Review of EPA's Environmental
Monitoring and Assessment Program
Forests and Estuaries
NATIONAL RESEARCH COUNCIL
-------
Review of EPA's Environmental Monitoring
and Assessment Program:
Forest and Estuaries Components
Committee to Review the ER% Environmental
Monitoring and Assessment Program
Board on Environmental Studies and Toxicology
Water Science and Technology Board
Commission on Life Sciences
Commission on Geosciences, Environment,
and Resources
National Research Council
Washington, D.C. 1994
-------
NOTICE: The project that is the subject of this report was approved by the Governing Board
of the National Research Council, whose members are drawn from the councils of the
National Academy of Sciences, the National Academy of Engineering, and the Institute of
Medicine. The members of the board responsible for the report were chosen for their special
competences and with regard for appropriate balance.
This report has been reviewed by a group other than the authors according to
procedures approved by a Report Review Committee consisting of members of the National
Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine.
The National Academy of Sciences is a private, nonprofit, self-perpetuating society of
distinguished scholars engaged in scientific and engineering research, dedicated to the
furtherance of science and technology and to their use for the general welfare. Upon the
authority of the charter granted to it by the Congress in 1863, the Academy has a mandate
that requires it to advise the federal government on scientific and technical matters. Dr.
Bruce Alberts is president of the National Academy of Sciences.
The National Academy of Engineering was established in 1964, under the charter of the
National Academy of Sciences, as a parallel organization of outstanding engineers. It is
autonomous in its administration and in the selection of its members, sharing with the
National Academy of Sciences the responsibility for advising the federal government. The
National Academy of Engineering also sponsors engineering programs aimed at meeting
national needs, encourages education and research, and recognizes the superior achieve-
ments of engineers. Dr. Robert M. White is president of the National Academy of Engi-
neering.
The Institute of Medicine was established in 1970 by the National Academy of Sciences
to secure the services of eminent members of appropriate professions in the examination of
policy matters pertaining to the health of the public. The Institute acts under the
responsibility given to the National Academy of Sciences by its congressional charter to be
an adviser to the federal government and, upon its own initiative, to identify issues of medical
care, research, and education. Dr. Kenneth I. Shine is president of the Institute of Medicine.
The National Research Council was organized by the National Academy of Sciences in
1916 to associate the broad community of science and technology with the Academys
purposes of furthering knowledge and advising the federal government. Functioning in
accordance with general policies determined by the Academy, the Council has become the
principal operating agency of both the National Academy of Sciences and the National
Academy of Engineering in providing services to the government, the public, and the
scientific and engineering communities. The Council is administered jointly by both
Academies and the Institute of Medicine. Dr. Bruce Albert and Dr. Robert M. White are
chairman and vice chairman, respectively, of the National Research Council.
Support for this project was provided by the U.S. Environmental Protection Agency under
Agreement No. 68C00082/C.
Copyright 1994 by the National Academy of Sciences. All rights reserved.
Cover by John Eberhard, Pittsburgh, PA.
Available from Board on Environmental Studies and Toxicology or Water Science and Tech-
nology Board, 2101 Constitution Avenue, N.W., Washington, D.C. 20418.
Printed in the United States of America
-------
This report is dedicated to the memory of committee members
Shirley Dreiss (1949-1993)
John Sutherland (1942-1993)
-------
COMMITTEE TO REVIEW THE ERA'S
ENVIRONMENTAL MONITORING AND ASSESSMENT PROGRAM
RICHARD FISHER, Chair, Texas A&M University College Station
PATRICK L BREZONIK, University of Minnesota, St. Paul
INGRID C. BURKE, Colorado State University, Ft. Collins
EDWIN H. CLARK, II, State of Delaware, Dover (February 1991 to October
1993)
LOVEDAY L CONQUEST, University of Washington, Seattle
ARTHUR COOPER, North Carolina State University (February 1991 to
October 1993)
SHIRLEY DREISS, University of California, Santa Cruz (February 1991 to
December 1993)
THURMAN L GROVE, North Carolina State University, Raleigh
JOHN E. HOBBIE, Marine Biological Laboratory Ecosystems Center,
Woods Hole, Massachusetts
CHARLES C. JOHNSON, Jr., U.S. Public Health Service, Washington, D.C.
(Retired)
TIM K. KRATZ, University of Wisconsin, Madison
ANNE E. MCELROY State University of New York, Stony Brook
JOHN PASTOR, University of Minnesota, Duluth
JAMES N. PITTS, JR., California State University, Fullerton
RAYMOND A. PRICE, Queen's University, Kingston, Ontario
SAUL SAILA, University of Rhode Island, Kingston
TERENCE R. SMITH, University of California, Santa Barbara
SUSAN STAFFORD, Oregon State University, Corvallis
DONALD R. STRONG, University of California, Bodega Bay
MICHAEL J. WILEY University of Michigan, Ann Arbor
Liaison to the Board on Environmental Studies and Toxicology
KRISTIN SHRADER-FRECHETTE, University of South Florida, Tampa
National Research Council Staff
SHEILA D. DAVID, Study Director, Water Science and Technology Board
DAVID J. POLICANSKY Study Director, Board on Environmental Studies
and Toxicology
ANITA A. HALL, Senior Project Assistant, Water Science and Technology
Board
SHIRLEY F. JONES, Project Assistant, Board on Environmental Studies and
Toxicology
IV
-------
BOARD ON ENVIRONMENTAL STUDIES AND TOXICOLOGY
PAUL G. RISSER, Chair, Miami University, Oxford, Ohio
FREDERICK R. ANDERSON, Cadwalader, Wickersham & Taft, Washington,
D.C.
MCHAEL J. BEAN, Environmental Defense Fund, Washington, D.C.
EULA BINGHAM, University of Cincinnati, Cincinnati, Ohio
EDWIN H. CLARK, Clean Sites, Inc., Alexandria, Virginia
ALLAN H. CONNEY Rutgers University, New Jersey
JOHN L EMMERSON, Eli Lilly & Company, Greenfield, Indiana
ROBERT C. FORNEY Consultant, Unionville, Pennsylvania
ROBERT A. FROSCH, Harvard University, Cambridge, Massachusetts
KAI LEE, Williams College, Williamstown, Massachusetts
JANE LUBCHENCO, Oregon State University, Corvallis
GORDON OR1ANS, University of Washington, Seattle, Washington
FRANK L PARKER, Vanderbilt University, Nashville, Tennessee
GEOFFREY PLACE, Consultant, Hilton Head, South Carolina
DAVID P. RALL, Consultant, Washington, D.C.
LESLIE A. REAL, Indiana University, Bloomington, Indiana
KRISTIN SHRADER-FRECHETTE, University of South Florida, Tampa,
Florida
GERALD van BELLE, University of Washington, Seattle, Washington
BAILUS WALKER, JR., University of Oklahoma, Oklahoma City
GERHARDT ZBINDEN, Eidgenossische Technisiche Hochschule Zurich,
Schwerzenbach, Switzerland
Staff
JAMES J. REISA, Director
DAVID J. POLICANSKY Associate Director and Program Director for
Applied Ecology and Natural Resources
RICHARD D. THOMAS, Associate Director and Program Director for Human
Toxicology and Risk Assessment
LEE R. PAULSON, Program Director for Information Systems and Statistics
RAYMOND A. WASSEL, Program Director for Environmental Sciences and
Engineering
SHARON HOLZMANN, Administrative Associate
BERNIDEAN WILLIAMS, Administrative Associate
CAROL M. KRESGE, Senior Program Assistant
-------
WATER SCIENCE AND TECHNOLOGY BOARD
DANIEL A. OKUN, Chair, University of North Carolina, Chapel Hill
A. DAN TARLOCK, Vice Chair, Illinois Institute of Technology, Chicago-Kent
College of Law, Chicago
J. DAN ALLEN, Chevron U.S.A., Inc., New Orleans, Louisiana
PATRICK L BREZONIK, University of Minnesota, St. Paul
KENNETH D. FREDERICK, Resources for the Future, Washington, D.C.
DAVID L. FREYBERG, Stanford University, Stanford, California
WILFORD R. GARDNER, University of California, Berkeley
WILLIAM L. GRAF, Arizona State University, Tempe
THOMAS M. HELLMAN, Bristol-Myers Squibb Company, New York,
New York
ROBERT J. HUGGETT, College of William and Mary Gloucester Point,
Virginia
CHARLES C. JOHNSON, Jr., U.S. Public Health Service, Washington, D.C.
(Retired)
WILLIAM M. LEWIS, JR., University of Colorado, Boulder
CAROLYN H. OLSEN, Brown and Caldwell, Atlanta, Georgia
CHARLES R. O'MELIA, Johns Hopkins University, Baltimore, Maryland
STAVROS S. PAPADOPULOS, S. S. Papadopulos & Associates, Inc.,
Rockville, Maryland
BRUCE E. RITTMANN, Northwestern University, Evanston, Illinois
JOY B. ZEDLER, San Diego State University, San Diego, California
Staff
STEPHEN D. PARKER, Director
SARAH CONNICK, Senior Staff Officer
SHEILA D. DAVID, Senior Staff Officer
CHRIS ELFRING, Senior Staff Officer
GARY KRAUSS, Staff Officer
JACQUELINE MACDONALD, Staff Officer
JEANNE AQUILINO, Administrative Specialist
ANITA A. HALL, Administrative Assistant
GREGORY NYCE, Senior Project Assistant
MARY BETH MORRIS, Senior Project Assistant
ANGELA BRUBAKER, Project Assistant
VI
-------
COMMISSION ON GEOSCIENCES, ENVIRONMENT,
AND RESOURCES
M. GORDON WOLMAN, Chair, The Johns Hopkins University, Baltimore,
Maryland
PATRICK R. ATKINS, Aluminum Company of America, Pittsburgh,
Pennsylvania
PETER EAGLESON, Massachusetts Institute of Technology, Cambridge
EDWARD A. FRIEMAN, Scripps Institution of Oceanography, La Jolla,
California
W BARCLAY KAMB, California Institute of Technology, Pasadena
JACK E. OLIVER, Cornell University, Ithaca, New York
FRANK L PARKER, Vanderbilt University, Nashville, Tennessee
RAYMOND A. PRICE, Queen's University at Kingston, Ontario *
THOMAS A. SCHELLING, University of Maryland, College Park, Maryland
LARRY L SMARR, University of Illinois, Urbana-Champaign
STEVEN M. STANLEY The Johns Hopkins University, Baltimore, Maryland
VICTORIA J. TSCHENKEL, Landers and Parsons, Tallahassee, Florida
WARREN WASHINGTON, National Center for Atmospheric Research,
Boulder, Colorado
EDITH BROWN WEISS, Georgetown University Law Center, Washington,
D.C.
