. de Vicente, L.F. Gomez, and P. Romero. 1983. Study of
microbial inactivation in the marine environment. J. Water Pollut. Control Fed.
55:297-302.
Brezenski, F.T. and J.A. Winter. 1979. Use of the delayed incubation membrane filter
test for determining coliformNbacteria in sea water. Water Res. 3:583.
Buras, N. 1974. Recovery of viruses from waste-water and effluent by the direct
inoculation method. Water Res. 8:19.
-------
B-220
Cabelli, VJ. 1983. Health Effects Criteria for Marine Waters. EPA 600/1-80-0431,
U.S. Environmental Protection Agency, Cincinnati, Ohio.
Cabelli, V.J., A.P. DuFour, M.A. Levin, L.J. McCabe, and P.W. Haberman. 1979.
Relationship of microbial indicators to health effects at marine bathing beaches.
Am. J. Public Health 69:690-696.
Cabelli, V.J., A.P. DuFour, L.J. McCabe, and M.A. Levin. 1982. Swimming-associ-
ated gastroenteritis and water quality. Am. J. Epidemiol. 115:606-61.
Cabelli, V.J., A.P. DuFour, L.J. McCabe, and M.A. Levin. 1983. A marine recre-
ational water quality criterion consistent with indicator concepts and risk analysis. J.
Water Pollut. Control Fed. 55:1306-1314.
CDC. 1979. Viral hepatitis outbreaks - Georgia, Alabama. Centers for Diseas
Control. Morbid. Mortal. Weekly Rep. 28:581.
Chang, S.L., G. Berg, K.A. Busch, R.E. Stevenson, N.A. Clarke, and P.W. Kabler.
1958. Application of the most Probable Number method for estimating concentrations
of animal viruses by tissue culture technique. Virology 6:27.
Clark, H.F., E.E. Geldreich, H.L. Jeter, and P.W. Kabler. 1951. The membrane filter
in sanitary bacteriology. Pub. Health Rep. 66:951.
Clark, J.A. 1969. The detection of various bacteria indicative of water pollution by a
presence-absence (P-A) procedure. Can. J. Microbiol. 15:771.
Clark, J.A. 1980. The influence of increasing numbers of nonindicator organisms
upon the detection of indicator organisms by the membrane filter and presence-
absence tests. Can. J. Microbiol. 26:827.
Clark J.A. and L.T. Vlassoff. 1973. Relationships among pollution indicator bacteria
isolated from raw water and distribution systems by the presence-absence (P-A) test
Health Lab. Sci. 10:163.
Clesceri, L.S., A.E. Greenberg, and R.R. Trussell (eds). 1989. Standard Methods for
the Examination of Water and Wastewater. 17th Edition. American Public Health
Association, Washington, DC.
-------
B-221
VM^A <ฃ} * ^ 'ง ป. ,. % www* -.^.w- ซซ * X v% -n-X__
'ป *- *^ _?ซ^ <^ ^-.v? * ^
^^^^
Cliver, D.0.1967. Detection of enteric viruses by concentration with polyethylene
glycol. In: Transmission of Viruses bv the Water Rnnte. G.Berg, ed. Interscience
Publ., New York, N.Y.
Cooney, M.K. 1973. Relative efficiency of cell cultures for detection of viruses.
Health Lab. Sci. 4:295.
Cowles, P.B. 1939. A modified fermentation tube. J. Bacteriol. 38:677.
Dalla Valle, J.M. 1941. Notes on the most probable number index as used in bacteriol-
ogy. Pub. Health Rep. 56:229.
Dobberkau, H.J., R. Walter, and S. Rudiger. 1981. Methods for virus concentration
from water. In: Viruses and Wastewater Treatment. M. Goddard and M. Butler, eds.
Pergamon Press, New York, N.Y.
DuFour, A.P., E.R. Strickland, and V.J. Cabelli. 1981. Membrane filter method for
enumerating Escherichia coli. Appl. Environ. Microbiol. 41:1152-1158.
Dutka, B.D. 1981. Membrane Filtration Applications. Techniques and Problems.
Marcel Dekkar, Inc., New York, N.Y.
Emerson, DJ. and V.J. Cabelli. 1982. Extraction of Clostridiumperftngens spores
from bottom sediment samplers. Appl. Environ. Microbiol. 44:1152-1158.
Ericksen, T.H., C. Thomas, and A. Dufour. 1983. Comparison of two selective
membrane filter methods for enumerating fecal streptococci in freshwater samples.
Abs. Annual Meeting, American Soc. Microbiology, p. 279.
Evans, T.M., C.E. Warvick.RJ. Seidler, and M.W. LeChevallier. 1981. Failure of
the most-probably number techniques to detect coliforms in drinking water and raw
water supplies. Appl. Environ. Microbiol. 41:130.
Farrah, S.R. 1982. Isolation of viruses associated with sludge particles. In: Methods
in F.nvironmp.ntal Virology. C.P. Gerba and S.M. Goyal, eds. Marcel Dekker, Inc.,
New York, N.Y.
Feingold, A.O. 1973. Hepatitis from eating steamed clams. J. Am. Med. Assoc.
225:526-527.
-------
B-222
Ferraro, S.P., F.A. Cole, W.A. DeBen, and R.C. Swartz. 1989. Power-cost efficiency
of eight macrobenthic sampling schemes in Puget Sound, Washington, USA. Can. J.
Fish. Aquat. Sci. 46:2157-2165.
Fifield, C.W. and C.P. Schaufus. 1958. Improved membrane filter medium for the
detection of coliform-organisms. J. Amer. Water Works Assoc. 50:193.
Gameson, A.L.N. 1983. Investigation of sewage discharges to some British coastal
waters. Water Resources Centre Technical Report, TR 193. Bucks, United Kingdom.
40pp.
Geldreich, E.E., P.W. Kabler, H.L. Jeter, and H.F. Clark. 1955. A delayed incubation
membrane filter test for coliform bacteria in water. Amer. J. Pub. Health 45:1462.
Geldreich, E.E., H.L. Jeter, and J.A. Winter. 1967. Technical considerations in
applying the membrane filter procedure. Health Lab. Sci. 4:113.
Gerba, C.P. and S.M. Goyal. 1982. Methods in Environmental Virology. Marcel
Dekker, New York.
Gerhards, P., ed. 1981. Manual of Methods for General Bacteriology. American Soc.
Microbiology, Washington, D.C.
Greenberg, A.E. and D.A. Hunt, eds. 1985. Laboratory procedures for the Examina-
tion of Seawater and Shellfish. 5th ed. American Public Health Assoc., Washington,
D.C.
Halvorson, H.O. and N.R. Ziegler. 1933-35. Application of statistics to problems in
bacteriology. J. Bacteriol. 24:101; 26:4331, 559, 29:609.
Hardy, J.T. 1982. The sea surface microlayer: Biology, chemistry, and anthropogenic
enrichment. Prog. Oceanogr. 11:307-328.
Homma, A., M.D. Sobsey, C. Wallis, and J.L. Melnick. 1973. Virus concentration
from sewage. Water Res. 7:945.
Hsiung, G.D. 1973. Diagnostic Virology, revised e.A Yale Univ. Press, New Haven,
Conn.
-------
B-223
Inhom, S.L. ed. 1977. Quality Assurance Practices for Health Laboratories. American
Public Health Assoc., Washington, D.C.
Jacobs, N.J., W.L. Zeigler, EC. Reed, T.A. Stukel, and E.W. Rice. 1986. Comparison
of membrane filter, multiple-fermentation-tube, and presence-absence techniques for
detecting total coliforms in small community water systems. Appl. Environ.
Microbiol. 51:1007.
Kaplan, J.E., R.A. Goodman, L.B. Schonberger, E.G. Lippy, and G.W. Gary. 1982.
Gastroenteritis due to Norwalk virus: An outbreak association with municipal water
system. J. Infect. Dis. 146: 190-197.
Kabler, P. W. 1954. Water examinations by membrane filter and MPN procedures.
Amer. J. Pub. Health 44:379.
Kelly, S. and W.W. Sanderson. 1962. Comparison of various tissue cultures for the
isolation of enteroviruses. Amer. J. Pub. Health 52:455.
Lee, L.H., C.A. Phillips, M.A. South, J.L. Melnick, and M.D. Yow. 1965. Enteric
virus isolations in different cell cultures. Bull. World Health Org. 32:657.
Lennette, E.H., A. Balows, WJ. Hausler, Jr., and H.J. Shadomy, ed. 1985. Manual of
Clinical Microbiology. 4th ed. American Soc. for Microbiology, Washington, D.C.