Staff
STEPHEN RATTEN, Executive Director
STEPHEN D. PARKER, Associate Executive Director
MORGAN GOPNIK, Assistant Executive Director
JEANETTE SPOON, Administrative Officer
SANDI FITZPATRICK, Administrative Associate
ROBIN ALLEN, Senior Project Assistant
VII
-------
COMMISSION ON LIFE SCIENCES
THOMAS D. POLLARD, Chair, Johns Hopkins Medical School, Baltimore,
Maryland
BRUCE N. AMES, University of California, Berkeley
JOHN C. BAILAR, III, McGill University, Montreal, Canada
J. MICHAEL BISHOP, University of California Medical Center, San Francisco
JOHN E. BURRIS, Marine Biological Laboratory, Woods Hole,
Massachusetts
MICHAEL! CLEGG, University of California, Riverside
GLENN A. CROSBY Washington State University, Pullman
LEROY E. HOOD, University of Washington, Seattle
MARIAN E. KOSHLAND, University of California, Berkeley
RICHARD E. LENSKI, Michigan State University, East Lansing
EMIL A. PFITZER, Hoffmann-LaRoche, Inc., Nutley, New Jersey
MALCOLM C. PIKE, University of Southern California School of Medicine,
Los Angeles
HENRY C. PITOT, III, University of Wisconsin, Madison
PAUL G. R1SSER, Miami University, Oxford, Ohio
JONATHAN M. SAMET, University of New Mexico School of Medicine,
Albuquerque
HAROLD M. SCHMECK, JR., Armonk, New York
CARLA J. SHATZ, University of California, Berkeley
SUSAN S. TAYLOR, University of California, San Diego, LaJolla
P. ROY \AGELOS, Merck & Company, Inc., Whitehouse Station, New Jersey
JOHN L VANDEBERG, Southwest Foundation for Biomedical Research,
San Antonio, Texas
TORSTEN N. WIESEL, Rockefeller University New York
Staff
PAUL GILMAN, Executive Director
AlVIN LAZEN, Director of Program Operations
BARBARA WENSUS, Administrative Officer
JACQUELINE PRINCE, Administrative Associate
SOLVEIG PADILLA, Administrative Assistant
NORMAN GROSSBLATT, Senior Editor
viii
-------
Contents
EXECUTIVE SUMMARY !
1 OVERVIEW 9
Introduction, 9
Goals and Objectives, 11
Associations and Causality, 12
Detection of Trends, 13
Indicators, 14
Information Management, 15
Cost, 17
Management and Coordination, 19
Intragency and Intra-Agency Cooperation, 20
2 ESTUARIES 21
Background, 21
Scope of Report, 22
Sampling Design, 24
Indicators, 27
Evolution of EMAP-Estuaries Indicators, 32
Data Collection, Analysis, and Management, 33
Application of Data, 36
Scientific Review, 37
Coordination with Agencies, 38
Recommendations, 39
3 FOREST HEALTH MONITORING PROGRAM 42
Background, 42
Theoretical Basis, 43
Selection of Indicators, 45
ix
-------
x Contents
Sampling Design, 47
Data Management, 49
Landscape Characterization, 50
Staffing and Personnel Training, 51
Results of Demonstration Projects, 52
Budget and Costs, 54
Recommendations, 54
REFERENCES 56
APPENDIX A: Review of EPA's Environmental
Monitoring and Assessment Program—Interim Report 60
APPENDIX B: EMAP Documents Reviewed by NRC Committee 82
APPENDIX C: Biographical Information 93
-------
Executive Summary
The Environmental Monitoring and Assessment Program (EMAP) of the
U.S. Environmental Protection Agency (EPA) is an ambitious, nationwide
effort to monitor the status and changes in the condition of the nation's
ecological resources to provide information for regulators, managers, and
policy makers. The United States is a large and diverse country, and
developing a nationwide program that retains sufficient spatial and temporal
coverage to provide meaningful information is thus a daunting challenge. To
assist it in meeting this challenge, EPA requested that the National Research
Council (NRC) provide an ongoing review of its developing program. In
response to that request, the NRC constituted the Committee to Review
EPAs Environmental Monitoring and Assessment Program in 1991. This is
the committees second report to EPA, and it focuses mainly on a review of
EMAP's forest health monitoring component and the estuaries component.
For this report, the committee has reviewed documents received from EPA
through June 1993.
In addition to having large goals, EMAP is an administratively complex
program. Its structure has evolved during the committees review, and now
consists of a headquarters in Washington, D.C., a central planning and
administration center in North Carolina (EMAP Center), and activities con-
ducted in many EPA laboratories and through private and university contrac-
tors throughout the United States. Conceptually, the foundations of EMAP
consist of a strategy for choosing indicators of the condition of various
different kinds of ecological resources; a landscape characterization to
provide information on the distribution and abundance of different kinds of
resource types, which utilizes a probability-based spatial grid; a sampling
design; and eight resource types defined by the program to represent eco-
system types or resources of national interest. The eight resource types are
agroecosystems, arid lands, forests, the Great Lakes, estuaries, inland sur-
-------
2 Review of EPA's EMAP
face waters, wetlands, and landscape ecology. In total, they include all the
land area of the United States and adjacent nearshore oceans.
The documentation of EMAP has produced a large and sometimes
unclear body of literature and—typical of many large programs—it has
lagged behind its initial schedule. Some of this delay is caused by internal
EPA review procedures and printing of reports, which can significantly delay
availability for publication in peer-reviewed journals and has also caused
delays in review of documents by the NRC committee. This committee's
schedule of reports and their subject matter have largely been determined
by the availability of documents to review. This report focuses on the
estuaries and forests components, which are the first resource types to have
had pilot studies. The committee concluded that its review of the two
resources must be done within the context of the overall program to be
meaningful (after all, the program is intended to be a unified, cohesive one),
and within the context of the program's likely ability to achieve its goals,
much as the committee provided in its first report. The context is provided
in Chapter 1 (Overview).
OVERALL ASSESSMENT
The committee continues to believe that EMAPs overall purpose and
basic goals are laudable, but questions the degree to which EPA can achieve
them in a timely and cost-effective manner. EMAP has clearly evolved and
progressed since the publication of the committee's interim report, but many
of the questions raised in that report have not yet been answered. Answers
to some of these questions will take additional time and money, others might
not be answerable, and some (e.g., how EMAP plans to manage the data
generated by the program) simply havent been addressed. The committee
has been impressed with the quality of the thought and information
generated in some of the resource groups and with EMAP officials' efforts to
obtain cooperation from other federal and state agencies. These matters are
discussed in Chapter 1. As in its earlier report, the committee is concerned
about the way that EMAP, as an environmental monitoring program,
addresses cause and effect relationships; about the ability of the program to
detect temporal trends; about the management and coordination of the
program; about the lack of adequate planning for information management;
and about the cost. A coherent, consistent, comprehensive strategic plan for
EMAP, with logical charts showing the flow of thought and activity through
the goals and objectives, should be an objective of the highest echelons of
EPA that are involved with EMAP.
-------
Executive Summary
In addition, although the indicator strategy and landscape charac-
terization components of EMAP are essential to it, according to EMAP's
designers, they are still only in early stages of development. Following is a
summary of recommendations of the committee's review of the activities and
plans of two EMAP resource groups—the estuaries component and the
forest health monitoring component.
Estuaries
The goals of the 1990 Virginian Province Demonstration Project were to
identify which indicators and design attributes are most effective for
assessing the ecological condition of estuarine resources on a regional scale
with limited financial resources. The committee believes that significant pro-
gress was made in many areas.
The grid-sampling scheme was successfully modified to better represent
discrete systems such as small estuaries and large rivers without com-
promising the acquisition of unbiased samples. A complex field-sampling
program was successfully mounted with a well-coordinated plan for quality
assurance of data acquisition, analysis, and management. Initial steps were
taken toward the development of a group of indicators of ecological condi-
tion. Subsequent efforts have been made to involve regional managers by
having them cooperate in future sampling and in the evaluation of the
applicability of data collected. The activities and results of the first year of
sampling (1990) have been issued in a well-written synthesis report de-
scribing the process of indicator development and containing an initial
interpretation of the data obtained (Weisberg, et al. 1992). Based on the
material in this report, the committee believes that EMAP has made a good
first step in getting the estuaries section of EMAP started.
Although there have been many positive accomplishments, the com-
mittee nonetheless believes that there are a number of areas needing signifi-
cant work. A more explicit conceptual model must be developed to drive
indicator development and set priorities. Continued work also is needed to
develop meaningful indicators that assess basic ecological condition (status
and functioning).
The combination of the EMAP probability-based sampling design and the
realities of national coverage with a limited budget severely limit the type and
number of indicator measurements that can be made. The committee
agrees with the review panel of the Estuarine Research Federation, which
doubts that the indices generated by EMAP will have the power to detect the
amount of environmental change expected. Environmental change can oc-
-------
4 Review of EPA's EMAP
cur at various rates. For example, the EMAP design standard is the ability to
detect a 20 percent change occurring over a decade. The published
information on changes in various indicators shows, however, that some
changes occur in estuaries at a much slower rate than this (Stanley, 1993).
As a result, it may take several decades for a 20 percent change to occur.
Therefore, it would take decades to be able to detect changes with the
current EMAP sampling design. It seems that in some cases EPA personnel
have not researched the published literature. By contrast, as is pointed out
in Chapters, changes in ecosystems can be quite sudden and catastrophic,
perhaps too fast to be adequately captured by EMAP's sampling scheme.
The committee believes that it is time for this issue to be clearly analyzed by
EMAP using extant data sets or similar proxy data. One alternative for
serious consideration is the adoption of intensive sampling at a number of
supplemental sites in conjunction with decreased sampling on the grid and
increased use of remote sensing.
The committee believes that large programs such as the estuaries
component of EMAP usually pay insufficient attention to analyzing exactly
what they have learned in their pilot and demonstration projects. There is a
temptation to think that the next challenge is to carry out pilot projects on
new provinces, one after another. However, as pointed out by the Estuarine
Research Federation review committee, the real challenge is in obtaining the
best possible set of indicators of ecological condition. Therefore, this
committee recommends that EMAP personnel stop and evaluate the
estuaries part of EMAP in detail before going on or adding additional
provinces. This evaluation should occur as soon as possible after the
Virginian Province demonstration completes its first four-year cycle and
should include a comparison of the EMAP information with other published
information on indicators of condition of estuarine resources with different
design attributes. A summary of other recommendations concerning the
estuaries component follow. Further detailed discussion of the recommen-
dations can be found in Chapter 2.
Indicators
1. The committee strongly recommends that the estuaries component
of EMAP include indicators of ecosystem function. These indicators are
difficult to monitor when studies are made only once a year, but can be
estimated to some extent indirectly. Lack of such indicators should be
addressed as soon as possible. An example of such an indirect approach
is that algal biomass can be used as a surrogate for primary production.
-------
Executive Summary 5
Also, remote sensing provides one possibility for chlorophyll measurements
on a regional scale.
2. Another measure of important coastal habitat, submerged aquatic
vegetation, was missing in the Virginian Province Demonstration Project.
Inclusion of submerged aquatic vegetation in the sampling scheme for all
estuaries demonstration projects should be considered.
3. The committee is concerned that insufficient effort has been
devoted to fish sampling to make the data obtained useful. The committee
believes that a relatively short trawl done once does not collect enough fish
for meaningful determination of population characteristics, contaminant body
burdens, or incidence of gross pathology. If quantitative information on
these indices is desirable, arrangements should be made with other agencies
with the experience and personnel on hand for more comprehen- sive
collection and analysis of data. If the level of planning and effort allocated
by EMAP for these activities cannot be significantly increased, the committee
recommends eliminating the fish sampling program.
4. In support of new indicators development, areas of research that
should be looked at include the analysis of long-term data sets from various
sites to examine indicator variability and its causes and the use of molecular
probes to look for the presence of enzymes indicating pollutant exposure or
changes in ecosystem function (e.g., nitrogen fixation).
5. The committee has found that no use has yet been made of a
number of historical data sets in the Virginian Province that have data com-
parable to those being collected by EMAP. Out of 18 studies investigated,
eight were found to contain information important to EMAP and in particular
to EMAP-estuaries. The committee recommends detailed analysis of this
material, which would help provide valuable information on spatial and
temporal variability and the power of certain indicators to detect trends within
a given period. It is past time for this work to have been completed.
Advice, Consultation, and Scientific Review
\
The committee recognizes EPA has sought advice from a wide variety of
scientists in developing EMAP, but is concerned about the effectiveness of
the present mechanisms for incorporating scientific expertise into the design
and execution of resource-group activities. Working groups, which have
been used by most of the resource groups, provide peer review but are not
necessarily efficient or adequate for ensuring that activities are based on the
best scientific approach. The committee recommends that continuing
oversight and review by groups of scientists from outside of EPA, built into
-------
Review of EPA's EMAP
the program at the highest levels, be implemented for EMAP center plan-
ning, for indicator development strategy, for landscape characterization, and
for all resource groups.
Forest Health Monitoring
EMAP's Forest Health Monitoring program (EMAP-Forests) proposes to
collect data on environmental factors that influence forest growth, as well as
additional response variables of the trees such as soil nutrients and canopy
structure. If this is implemented, the resulting data sets will be valuable for
ecologists and foresters seeking to understand basic ecological patterns and
for policy makers who require information for the evaluation of future
environmental impacts on the nation's forests.
The multi-agency partnership of EPA, the Forest Service, state forestry
agencies, the National Park Service, the Fish and Wildlife Service, the
Tennessee Valley Authority, and the Bureau of Land Management is exactly
the type of cooperation that EPA and this committee see as vital to EMAP's
national monitoring effort. Many of the positive features of the program
derive from the previously established U.S. Forest Service Health Monitoring
Program. Following is a summary of specific recommendations based on
the committees review. Further detailed discussion on the recommenda-
tions can be found in Chapter 3.
LacA of a Theoretical Basis
Elements of a theoretical basis for EMAP-Forests included hierarchy
theory, sampling theory, epidemiological theory, and the stand-development
theory of Oliver and Larson (1990). However, the logical basis by which
these theories explain the responses of forests to stress and the subsequent
responses of surface waters to changes in forests and by which the theories
indicator development and sampling protocols is not clear. Heavy reliance
appears to be placed on a purely epidemiological model. Epidemiological
models describe how diseases spread through populations. Such models
appear to have little utility in predicting how nutrient cycles, nutrient losses,
or biodiversity of ecosystems change in response to stress. Therefore, the
committee strongly recommends that EMAP personnel continue develop-
ment of a theoretical basis for EMAP-Forests from which predictions can be
made of general types of forest response to different types of stress. The
theory should encompass at least productivity and diversity.
-------
Executive Summary 7
Select a Set of Indicators
It is essential that EMAP-Forests choose a set of indicators as soon as
possible and then conduct the staff work necessary to establish sampling
methods and convey these to field crews. The next step is to develop the
process for interpreting the results derived from field studies. Priority should
be given to the evaluation of measurements that integrate limiting factors
over the growing season, such as resin bags, and that can be performed
quickly using standardized procedures.
Revise Sampling Design
The committee recommends that the current design of four-year plot
rotations be replaced or augmented by a design in which some plots are
revisited every year. Revisiting a site only once every four years prevents
EMAP from making site-specific estimates of changes in these frequencies.
This is unfortunate in the case of known cyclical events. An augmented
sampling scheme that would permit some plots to be revisited every year
would maximize temporal coverage. Some plots could be sampled on a
rotating basis to maximize spatial coverage. This is essentially the recom-
mendation also reached by the statistical sampling design team.
Information-Management System
EMAP-Forests should develop a comprehensive information-manage-
ment plan that outlines user requirements examines long-term implementa-
tion (of hardware and software), and fits in with the overall plan for the
information-management system.
Publish Study Results
The committee recommends that EPA encourage publication of study
results in peer-reviewed science journals to gain credibility in the scientific
community and to ensure accessibility of information.
-------
8 Review of EPA's EMAP
Delay Full Implementation Until Results of
Demonstration Projects are Evaluated
For all the reasons described above, the committee recommends that the
EMAP-Forests program not be fully implemented until the results of
demonstration projects have been thoroughly evaluated and a realistic esti-
mate of its costs to EPA and other agencies is available.
-------
1
Overview
INTRODUCTION
In response to a request from the U.S. Environmental Protection Agency
(EPA), the National Research Council (NRC) Committee to Review EP/%
Environmental Monitoring and Assessment Program (EMAP) has been
reviewing EMAP's overall design and objectives and considering ways to
increase its effectiveness. EMAP was clearly intended to provide information
required by Congress and the Administrator of EPA to assist in making
public policy decisions. The committee has reviewed many EMAP public
documents (see Appendix B) and has received briefings and held
conversations with EMAP officers, including the technical directors of the
EMAP resource components. In June 1992, the committee prepared an
interim report (NRC, 1992; Appendix A), which raised substantive questions
about the design and implementation of EMAP.
As described by EPA in many documents, EMAP is unified by its
approach to landscape characterization, the development of a strategy for
choosing indicators of ecosystem condition, and the use of a probability-
based sampling approach that uses a hexagonal grid for identifying sam-
pling sites. The eight resource types identified by the program (agroeco-
systems, arid lands, forests, the Great Lakes, estuaries, inland surface
waters, wetlands, and landscape ecology) are intended to represent eco-
system types or resources of national interest and to provide a basis for
incorporating ecological knowledge into the design of indicators and
sampling programs.
This report provides a review of the activities and plans of two EMAP
resource groups: Estuaries and Forests. These resource groups are two of
eight for which EMAP plans or is currently running demonstration projects
to test the program. Because these component groups are part of a greater
whole, they cannot be evaluated in a vacuum. Thus, the committee's review
-------
10 Review of EPA's EMAP
of these individual components of EMAP considered the relationships
between those components and current perceptions of the basic goal and
the strategic objectives of EMAP. The committee evaluated the two groups
in the context of the whole program.
Much of the work of the estuaries and forest-monitoring resource groups
has been well conceived and executed; many of the results of the demon-
stration projects seem to be of considerable interest. However, the
importance and uniqueness of EMAP as designed and presented to Con-
gress, the scientific community, and the public depends on its being an
integrated, coordinated, national program. If it is not, its value is diminished.
The possibility of integrated descriptions of environmental trends across
several resource types is what sets EMAP apart from intensive surveys in
other agencies (e.g., the Forest Inventory and Analysis Program). For
example, trends in forest conditions should be considered in light of trends
in condition of adjacent streams and lakes. However, no pilot studies have
attempted any such integration, and there has been little thought given to
the scientific underpinnings of cross-resource analysis. Consequently, the
committee questions whether EMAP can accomplish its purpose, regardless
of how well the trend analyses of each resource group are performed.
In its interim report (NRC, 1992), this committee concluded that EMAP's
overall goal as stated by EPA was "laudable and endorsed by the com-
mittee." The committee, however, expressed considerable concern about
the ability of EMAP to achieve this goal, specifically "the assessment of
[ecological] status, the detection and characterization of changes or trends
in the status of ecological resources of the country, and the establishment
of associations between ecological condition and human-induced stresses."
The committee has been concerned that the development of an indicator
strategy and of a landscape characterization component (the cornerstones
of EMAP) is lagging far behind other components of the program. Instead,
EMAP has evolved from the bottom up and only recently has the program
begun work on the indicator strategy development and landscape
characterization components. The committee believes that the lag in
developing these cornerstones could lead to a fragmented monitoring
program with little hope of integration and significant environmental
evaluation.
One comment made in EMAP's response (EPA, 1993a) to the Interim
Report was that the committee was not adequately aware of program
developments. However, the committee believes that most of the
comments made in the committee's interim report to EPA addressed
fundamental issues that remain unresolved. Some 18 months after the
Interim Report was written and after many additional reports and briefings
from EMAP, most of the concerns of the committee remain valid.
-------
Overview 11
There has been, of course, incremental progress in many areas, for
example, development of EMAP Center, interagency cooperation, and the
hiring of scientists and managers to head resource groups. One area of
fundamental concern is the need for better use of scientists from outside
EMAP in the review process. Scientific review committees for the resource
groups are not yet in place and there is no mechanism to ensure that the
best suggestions are incorporated into the development of programs. The
committee endorses the comments in Dr. Paul Risser's letter to Erich
Bretthauer (Risser, 1992) that "the credibility of EMAP will rise with greater
use of competitive funding and open decision-making processes."
GOALS AND OBJECTIVES
The overall mission statement of EMAP is "to monitor indicators of the
condition of our nation's ecological resources to respond to the growing
demand for information characterizing the condition of our environment and
the type and location of changes in our environment. Simultaneous
monitoring of pollutants and environmental changes will allow us to identify
likely causes of adverse changes." EMAP was thus clearly intended to
provide information for the Administrator of EPA, Congress, and the public
to inform policy decisions; this seems appropriate—even essential—for a
federal regulatory agency such as EPA. The committee still believes that
this overall goal is laudable.
The objectives as stated in a revised form in the May 1993 EMAP
Program Guide (EPA, 1993b) are
1. Estimate the current status, trends, and changes in selected
indicators of condition of the nations ecological resources on a regional
basis with known confidence.
2. Estimate the geographic coverage and extent of the nations ecol-
ogical resources with known confidence.
3. Seek associations between selected indicators or natural and
anthropogenic stresses and indicators of the condition of ecological
resources.
4. Provide annual statistical summaries and periodic assessments of
the nations ecological resources.
The primary difference between this statement of objectives and earlier
ones is that the analysis of cause-effect relationships has been downplayed.
Instead, EMAP now states (perhaps more realistically) that it only will "seek
-------
12 Review of EPA's EMAP
associations" between indicators of stress and indicators of ecological
condition. Whether this meets the needs of policy makers or not is an
arguable point. Further discussion concerning indicators can be found on
page 14.
ASSOCIATIONS AND CAUSALITY
Even though EMAP is not designed to study cause-effect relationships,
no meaningful monitoring program can be completely devoid of cause-effect
considerations, in theory, if EMAP could thoroughly monitor all, or a great
many, environmental variables, it could identify many changes or trends.