Levin, M.A., J.R. Fischer, and V.J. Cabelli. 1975. Membrane filter technique for
enumeration of enterococci in marine waters. Appl. Microbiol 30:66.
Lin, S. 1973. Evaluation of coliform test for chlorinated secondary effluents. J.
Water Pollut. Control Fed. 45:498.
Lin, S.D. 1976. Evaluation of Millipore HA and HC membrane filters for the enu-
meration of indicator bacteria. Appl. Environ. Microbiol. 32:300.
Lund, E. and C.E. Hedstrom. 1969. A study on sampling and isolation methods for
the detection of virus in sewage. Water Res. 3:823.
Lydholm, B. and A.L.Nielsen. 1979. Methods for detection of virus in wastewater
applied to samples from small scale treatment systems. Water Res. 14:169.
-------
B-224 i
McCarthy, J.A., H.A. Thomas, and J.E. Delaney. 1958. Evaluation of reliability of
coliform density test. Amer. J. Pub. Health 48:12.
McCarthy, J.A. and J.E. Delaney. 1958. Membrane filter media studies. Water
Sewage Works 105:292.
McCarthy, J.A. J.E. Delaney, and R.J. Grasso. 1961. Measuring coliforms in water.
Water Sewage Works 108:238.
McCarthy, J.A. and J.E. Delaney. 1965. Methods for measuring the coliform content
of water. Sec. Ill Delayed holding procedures for coliform bacteria. PHS Res. Grant
WP 00202 N1H Rep.
McCrady, M.H. 1915. The numerical interpretation of fermentation tube results. J.
Infect. Dis. 12:183.
Morris, R. and W.M. Waite. 1980. Evaluation of procedures for recovery of viruses
from water - II detection systems. Water Res. 14:795.
NOAA. 1988. National Marine Pollution Program Federal Plan for Ocean Pollution
Research. Development and Monitoring. U.S. Department of Commerce, National
Oceanic and Atmospheric Administration, Washington, D.C.
Ohara, H., H. Naruto, W. Watanabe, and I. Ebisawa. 1983. An outbreak of Hepatitis A
caused by consumption of raw oysters. J. Hyg. Camb. 91:163-165.
Olson, B.H. 1978. Enhanced accuracy of coliform testing in seawater by a modifica-
tion of the most-probable-number method. Appl. Microbiol. 36:438.
OTA. 1987. Wastes in Marine Environments. Office of Technical Assessment. US
Congress. OTA 0-334. 313pp.
Panezai, A.K., T.J. Macklin, and H.G. Coles. 1965. Coli-aerogenes and Escherichia
coli counts on water samples by means of transported membranes. Proc. Soc. Water
Treat. Exam. 14:179.
Payment, P., C.P. Gerba, C. Wallis, and J.L. Melnick. 1976. Methods for concentrat-
ing viruses from large volumes of estuarine water on plated membranes. Water Res.
10:893.
-------
B-225
Pederson, D.C. 1980. Density levels of pathogenic organisms in municipal wastewa-
ter sludges: A literature review. Prepared by Camp Dresser and McKee, Inc. for
ORD, US EPA, Cincinnati, OH.
Ramia, S. and S.A. Sattar. 1979. Second-step concentration of viruses in drinking and
surface waters using polyethylene glycol hydroextraction. Can. J. Microbiol 25:587.
Rao, V.C., U. Chandorkar, N.U. Rao, P. Kumaran, and S.B. Lakhe. 1972. A simple
method for concentrating and detecting viruses in wastewater. Water Res. 6:1565.
Rehm, R., S. Duletsky, J. Pierce, and R. Sommer. 1983. Contaminants of concern in
sewage sludge. Draft prepared for US EPA, Office of Program Policy and Evaluation,
Washington, D.C.
Rovozzo, G.C. and C.N. Burke. 1973. A Manual of Basic Virological Techniques.
Prentice-Hall, Englewood cliffs, NJ.
Schmidt, N.J. H.H. Ho, J.L. Riggs, and E.H. Lennette. 1978. Comparative sensitivity
of various cell culture systems for isolation of viruses from wastewater and fecal
samples. Appl. Environ. Microbiol. 36:480.
Seidler, R.J. T.M. Evans, J.R. Kaufman, C.E. Warvick, and M.W. LeChevallier. 1981.
Limitations of standard coliform enumeration techniques. J. Amer. Water Works
Assoc. 73:538.
Selna, M.W. and R.P. Miele. 1977. Virus sampling in wastewater-field experiences.