Then, either EMAP or another program could study the changing variables
and attempt to find out why they were changing. But such a program would
be enormously ambitious, expensive, and inefficient, because it would not
explicitly incorporate prior scientific knowledge in the selection of variables
and sampling design. An alternative approach, the one EMAP purports to
use, is to select measurement variables based on explicit consideration of
the factors that would cause them to change over time or space. The
advantage of this approach is that a select subset of variables are measured
and that these variables can be tied to known or suspected cause-effect
relationships with human-induced or natural stressors. Although detection
of changes in these variables would not in and of itself formally establish
cause-effect relationships, additional focused research could address cause
and effect.
How assumptions of causality are made and the roles those
assumptions play in the overall design of EMAP are unclear at this point, but
they may be critical to the success of EMAP. The committee is concerned
with the degree to which EMAP is following its own guidelines for indicator
selection. Indicators are supposed to be selected in a "stressor cognitive"
manner. This means that an indicator is selected for measurement because
it can be linked to known or suspected cause-effect relationships. This
linkage requires a conceptual model or models of how the ecosystem type
functions, what the likely stressors are, and how the stressors interact with
ecosystem state or function. This is a critical step that infuses science into
the monitoring program. Indicators must be chosen based on a conceptual
model of ecosystem structure and functioning, in addition, for policy
relevance, stressor cognition is needed. A theoretical underpinning has not
been consistently described and used within EMAP.
The importance of causality and the interdependency of objectives is
again apparent when considering assessment questions, if, as stated in
-------
Overview 13
Table 3.3 of EPAs Office of Research and Development Workshop Summary
of January 7,1993, 'Assessment activities are critical to EMAP..." and "EMAP
is driven in large measure by assessment questions," then objective #4
("Provide annual statistical summaries and periodic assessments of the
nation's ecological resources") must be a high priority (EPA, 1993C).
Presumably, the monitoring is done to permit assessment of environ-
mental trends and to help evaluate policies. Moreover, the assessment
questions should be defined by both policy needs and state-of-the-art
scientific knowledge about causality. Currently, these questions and
assumptions seem to have been made on an ad hoc basis within individual
resource groups with little guidance, integration, or oversight between
groups. It appears to the committee that the question of causality needs
more focused planning by EMAP officials.
DETECTION OF TRENDS
In its interim report, this committee expressed concern that EMAP would
have difficulty in detecting temporal trends because of an overemphasis on
spatial sampling. The committee continues to have this concern. Even the
most recent version of EMAP's objectives (EPA, 1993a) does not address
the substantial difficulties in establishing trends in resource condition at the
broad spatial scale proposed to be addressed by EMAP.
The committee foresees many technical difficulties in detecting
meaningful trends at scales relevant to policy decisions (semi-decadal and
regional). These are exacerbated by the coarseness of EMAP's interpreta-
tion of "regional basis" and the political, rather than ecological, basis of the
proposed regions. EMAP has chosen to use standard federal regions estab-
lished by the Office of Management and Budget in its documentation, al-
though the relationships of those regions to what the actual pilot studies are
using is unclear. These regions are not standard between federal agencies,
as a comparison of the regions defined by the National Park Service, the
National Marine Fisheries Service, the Fish and Wildlife Service, and EPA will
show. In addition, they do not capture known geographic, climatological,
or ecological regimes or processes.
The environments within EMAP's resource types and regions are extra-
ordinarily heterogeneous. Even in more homogeneous ecological regions,
trends at different sample points are often likely not to be concordant. This
natural variability will mean that a large-scale regional approach will fail to
detect many important subregional trends. The committee is concerned that
EMAP is focused on broad spatial assessments to the detriment of assess-
-------
14 Review of EPA's EMAP
ment of trends at spatial scales that are meaningful in both an ecological
sense and a resource-management context.
For some assessment questions, the appropriate scale of resolution may
be an ecoregion or cluster of ecoregions. (Nationwide, there are 78 eco-
regions. These were developed by an EPA scientist (Omernik, 1987) based
on similarities in soil, climate, and natural vegetation cover, and they are
being used increasingly by resource managers to delineate geographic
regions that are fairly homogeneous ecological units.) For other assessment
questions, the appropriate level of aggregation may be a set of subsystems
within an ecoregion, cluster of ecoregions, or even a group of EPA regions.
For example, with respect to eutrophication problems in lakes, pertinent
assessment questions may include: Are trophic conditions improving or
worsening in urban lakes of temperate regions? In lakes where row crop
production is the dominant land use? In lakes with substantion recreational
development?
Additionally, the committee remains convinced that EMAP has focused
its programmatic design efforts too narrowly on achieving broad spatial
coverage at the cost of temporal trend detection. In continuing to make this
criticism, the committee wishes to emphasize that it is not diminishing the
importance of spatial assessments, and it supports the concept of
probability-based sampling to achieve unbiased population estimates for
such assessment of status. However, assessment of temporal trends is an
equally valuable goal for EMAP, and additional sampling strategies may be
needed to achieve this goal in a cost-effective manner and in time-frames
that are meaningful for resource policy analysis and management. This
apparent lack of commitment to evaluating the best approach for temporal
trend detection is a serious flaw in EMAP program development.
INDICATORS
Indicators are the very heart of EMAP, but the development of an
indicator strategy for the entire program has still not been achieved.
Although indicators have been chosen within the resource groups, a
comprehensive theory that can be applied to the choice of indicators in all
resource areas is missing. EMAP cannot achieve its objectives without
satisfactorily resolving the problem of how to choose indicators that will
provide the information required of the program. The choice of indicators
has both scientific and policy aspects.
To estimate trends or changes "with known confidence" over multiple
periods (EPA, 1993a) will require long-term monitoring records; thus indica-
-------
Overview 15
tors cannot change on a yearly basis. However, rapid advances in science
and rapid evolution in public perceptions of environmental values and in
public policies on the environment can be expected to create the demand
for assessments involving new questions and new values. A set of "perma-
nent" indicators should be perpetually monitored with new ones added oc-
casionally but not necessarily at the expense of dropping the original ones.
Currently, there is no comprehensive philosophy or theory to guide the
choice of indicators, receptors, or stressors that can be applied to all the
resource groups, because the Indicator Development Strategy is still
incomplete.
INFORMATION MANAGEMENT
EMAP will have major information management and processing needs,
and the success of EMAP will depend to a significant degree on an ade-
quate information system. To date, however, the committee has received
only one document that deals with this set of issues (Information Manage-
ment Strategic Plan; EPA, 1992a). The committee finds the plan to be vague
and overly general. The document does not describe how EMAP will incor-
porate current and developing technology. There is not a single reference
in this document to any substantive articles that deal with the current state
of research on scientific data base management. In particular, the com-
mittee notes the following concerns.
1. There apparently has been no analysis of user requirements that
details to any satisfactory degree (a) the various classes of users, their
priorities, and requirements, (b) the numbers, sizes, distribution, and hetero-
geneity of data sets produced internally by EMAP or imported from other
sources; and (c) the processing needs of scientists and policy makers who
will use EMAP data.
2. There is currently no real system "design" nor specific information
concerning what the system will do and for whom. Neither the Zachman
framework nor the conceptual systems architecture diagram is a design or
plan, and it is not clear how they will lead to a design or plan. The EMAP-
information management personnel have spent more time thinking about an
approach that may lead to a design than designing a system. A first priority
should have been a thorough analysis of requirements.
3. There is, as yet, no concrete plan ,for short-term or long-term
implementation of specific software and hardware systems. EMAP needs
such a plan for either the choice or design of an adequate data base
-------
16 Review of EPA's EMAP
management system (such as an object-oriented system) and the pro-
cessing software that will be supported by such a system.
4. It is the general consensus of the committee that current relational
data base management systems (e.g., ORACLE) and Geographical informa-
tion Systems (GISs) (e.g., ARC/INFO) are inadequate for the storage,
access, processing, and display of large scientific data bases containing
spatiotemporal information. In particular, the current use of the Statistical
Analysis System is inadequate for managing scientific data bases as large
and complicated as the one EMAP will develop. The acquisition of commer-
cial relational data base management systems is also inappropriate for the
expected sizes of data bases (0.35 terabytes by 1996) and the committee's
understanding of the processing needs.
Any information system plan for EMAP should have both long-term and
short-term components. Both should be based on an adequate analysis of
requirements, which the committee has not seen. The people who collect
the data, those who will use the data, and those who design and manage
the information system must communicate with each other so that the
system will be of greatest use. Therefore, the plan and solutions should be
well developed, well described, and well understood by all of the managers
and the resource groups of EMAP, and it should be common to all of them,
otherwise there will be data-management chaos.
The short-term component should solve the data-management issues for
the current planning and exploratory stages of EMAP (say 2-3 years). This
probably requires the use of relational technology (e.g., ORACLE, SYBASE)
or GIS technology that is based on relational technology (e.g., ARC-INFO).
The plan should also address short-term statistical and other processing
requirements. The above relational systems currently provide adequate
interfaces to statistical packages such as SAS, as do some other systems.
Relational technology, however, was developed for business and not for
science applications, and has serious shortcomings for science. For this
reason, a great deal of research activity is currently taking place with respect
to scientific database systems.
The long-term component of the plan should have at least the following
parts:
1. In particular the long-term component of the plan should address
issues concerning the incorporation of, and access to, tertiary mass-storage
devices into the computational support system. Since the ultimate size of
the EMAP database is likely to be in the (low) terabyte range, the use of
such devices is probably unavoidable and questions concerning adequate
-------
Overview 17
access to the data stored on such devices are important. Similar issues are,
for example, a major focus of attention in current research, and in particular,
in research concerning NASAs EOS-DIS.
2. A common data model and associated manipulation and query
languages that permit the representation of most of the important concep-
tual entities. The model should also permit the expression of the operations
on these entities. Such a model should probably be based on some com-
bination of object-oriented, semantic, and deductive database ideas. Appro-
priate models currently exist. There are also emerging federal standards for
spatial data and meta-data exchange forts (e.g., Spatial Data Transfer
Standard (SDTS).
3. An open and extensible meta-database system (MDBS) architecture
that can support the data model and provide the standard services of an
MDBS, and that also provides ease of interfacing with respect to the various
statistical, GIS, and other analytical support modules that EMAP scientists
require for their work. There are now some well-respected MDBSs
becoming commercially available (e.g., 02, the commercial version of
POSTGRES), and within two years or so, it is likely that systems will be
adequate for EMAP's needs.
4. Storage, archiving, and access facilities that are adequate for both
the large data entities and the large total volume of data.
5. Adequate network and communications architecture.
6. Appropriate (visual) interface designs.
An important part of EMAP's challenge is to have a plan for the transition
from the short-term system to the long-term one. Finally, as much as
possible, EMAP should advantage of the expertise in spatial data processing
and handling that is being developed at other federal agencies, such as the
Defense Mapping Agency and the USGS.
COST
Although the committee's primary concern is with reviewing the scientific
and technical foundation of the EMAP program, it remains concerned about
the programs cost-effectiveness. Ultimately, someone will have to answer
the question: "Is it worth it?" The committee does not intend to provide a
judgement on this question. However, it does believe that it is important that
the proper information on cost be collected to ensure that this question can
be answered intelligently.
-------
18 Review of EPA's EMAP
In this respect, the committee has been concerned about its inability to
obtain accurate cost estimates. The committee recognizes that budget
estimates cannot be precise; it does not expect them to be. But the
committee has received conflicting and undocumented responses, and it
does not appear that an adequate and thorough attempt has been made to
arrive at a realistic budget estimate. Furthermore, the committee has never
seen a first approximation of a budget justification. This committee, of
course, is not in a position to make such an estimate in any detail; so its
concerns over costs are communicated with substantial uncertainty
Nonetheless, the experience of many committee members with a variety of
large environmental research and monitoring programs provides at least a
rough basis for the committee's concerns.
When accurate cost data are available, two fundamental questions need
to be addressed. The first is whether the EMAP program is the most cost-
effective manner of obtaining the information it will collect. The second
question is; How and at what scale will EMAP integrate geographically
referenced data into the landscape characterization, and how will it manipu-
late and make such data available for use?
With respect to the first question, the committee has, throughout its
review, raised questions regarding whether alternative approaches and
improved integration with other monitoring programs might not provide data
that are at least as useful as those that EMAP will provide but at substantially
reduced cost.
One alternative approach to sampling that might be less costly is to
coordinate field sampling among the various resource groups within EMAP
hexagons. This might be done, for example, by randomly choosing
watersheds, or portions of watersheds, within a sampling hexagon and
having each appropriate resource group locate its sample site within the
watershed. This type of approach would have the advantage of co-locating
sampling sites, thereby increasing the likelihood that changes in one
resource type within a watershed might be linked with changes in another
resource type. Another advantage of this approach is that it would force
greater communication and cooperation among the diverse resource
groups.
Although the committee judges that Objective #1, "estimating the
current status, trends, and changes in selected indicators of the condition
of the nation's ecological resources on a regional basis with known
confidence," is attainable, it believes that this would be a very costly
environmental program. It seems likely that Objective #2, "estimating the
geographic coverage and extent of the nations ecological resources with
known confidence," would be at least as costly. Furthermore, Objective #1
-------
Overview 79
really cannot be achieved without achieving Objective #2. For EMAP to
establish the status of resources, it must know what and where the
resources are.
With respect to the second question concerning geographic coverage,
one of the greatest expenses is the integration of all data across resources
in the landscape component of EMAP. At this stage, the committee is
unable to understand how status data from the probability-based sampling
will be Integrated with cover-type data from thematic mapper imagery. Even
after repeated inquiries, the committee is still quite uncertain as to how and
at what scale EMAP will integrate geographically referenced data into the
landscape characterization and how it will manipulate and make such data
available for use. It is not clear that resources would be sampled within the
same watershed, and thus there is little basis for integrating data.
Although EMAP may be an expensive monitoring program, it will also
provide information that is not currently available elsewhere. EPA and
Congress will have to make the final judgement about whether this
additional information is worth the cost of obtaining it.
MANAGEMENT AND COORDINATION
The issue of coordination is perhaps as serious as any the committee
has addressed. Evolution of scientific programs is essential, and the com-
mittee in no fashion suggests that EMAP should avoid this healthy and invig-
orating process. But EMAP thus far appears to have evolved in a piecemeal
fashion, without clear guidance from a center or from its cornerstones—the
indicator development strategy and the landscape characterization com-
ponents.
There are signs that this situation is changing, such as the development
of EMAP Center in North Carolina and the hiring of scientist-managers to
head the resource groups. These are significant signs of progress toward
a better managed and coordinated program. However, individual resource
groups are conducting pilot studies with indicators that do not always meet
the indicator criteria that have been described. The revised indicator
strategy is not available, and the committee does not know what the new
criteria will look like. Presumably, the people in the resource groups do not
know either. It is not clear whether the individuals who are developing or
refining the indicator strategy fully appreciate the difficulties and costs
incurred by those performing the pilot studies in measuring indicators.
Furthermore, different resource groups are using different classification
schemes; for example, the surface waters resource group classifies the land-
-------
20 Review of EPA's EMAP
scape surrounding lakes and streams using a procedure that differs from
that used by the forests group. Thus, data from the two are incompatible,
and EMAP will fail to integrate trends across these two resources that
obviously interact because changes in forests affect the water quality of
streams and lakes. Procedures that ensure integration across resource
groups are therefore of as high priority as those that ensure quality of data
within resource groups. Yet, EMAP has placed much greater priority on tf?e
latter than on cross-resource procedures. A coherent, consistent,
comprehensive strategic plan for EMAP, with logical charts showing the
flow of thought and activity through the goals and objectives, should be the
objective of the highest echelons of EPA that are involved with EMAP.
1NTERAGENCY AND INTRA-AGENCY COOPERATION
The committee is encouraged by the extent to which the EMAP program
has been pursuing opportunities for interagency cooperation. Both the
estuaries and the forest studies have been cooperative efforts with other
agencies, and the EMAP staff has been exploring other opportunities as well.
Such cooperation can reduce the cost and significantly increase the
usefulness of the monitoring programs. It will also result in the different
government agencies basing their natural resource programs on consistent
information.
The EMAP program should make better use of sampling and monitoring
programs already developed by other agencies and other parts of EPA.
Further development of the Regional EMAP program (REMAP) also could
lead to increased efficiency while actually enhancing the detail at which
information is obtained in critical areas. More use of these programs would
significantly reduce the cost of the EMAP monitoring program without sacri-
ficing the quality of the information obtained. The committee recommends
that EMAP personnel continue to aggressively seek out such opportunities.
As examples of EMAP's strengths and weaknesses, the committee looks
next at the forests and estuaries components of the program in chapters 2
and 3.
-------
Estuaries
BACKGROUND
The estuaries component of EMAP is a joint EPA/National Oceanic and
Atmospheric Administration (NOAA) program designed for long-term
monitoring of the waters, sediment, and biota from the head of tide to the
Outer Continental Shelf. EMAP-Estuaries is well developed and is the major
focus of this chapter. Two other EMAP components, focusing on the Great
Lakes and on coastal regions, are still embryonic. The program is intended
to complement NOAA% ongoing Status and Trends Program for marine
environmental quality. Eventually, these two programs may merge to
produce a single, cooperative coastal and estuarine monitoring program.
The original objectives for EMAP-Estuaries were presented in the
EMAP-Near Coastal Program Plan for 1990: Estuaries (EPA, 1990). These
objectives have since been refined based upon interactions with various
review groups, including this NRC committee. The current objectives are as
follows.
• Provide estimates of status and trends in the condition of estuarine
and coastal resources by measuring selected indicators of ecological condi-
tion.
• Measure changes in the extent of the nation's estuarine and coastal
resources on a regional basis.
• Identify and evaluate associations between ind icators of ecological
condition for the nation's estuarine and coastal resources and selected
indicators of environmental stress, both natural and anthropogenic.
• Evaluate the progress of pollution control actions and environ-
mental policies in achieving national goals for the condition of estuarine and
coastal resources.
21
-------
22 Review of EPA's EMAP
In addition to meeting the overall objectives of EMAP, the estuaries
component of EMAP addresses general regional-scale environmental
problems by measuring indicators of ecological condition associated with
such problems. Problems applicable to estuarine waters are low dissolved-
oxygen concentrations, eutrophication, chemical and biological contamina-
tion, habitat modification, and cumulative impacts of these stresses.
EMAP-Estuaries is intended to provide a different type of information
from that already available or currently being provided. One difference is
spatial; the ecological condition of entire estuaries and rivers is to be
characterized on a regional basis, instead of at separate sites in estuaries or
rivers. What is more, the data are to be reported as a cumulative distribution
function that gives the estuarine area (as a percent area plus an error
estimation) for each value of each indicator of ecological condition. For
example, EMAP will be able to state that 50 percent of the area of large
estuaries had a benthic index of 5.0 with 90 percent confidence limits of 4.5
to 5.5. A second difference is that similar measurements will be made in
estuaries and coastal rivers across broad regions and even over the entire
U.S. coastline. Although some detailed extent and condition data exist for
portions of specific estuaries, this is the first systematic attempt to assess
condition in a consistent way, nationwide, for all estuaries. Finally, the
program will be long term so that changes in the condition (if any) can also
potentially be evaluated.
Although one of the currently stated goals of EMAP-Estuaries is to
"measure changes in the extent of the nation's estuarine and coastal
resources on a regional basis," the program will not measure changes in the
areal extent of estuaries except in the case of coastal wetlands. The
committee has not been given any information about how even this wetland
area determination will be done, but it is not an especially difficult task. It is
unfortunate that 20 years of data do not already exist so that informed
statements could be made now about the condition of and changes in this
important ecological resource of the near-coastal region of the United States.
Careful consideration should be given to whether or not assessment of
changes in the area! extent of other aspects of estuarine and coastal
resources should also be addressed.
SCOPE OF REPORT
EMAP-Estuaries was the first portion of the overall EMAP program to be
implemented. Shown in Figure 2-1 are the geographic provinces that make
up the nation's coastal zone. A full-scale demonstration project was initiated
-------
Columbian
C«lllornl«n v>;£
Acadian
FIGURE 2-1 EMAP biogeographical provinces for estuaries.
Source: Weisberg et al., 1992.
CO
-------
24 Review of EPA's EMAP
in the summer of 1990 in the Virginian Province at 217 sites in estuaries
between Cape Cod and the mouth of Chesapeake Bay (see Figure 2-2).
The Virginian Province research has continued; demonstration projects
were carried out in 1990, 1991,1992, and 1993, but only the report for the
1990 research has been issued. In 1991, a demonstration project was
initiated in the Louisianian Province, and this continued in 1992 and 1993 with
a data report issued for 1991. A pilot project for the Carolinian Province
began in 1993, and this is expected to become a demonstration project in
1994. Because of its early start, evolution of EMAP-Estuaries has been
influential in the development of the entire EMAP program.
This review is primarily based on the results of the 1990 Virginian
Demonstration Project. It is the only estuarine project for which the com-
mittee has been provided a final report, although a data report from the
Louisianian Province is published. The primary goals of the 1990 Virginian
Demonstration Project were to (1) determine the precision with which status
can be estimated using the present design, (2) estimate sources of variability
for selected indicators (dissolved oxygen and a benthic index), and (3)
determine the appropriate sampling period. These goals for 1990 were
adopted to support the larger goals of the overall EMAP-Estuaries program
to assess the status and trends in the condition of the nation's estuaries.