J. Environ. Eng. Div., Proc. Amer. Soc. Civil Eng. 103:693.
Shuval, H.I., S. Cymbalista, B.Fattal, and N. Goldblum. 1967. Concentration of
enteric viruses in water by hydro-extraction and two-phase separation. In G. Berg, ed.
Transmission of Viruses by the Water route. Interscience Publ., New York, NY
Shuval, H., B. Fattal, S. Cymbalista, and N. Goldblum. 1969. The phase-separation
method for the concentration and detection of viruses in water. Water Res. 3:225.
Slanetz, L.W. and C.H. Bartley. 1957. Numbers of enterococci water, sewage and
feces determined by the membrane filter technique with an improved medium. J.
Bacteriol. 74:591.
-------
B-226!
Sobsey, M.D. 1976. Methods for detecting enteric viruses in water and wastewater.
In: Viruses in Water. G. Berg, H.L. Bodily, E.H. Lennette, J.L. Melnick and T.G.
Metcalf, eds. American Public Health Assoc., Washington, D.C.
Sobsey, M.D. 1976. Field monitoring techniques and data analysis. In: Virus Aspects
of Applying Municipal Waste to Land. L.B. Baldwin, J.M. Davidson and J.F. Gerber,
eds. Univ. Florida, Gainesville.
Sobsey, M.D. 1982. Quality of currently available methodology for monitoring
viruses in the environment. Environ. Internal. 7:39.
Sobsey, M.D., C.P. Gerba, C. Wallis, and J.L. Melnick. 1977. Concentration of
enteroviruses from large volumes of turbid estuary water. Can. J. Microbiol. 23:770.
St. John, E.W., J.R. Matches, and M.M. Wekell. 1982. Use of iron milk medium for
enumeration of Clostridiwn perfringens. J. Assoc. Off. Anal. Chem. 65:1129-1133.
Strandridge, J.H. and J.J. Delfino. 1981. A-1 Medium: Alternative techniques for
fecal coliform organism enumeration in chlorinated wastewaters. Appl. Environ.
Microbiol. 42:918.
Thomas, H.A., Jr. 1942. Bacterial densities from fermentation tube test. J. Amer.
Water Works Assoc. 34:572.
Taylor, R.H., R.H. Bordner, and P.V. Scarpino. 1973. Delayed incubation membrane-
filter test for fecal coliforms. Appl. Microbiol. 25:363.
Thomas, H.A. and R.L. Woodward. 1956. Use of molecular filter membranes for
water potability control J. Amer. Waterworks Assoc. 48:1391.
U.S. EPA. 1978. Microbiological methods for monitoring the environment. EPA
600/8-78-017. Environmental Monitoring and Support Laboratory. Office of Research
and Development. 337 pp.
U.S. EPA. 1985. Test Methods for Escherichia coli and Enterococci in Water by the
Membrane Filter Procedure, U.S. Environmental Protection Agency, Environmental
Monitoring and Support Laboratory, Cincinnati, OH.
U.S. EPA. 1986. Ambient Water Quality Criteria for Bacteria -1986. EPA 440/5-84-
002, U.S. Environmental Protection Agency, Washington, D.C.
-------
B-227
U.S. EPA. 1987. Puget Sound protocols. Final Report. 31pp. Prepared for
Region X, Office of Puget Sound.
U.S. EPA. 1988. Water Quality Standards Criteria Summaries: A Compilation of
State/Federal Criteria - Bacteria. U.S. Environmental Protection Agency, Office of
Water Regulations and Standards, Washington, D.C.
Wallis, C. and J.L. Melnick. 1967. Virus concentration on aluminum and calcium
salts. Amer. J. Epidemiol. 85:459.
Wallis, C. and J.L. Melnick. 1967. Concentration of viruses on aluminum hydroxide
precipitates. In: Transmission of Viruses bv the Water Route. G. Berg, ed.
Interscience Publ., New york, N.Y.
Wellings, P.M. A.L. Lewis, and C.W. Mountain. 1976. Viral concentration tech-
niques for field sample analysis. In: Virus Aspects of Applying Municipal Waste to
Land. L.B. Baldwin, J.M. Davidson and J.F. Gerber, eds. Univ. Florida, Gainesville.
Weiss, J.E. and C.A. Hunter. 1939. Simplified bacteriological examination of water.
J. Amer. Water Works Assoc. 31:689.
-------
B-228
------- |