SAMPLING DESIGN
in developing the sampling approach for the estuaries component of
EMAP, a list of estuaries was used to define the population of estuaries in the
United States. From this, a new list was assembled to represent all estuarine
systems larger than one square mile within the Virginian Province. The
estuaries on this Virginian Province list were then divided into three sampling
strata representing small estuaries and tidal rivers, large tidal rivers, and large
estuaries. Such stratification into more homogeneous subgroups is a well-
accepted practice in statistical sampling procedures. The EMAP-Estuaries
sampling design is based on a single, annual sampling season (called an
index period) of each of these classes. Additional information on the
sampling program is found in the Virginian Demonstration Project Report
(Weisbergetal., 1992).
The general EMAP sampling grid is hexagonal; it allows systematic
sampling from a random start (i.e., a random orientation of the grid design).
The EMAP grid was applied directly to large estuaries to select 54 sampling
sites; these sites were the center points of hexagons 18-km apart. For large
tidal rivers, a "spine and rib" approach, a linear analog of the spatial sam-
-------
Estuaries
25
t Jl'l I f * ^J,
:^*xS/,
i
FIGURE 2-2 Location of the EMAP 1990 Virginian Demonstration Project
sample sites in the estuaries of the Virginia Province.
Source: Weisbergetal., 1992.
-------
26 Review of EPA's EMAP
pling grid, was used. The mouth of the river provided the starting point of the
"spine," and the first transect of "rib" was located randomly between 0- and
25-km downriver. Additional ribs were located 25 km from the first. Sam-
pling sites were then selected at random along the rib of each transect to
yield a total of 25 sampling sites per river. The 137 small estuarine systems
were listed from north to south and combined into groups of four; one
system was selected at random from each group. This ensured that the 32
sampled systems were dispersed geographically (Weisberg et al., 1992).
Landscape characterization does not appear to have been required for
development of this local sampling scheme.
Samples are to be collected on one quarter of the sampling sites each
year, returning to any particular site every fifth year. EMAP has considered
at least one other sampling alternative to the interpenetrating design
discussed above in its evaluation of designs. Urquhart et al. (1991) com-
pared the EMAP "augmented serially alternating design," where each set of
sites is visited every four years, with one set of sites being sampled every
year, with an "augmented rotating panel design," where a series of sites is
visited for several consecutive years (four in this example) and then removed
from future consideration. They found that the EMAP design had equal or
smaller variances for estimation of trend. They also found that the number
of required augmented sites (those sites measured each year) drops off
considerably after the first four years; hence the main value of augmentation
occurs during the initial cycle.
The interpenetrating sampling scheme was designed to detect long-term
monotonic trends in regional mean values. This is an appropriate design for
detecting many expected changes. It is possible, however, that the changes
might occur as step functions after some threshold is reached, or some other
nonlinear process might be operating. The temporal aspect of the design
could be significantly enhanced by also investigating methods for detecting
nonlinear trends. This would be an excellent area for further investigation,
especially for investigator-instigated research outside EPA.
It is the committee's opinion that EMAP has not given adequate
consideration to intensive site-based sampling for the analysis of trends.
Better resolution for trends might be gained if the sampling program
incorporated analysis and detailed site-based sampling at a representative
subset of locations within a region. Such spatial and time-intensive sampling
is not possible within the constraints currently embodied in the once-every-
fourth-year, interpenetrating grid design currently used by EMAP. Although
such an approach would not provide unbiased estimates of condition for an
entire region, It could provide information of sufficient resolution and quality
to detect trends in condition and evaluate the likely causes of any observed
-------
Estuaries 27
changes. Adoption of such an approach would not necessarily mean aban-
doning the interpenetrating grid design. Within the constraints of the
projected EMAP budget, limited sampling on the grid could continue to pro-
vide unbiased estimates of condition, while detailed sampling for temporal
trends would occur at a restricted number of representative sites.
It appears that other monitoring efforts also are adopting the EMAP
design (e.g., in Chesapeake Bay, in New York-New Jersey Harbor, in Long
Island Sound, and in Delaware Bay). By its very existence, the EMAP design
is becoming a national standard, at least for coastal waters in the demon-
stration provinces. Delaware and Maryland are proposing pilot projects, with
state-contributed matching funds, to test the application of the EMAP pro-
tocols to the assessment of inland bays and tidal rivers along the Atlantic
coast (Delaware Department of Natural Resources and Environmental Con-
trol and Maryland Department of the Environment, 1992). The two states
plan to use the EMAP design concepts but with much more intensive spatial
coverage.
In conclusion, the committee considers the basic EMAP probability-
based sampling approach to be generally satisfactory for the determination
of spatial extent of indices measured (but see comments on constraints in
the next section, Indicators). Nonetheless, the utility of the probability-
based, four-year interpenetrating design in assessing trends on the condition
of estuarine or coastal resources or in making associations with information
on stressors has yet to be adequately addressed and may be problematic.
The approach EMAP has selected provides the necessary spatial coverage
and has proved flexible enough to meet the problem of how to sample small
estuaries and rivers. The basic design has been shown to have the same or
smaller variances concerning estimation of trends when compared with the
augmented rotating-panel design. The adoption of the EMAP-Estuaries sam-
pling design by some other monitoring programs is a positive development
that will aid in the comparability of data obtained.
INDICATORS
Overview of EMAP-Estuaries Indicators
The environmental indicators measured by EMAP are at the heart of the
entire program. They are the key to evaluating the ecological condition of
the near-coastal ecosystems and to identifying and evaluating associations
between ecological condition and pollutant exposure. The choice of indica-
tors is difficult in any ecosystem, but estuarine and coastal ecosystems
-------
28 Review of EPA's EMAP
present extraordinary problems of variation caused by changes in the input
of freshwater, changes in circulation and flushing related to coastal storms,
and migration and seasonally of many of the organisms.
The EMAP sampling design greatly constrains the choice of indicators.
Probability-based sampling, broad geographic coverage, and limited funds
mean that relatively few samples can be collected over a very large and
diverse area. The decision was made to restrict the sampling over an entire
province to a single summer month when populations and activity of indig-
enous organisms are high and oxygen stress is at a maximum.
A reasonable set of trial measurements were chosen to begin the 1990
sampling seasons. A great deal of effort went in to testing one of the indica-
tors, dissolved oxygen concentration. Another indicator, an index calculated
from measurements of benthic populations, is promising but needs addi-
tional testing and development. While a complete suite of indicators is not
yet at hand, EMAP-Estuaries has made a reasonable start.
The Conceptual Framework for
EMAP-Estuaries Indicators
The EMAP-Estuaries indicators were not developed in a vacuum. In-
stead, there was an assumed overall plan or conceptual model of how the
coastal and estuarine components fit together as a system. In general,
reviewers (Including this committee) have agreed with EMAP's views of what
is /mportanf to measure and, ty implication, with EMAP's conceptual view
of the system. Now it is time for a more formal presentation of the con-
ceptual scheme or model of these coastal and estuarine systems that incor-
porates the probable linkages and controls. This part of the conceptual
framework will help direct indicator research and will make integration across
different estuaries and provinces possible.
The basic indicators used in the Virginian Demonstration Project evolved
from an Indicator Development Workshop (attended primarily by EPA scien-
tists) held in December 1989. In its current form, the EMAP indicator strategy
as presented by Weisberg etal. (1992) calls for the development of indicators
of ecological response, exposure indicators, habitat indicators, and stressor
indicators. The response indicators include the abundance, biomass, and
composition of benthic and fish communities; the exposure indicators in-
clude the concentration of dissolved oxygen, bioassays of water and sedi-
ment, and contaminant concentrations; the habitat indicators include water
depth, salinity, and sediment characteristics; and the stressor indicators
include freshwater discharge rates, climate, pollutant loadings, and human
-------
Estuaries 29
population and land use parameters. Stressor indicators are not discussed
further in the Virginian report.
The committee strongly believesthat EMAP-Estuaries should also include
indicators of ecosystem functioning. These indicators are difficult to mea-
sure when studies are made only once a year, but they can be approached
to some extent indirectly. For example, algal biomass can be used as a
surrogate for primary production. Chlorophyll a (related to algal biomass)
was scheduled to be assessed in the Virginian Province, but because of
problems experienced with sample handling, this measurement was dropped
after the first year. In the newly revised description of appropriate response
indicators for the whole of EMAP (presented by EMAP Director Martinko at
an NRC committee meeting in October 1992), measures of ecosystem pro-
ductivity were given high priority. Lack of such measures in EMAP-Estuaries
should be addressed as soon as possible. Remote sensing provides one
possibility for chlorophyll measurements on a regional scale. This has been
done in Chesapeake Bay and could be easily carried out for an entire
province.
Also missing in the Virginian Demonstration Project is another measure
of important coastal habitat, submerged aquatic vegetation. The submerged
aquatic vegetation is also a candidate for measurement in the EPA report
Monitoring Guidance for the National Estuaries Program (EPA, 1991 a).
Inclusion of the submerged aquatic vegetation into the sampling scheme for
all estuaries demonstration projects should be addressed. Remote sensing
may be a good technique for mapping it regionally.
Finally, it should be kept in mind that most of the base indicators chosen
so far respond similarly to natural and human-produced stressor inputs. This
fact confounds both indicator development and interpretation. The com-
mittee points out that with general indicators, such as are currently in use by
EMAP-Estuaries, it may be impossible to separate the source of the stress-
ors. The differentiation between results due to anthropogenic inputs and
those due to natural stressor inputs will have to come from other information
and perhaps from intensively studied sites.
Description of the EMAP-Estuaries Indicators
During the 1990 Virginian Demonstration Project, a wide variety of
samples were taken for development of response, exposure, and habitat
indicators. These included species composition, number, and biomass of
infaunal organisms (i.e., those living within the substrate); profiles of
sediments (cross-sectional images taken with a camera inserted into the top
-------
30 Review of EPA's EMAP
portion of the sediment); species composition, number, biomass, and visual
pathology of fish; abundance of large bivalves; temperature/salinity/depth
profiles of the water column; concentration of dissolved oxygen and chloro-
phyll a in the water column samples; toxicity of the sediment and water
column to standard test organisms; abundance of Clostridium spores and
sediment grain size; presence or absence of marine debris; and concentra-
tion of contaminants in sediments, fish muscle, and bivalve tissues.
The logic behind the types of chemical analyses chosen is not clear. It
appears that EMAP simply adopted the NQAA Status and Trends chemistry
program without very much, if any, consideration of what it would or would
not provide in the way of data or answers. There is no provision for seeking
or identifying any substance not on the preselected list of compounds and
the relevance of any samples taken to EMAPs goals should be made clear.
As might be expected in a demonstration study, only some of the
measurements and analyses proved to be useful indicators. Logistic
difficulties precluded the completion of some measurements; others were
completed but have not yet been analyzed in detail. EMAP-Estuaries
attempted to develop two integrated indicators of response, one for benthic
communities and one for fish. Of these indicators, the benthic index showed
the greatest promise in assessing the regional status of biota.
The index assigns a numeric value to a sample site based on a scale
developed from benthic animal characteristics from nearby undegraded
versus degraded sites. Criteria for a degraded environment included low
dissolved oxygen, elevated levels of at least one chemical contaminant
measured against a scale developed by Long and Morgan (Weisberg et al.,
1992), and sediment toxicity as measured by a standard amphipod bioassay.
EMAP investigators designated the degraded and undegraded sites accord-
ing to an overall assessment of condition by experienced benthic ecologists.
Unfortunately, the assessment included some of the measures that were then
used to develop the benthic index. This does introduce some circularity in
the interpretation of the results. EMAP was unable to develop an indepen-
dent measure of undegraded or degraded environments. From this analysis,
a benthic index composed of expected numbers of species, numbers of pol-
lution-tolerant and sensitive species, abundance of bivalves, and weights of
individual polychaetes was constructed. Depending on the specific formu-
lation used (i.e., certain predictors "forced" into the equation first, or not), the
benthic index was able to correctly classify undegraded or degraded sites 79
to 84 percent of the time. The values of the index computed on the stations
from September 1990 samples agreed with the values computed from the
August 1990 samples 80 to 84 percent of the time, depending upon which
index was used.
-------
Estuaries 31
The index is attractive because it is relatively simple to use and makes
use of large amounts of data from a wide range of organisms that are
exposed to the worst of contaminant-imposed stress. However, since
infaunal populations are notoriously seasonal and site-specific, analysis of
seasonal and region-specific variance among both undegraded and de-
graded sites is essential. Furthermore, the validity of using weighing factors
(applied to correct for community structure changes in some parameters
correlating with salinity) should be examined more thoroughly. Although
variance may be reduced by the use of weighing factors, such manipulations
may artificially skew the data and support erroneous conclusions about what
constitutes a "degraded" population.
The EMAP-Estuaries program is also attempting to develop an indicator
index based upon fish population structure, pathology, and contaminant
burden. While there can be no doubt that fish could be important and
sensitive indicators, it is doubtful that the limited sampling planned for
estuarine waters will allow these mobile and migratory animals to be used.
A major problem is the limited number of samples that can be taken at a
single time of the year. Given this limitation, EMAP cannot collect acceptable
population data and assess population structure.
Another aspect of the sampling program is the gross pathology of fishes.
Sampling wild fish and recording their gross anomalies is subject to a high
degree of variability. The diagnosis itself is subjective and next to impossible
to do with relatively untrained field crews. Even if the field measurements
were excellent, questions would remain about interpretation of gross pathol-
ogy that may be related to nutrition, physical injury, or infectious disease as
well as toxicity. Additional problems arise because there are significant
seasonal changes of infestation rates offish diseases. If the gross pathology
of fish is to be used as an indicator, research must be pursued on ways to
differentiate the various causes of the pathology.
The analysis of fish flesh samples for contaminants was not completed
for the 1990 samples. The committee believes, however, that there is a
fundamental flaw in the fish sampling program for population-level effects,
and contaminant burdens cannot be effectively estimated without adjusting
for the size (age) of individuals. This work should not be continued in its
present form. Examples from other large studies, in particular those of the
International Council for the Exploration of the Sea, demonstrate that an
extremely large sampling effort is needed for effect and contaminant studies
in estuarine systems. Unless a large amount of additional effort is expended,
the value of the data obtained will be minimal.
Dissolved oxygen was assessed in detail during the 1990 demonstration
project. Single-time-point measurements were supplemented with contin-
-------
32 Review of EPA's EMAP
uous measurements at 13 sites. Time-series analyses indicated that although
dissolved oxygen is subject to large diurnal and tidal variations, single-time-
point measurements were adequate to describe the average dissolved oxy-
gen for an entire region. In contrast to this regional estimate, an estimate of
the variance in dissolved oxygen conditions for a single site would require
continuous recording for a period of 14 days. Although there are limitations
to these measurements, the thorough evaluation of the variation in dissolved
oxygen is to be commended. Unfortunately, a similarly detailed evaluation
will need to be performed in all new regions to be assessed before single-
time-point measurements can be adopted.
EVOLUTION OF EMAP-ESTUARIES INDICATORS
At a meeting in Narragansett, Rhode Island in September 1992, a
subgroup of this committee asked EMAP staff how results of the demon-
stration projects affected what was done in subsequent years. In the
Virginian Province, EMAP staff made the decision to keep the base set of
indicators initiated in 1990 throughout a complete four- year cycle.
Indicators were dropped only when they proved to be too difficult for field
preservation or logistically impossible (e.g., chlorophyll a, water-column
toxicity, number of large bivalves). Results from previous years were used
to improve the sampling scheme. For instance, approximately 10 percent of
the sites were revisited in the second year to provide data on interannual
variation.
Lessons learned in the Virginian Province also led to development of
additional indicators in the Louisianian Province Demonstration Project,
where more experimental indicators are being assessed, particularly those
dealing with biomarkers and functional attributes of the ecosystem.
Apparently the same will be true of the Carolinian Province. Design
workshops in 1991 and 1992 were used to evaluate the design and imple-
ment enhancements in the base sampling-strategy in the Virginian Province.
The 1992 workshop was attended by representatives from EMAP-Estuaries
as well as representatives from the EMAP Design and Statistics Task Group.
At this stage in the evolution of the indicators, EMAP-Estuaries is
exhibiting a laudable degree of flexibility and experimentation. For example,
each demonstration project has been implemented in a different way. The
Virginian Demonstration Project was planned and implemented primarily by
consulting firms (Versarand Science Application International Corporation),
the Louisianian project by a consortium of universities, and the [Carolinian
-------
Estuaries 33
project jointly with NOAA through the Status and Trends Program and Sea
Grant. The next step has not been stated by EMAP-Estuaries, but the
committee recommends detailed analyses of the successes and failures of
these efforts and incorporation of the best ideas into the indicator plan for
EMAP-Estuaries before proceeding with further demonstration projects or
with full-scale implementation.
One part of EMAP-Estuaries' effort has been to identify of a number of
historical data sets in the Virginian Province (Menzie Cura and Associates,
Inc., 1990). Out of 18 studies investigated, eight were found to contain
information important to EMAP and in particular to EMAP-Estuaries (Menzie
Cura and Associates, Inc., 1990). These eight cover estuaries of the East
Coast from North Carolina to Massachusetts, followed documented proce-
dures for field collections and chemical analyses, and have data available on
computer. Unfortunately, no use has yet been made of these data sets.
They should be used to provide temporal and spatial information for various
indicators; this would undoubtedly enhance the indicator development work
and simulation results on general properties of the data. Detailed analysis
of this material would provide valuable information on spatial and temporal
variability and the power of certain indicators to detect trends within a given
time period, ft is past time for this work to be completed.
EMAP-Estuaries recently began investigator-initiated, competitively
reviewed research on new indicators and on the further development of
those already being used. Areas of research that should be examined
include the analysis of long-term data sets from various sites to examine
indicator variability and its causes and use of molecular probes to look for
the presence of enzymes indicating pollutant exposure or changes in eco-
system function (e.g., nitrogen fixation). Research aimed at developing and
validating indicators of ecological function and stressor impact would be an
excellent way to involve non-EPA scientists in the EMAP program.
DATA COLLECTION, ANALYSIS, AND MANAGEMENT
Training of field crews and quality assurance on data collected appear to
be exemplary. During the second and third sampling periods of the 1990
demonstration project, samples were collected at more than 95 percent of
stations visited for most parameters of interest. Analysis of samples was also
completed with a high degree of success. Serious problems were only ob-
served with organic analysis of bivalve and fish tissue and chlorophyll a (< 10
percent passed quality assurance); less serious problems were observed
with organic analysis of sediment and datasonde deployments (73 percent
-------
34 Review of EPA's EMAP
passed quality assurance). All other measurements passed quality assur-
ance 85 percent of the time or more.
The EMAP-Estuaries Information Management group has developed a
comprehensive approach to information management, including protocols
and apparatus for data collection in the field, data verification and validation,
and data-base management. All field data are available to logistics personnel
in data bases the day after sampling. Three levels of data aggregation have
been developed and documented for the EMAP-Estuaries programs: raw
field and analytical data, verified and validated data bases with associated
documentation, and station/date aggregates of results for indicators and
supporting data. Raw data are stored and maintained regionally. This is a
rational approach, because monitoring data should be stored and main-
tained by the groups who generate the data and who have an interest in and
a familiarity with the data The aggregated data are stored at the Central
Estuaries Information Center located in Narragansett, Rhode Island. The
aggregated data will be used to make regional estimates of status, which will
be stored on a central EMAP information system.
Data-base management currently uses the Statistical Analysis System,
because contractual vehicles for the procurement of relational data-base
management tools have not been in place within EPA. While this may be
satisfactory in the short term, in the long term it will pose many limitations in
data management and data distribution. The committee recommends that
EMAP change from the Statistical Analysis System to a standard data-
storage and retrieval-management package in the near future.
At the moment, specific subsets of the EMAP-Estuaries raw data base
can be obtained (on electronic media) by request to EMAP-Estuaries data
center at the EPA Environmental Research Laboratory in Narragansett,
Rhode Island. An on-line system has also been developed that will allow any
user on the EPA Wide Area Network or on a personal computer with a
modem to gain access to the EMAP-Estuaries data base. This system is
currently in a prototype form and is being used by researchers at six
locations. Plans have been made to distribute data via an on-line interface,
CD-ROM (compact disk-read-only memory), and a central distribution
system. Given the wide audience for these data, it is unlikely that any one
vehicle of data distribution will meet the needs of the majority of users.
Weaknesses still exist in the Interface between the EMAP-Estuaries
information system and other resource groups within EMAP, other informa-
tion systems within EPA, and information systems in other federal and state
agencies. EMAP should be more active in building interfaces to these other
systems to allow easy exchange. The EMAP-Estuaries Information Manage-
ment group has started to work with NOAA to develop integrated data bases.
These efforts should be accelerated.
-------
Estuaries 35
The committee suggests that other multivariate analysis techniques be
carefully explored in addition to the linear discriminant function that was
applied to data for the classification of estuarine habitats in the 1990
Virginian Demonstration Project report. Logistic regression may be appro-
priate in cases where one is interested in determining how observed pro-
portions (e.g., the probability that a site is degraded) depend on certain
predictor variables. Unlike discriminant analysis, these predictor variables
are not assumed to follow a multivariate normal distribution and may even be
categorical. Feinberg (1980) provided a good introduction to this type of
analysis. Other multivariate analyses, such as neural-network analysis,
multidimensional scaling, and various clustering and pattern-recpgnition
algorithms should be screened for their potential value in classifying estua-
rine environments. EMAP might also try methods of statistical computing
using visualization of multidimensional data (e.g., Cleveland and McGill,
1988; Becker et al., 1987). This might help to provide further insights into
relationships and underlying processes.
The committee concurs with some of the conclusions from the review of
EMAP-Estuaries by the Estuarine Research Federation Committee (Estuarine
Research Federation, 1992) concerning the consideration of several spatial-
temporal models reviewed by Cressie (1991). For example, predictors based
on the method of kriging (a method of spatial statistical analysis that takes
spatial correlation into account) could be used to obtain estimates of the
proportional area! coverage of degraded habitats (see also Gilbert and
Simpson, 1985). The Estuarine Research Federation also recommended that
EMAP-Estuaries consider nonlinear models for trend.
Synthesizing the data and providing it to management and scientific
users is another level of data analysis. The EMAP-Estuaries program has
provided several examples of how this might be done. One comes from the
Example Environmental Assessment Report for Estuaries (Frithsen et al.,
1991), which gives hypothetical conclusions based on an artificial data set
in order to exemplify the type of information that EMAP-Estuaries would
provide managers. The EMAP-Estuaries Virginian Province 1990 Demonstra-
tion Project Report (Weisberg et al., 1992) is a good first step in data
reporting. Cumulative distribution frequencies and regional averages in the
areal extent of nominal and subnominal values are presented for many
indices, as well as a thorough discussion of data supporting the development
of a few indices; nonetheless, this is only the first step. Much additional infor-
mation could be obtained from further analysis of these data. A mechanism
for ensuring that this happens needs to developed. The committee recom-
mends that the scientific and regulatory communities be included in the
analytical process. At a minimum, there should be review by an outside sci-
-------
36 Review of EPA's EMAP
entitle group to ensure that all likely biological relationships are addressed
based on understanding of ecosystem functioning and relation of data to
regulatory needs. Wherever appropriate, EMAP results should be published
in the peer-reviewed scientific literature.
Originally, the EMAP-Estuaries program planned to incorporate available
information on stressors into the interpretive reports. Some of this type of
information was presented in the Example Environmental Assessment Report
for Estuaries (Frithsen, et al., 1991), but so far there have been no examples
of how real information will be incorporated. An updated plan needs to be
prepared, especially in view of the reduced emphasis on stressors for the
entire EMAP as described to the committee by EMAP officials (EPA, 1993c).
APPLICATION OF DATA
How well will the data and information collected under EMAP-Estuaries
be applied to environmental needs? The answer to this question centers on
the use of the information by managers. Exactly how useful this information
will be can only be judged over five or ten years. For this report, the
committee can only judge whether or not the type of information is likely to
be useful, and if it is likely to be provided in a reasonable time and in an
understandable format. A secondary question concerns its possible use in
answering scientific questions.
The indicators chosen by EMAP-Estuaries, including the benthic index,
sediment toxicity and contaminants, and oxygen, will be useful to the
regulators and assessors of estuarine conditions. They are reasonable
indicators to measure at this stage. This endorsement does not say that
these particular indicators could not be improved.
The first step toward making the collected data applicable to the needs
of managers comes in planning just what to measure. EMAP-Estuaries had
several managers involved in the initial planning. Attendees at the indicator
workshops were a mix of scientists and EPA consultants. EPA regional
officials and state officials were involved later in planning the Louisianian
Demonstration Project.
The report on the Virginian Demonstration Project was well written,
making the information it covered accessible to all scientifically literate
readers, and the executive summary was aimed directly at managers. More
than 100 managers of estuarine resources also participated in a workshop
run by EMAP to answer the question "How useful are the indicator and
monitoring data being collected?"
-------
Estuaries 37
One interesting and potentially useful development has been the
Regional EMAP. In this program, EMAP will award a number of grants to
states through individual EPA regional offices to use the EMAP sampling
design and sampling protocols with increased sampling density to address
regional problems, such as pollution in New York Harbor.
Finally, the EMAP-Estuaries has developed a data base and data-
handling ability so that the data will be available to managers and scientists
through a data network. The prototype system, working at the EPA Environ-
mental Research Laboratory in Narragansett, Rhode Island, is accessible
through five nodes. Although the scientific applications of the data are not
yet clear, scientists will be interested in these data and also, in the
forthcoming Thematic Mapper coverage. The demonstration project of
complete Thematic Mapper coverage of the Chesapeake Bay watershed is
now available to scientists outside the EPA on tape and eventually will be on
compact disk. The open availability of EMAP-Estuaries data must be
maintained and expanded.
At the start of the Virginian Demonstration Project there evidently was
little communication between the EMAP planners and the users. As noted,
this has greatly improved, and the involvement of the managers should be
continued.
Currently the largest impediment to the use of the EMAP-Estuaries data
is the amount of time it is taking to release information synthesized from the
demonstration projects. The final version of the 1990 Virginian Demonstra-
tion Project was released only in June 1992. The data report for the
Louisianian 1991 Demonstration Project was released in 1993. If the results
of f/?ese demonstration projects are to be useful in leading the development
of EMAP-Estuaries, they must be accessible to review panels, the scientific
community, and planners and managers in a more timely manner.
SCIENTIFIC REVIEW
For the EMAP-Estuaries project the Estuarine Research Federation
convened a panel made up of well-respected and critical scientists. The
panel met three times over a two-year period (the last meeting was in March
1992) with EMAP personnel. This gave the panel the necessary continuity
and experience with EMAP as it evolved, ft was evident from the panel's
report that it felt its constructive comments and criticisms had been taken
seriously and that substantive progress had been made over the two years
of the panel's existence. Some of their questions, such as those about the
power of the indicators to measure the small changes expected, remain
-------
38 Review of EPA's EMAP
incompletely answered, and are also of concern to this committee. However,
in general, EMAP-Estuaries' response to the Estuarine Research Federation
panel was serious and effective. The Estuarine Research Federation panel
could be used as a successful prototype for the resource panels (one for
each component of EMAP-Estuaries) that are to be developed to incorporate
scientific input to the project.
The annual design workshops held in 1991 and 1992 were a potentially
useful device to evaluate the previous year's results and incorporate con-
structive changes into the next year's sampling scheme. However, reports
of these workshops are very brief and do not present much detail. Although
sampling design and statistical power of some analyses were thoroughly
discussed, the real ability of the indicators measured to assess ecosystem
degradation or restoration likely to be encountered was not addressed. The
composition of the workshop attendees also differed greatly between 1991
and 1992. In 1991, most participants were from the EPA and the EPAs prime
EMAP-Estuaries contractor, VERSAR, but a number of academic estuarine
scientists also attended. In 1992, only a small number of EMAP-Estuaries
and EMAP design and statistical personnel participated. Thus, the level of
scientific review and input into design and analysis of the program is
inconsistent, with respect to both depth and percentage of outside
scientists involved.
COORDINATION WITH AGENCIES
NOA/fe Status and Trends (NS&T) program is often cited as an example
of how well EMAP is cooperating with other agencies. However, the NS&T
sampling is not probability based. EMAP-Estuaries and NS&T have
cooperated to the extent of standardizing the collection of certain types of
data and may produce a handbook to facilitate sampling. NS&T and EMAP
recently opened a joint office in Maryland from which to cooperatively run the
Carolinian Province demonstration project. How well this approach works
has yet to be determined. At this date, the committee knows of no
agreements between NS&T and EMAP-Estuaries to coordinate their sampling
approach.
The EPAs National Estuarine Program is another program that should be
cooperating with EMAP. Such cooperation has been repeatedly called for
by the Estuarine Research Federation review panel. The emerging Regional
EMAP program could effectively make the link between the National
Estuarine Program and EMAP. Proposals currently under consideration from
Maryland and Delaware, and another proposal from the New York-New
-------
Estuaries 39
Jersey Harbor Estuaries Program, are examples of coordination that would
assist states and regions in monitoring the health of their own waters and
provide supporting data (grid and temporal enhancement) for EMAP-
Estuaries. These projects require some matching with local funds.
RECOMMENDATIONS
Sampling Design
1. The committee considers the probability-based sampling approach
to be satisfactory and flexible enough to meet the problem of how to sample
large estuaries, small estuaries, and rivers. However, the value of the
resulting data could be significantly enhanced by the addition of methods for
detecting step functions of change, including threshold effects. This is an
appropriate area for further research.
2. The committee is concerned that the proposed probability-based
sampling is not adequate for meaningful trend analysis. Consideration
should be given to development of an alternative regional-sampling scheme
that includes detailed site-based sampling, which would lend itself to more
effective trend-analysis procedures.
Indicators
3. A more explicit conceptual model is needed to direct indicator
research and provide integration across different estuarine provinces.
4. Fish sampling should be re-examined carefully and either expanded
significantly or dropped. In particular, the measurement of the contaminant
body burden as currently implemented is fundamentally flawed because no
account was taken of temporal (seasonal) variability, and more importantly
no account was taken of the relation between body burden and the size
(age) of the fish.
5. The evaluation of current indicators should continue. The deter-
mination of the variability of the dissolved oxygen measurements carried out
by EMAP-Estuaries is a good example of the care that must be taken in
evaluating even a relatively simple indicator.
6. Whenever possible, EMAP-Estuaries should develop remotely-
sensed indicators. Candidate measures would include surface chlorophyll
and extent of submerged aquatic vegetation.
-------
40 Review of EPA's EMAP
7. EMAP-Estuaries should develop indicators of ecosystem function,
such as productivity.
8. The benthic index should continue to be examined, particularly the
validity of the use of weighing factors in the discriminant analysis.
9. Research must continue on the best way to make measurements of
fish pathology and on the best way to separate the various causes of the
pathology.
10. A detailed analysis of the successes and failures of the various pilot
and demonstration projects should be undertaken and the best ideas should
be incorporated into the indicator plan for EMAP-Estuaries.
11. Investigator-initiated research should continue to be supported to
validate indicators in use and develop new ones.
12. EMAP-Estuaries should use the detailed historical data sets already
identified in the Virginian Province to provide temporal and spatial
information on indicators and their ability to detect change.
13. EMAP-Estuaries should give priority to determining the power of
these indices to detect the changes that are reasonably expected over the
next decade.
Data Collection, Analysis, and Management
14. EMAP-Estuaries should change to a standard relational data-base
management tool for storage and retrieval in the near future.
15. Alternative multtvariate analysis techniques should be carefully
explored in addition to the linear discriminant function that was applied to
data for the classification of estuarine habitats. EMAP might also explore
methods of statistical computing concerning visualization of multi-
dimensional data.
16. Compatible data sets and efficient interfaces must be developed
between EMAP-Estuaries and other segments of EMAP, NOAA, and other
federal agencies with large-scale monitoring information. This would allow
EMAP-Estuaries to be used as a model for development of data acquisition
and management in other EMAP resource groups.
17. A mechanism needs to be developed to ensure thoughtful, detailed
analyses and interpretation of data. Outside scientists and managers should
be involved at all stages, including review.
18. An updated plan needs to be prepared as to how the reports and
analyses will incorporate the data on stressors. There has been a change
from the original plans, and the entire EMAP now appears to be reducing its
emphasis on stressors.
-------
Estuaries 41
19. It is imperative to assess EMAP-Estuaries in detail before going on
or before adding additional provinces. This assessment should begin as
soon as the Virginian Province demonstration completes its first four-year
cycle.
Application of Data
20. EMAP-Estuaries has begun development of indicators that should
be useful to managers of estuarine resources. It now must begin to develop
a program for continual involvement of these managers in improving the
indicators and reporting and
interpreting results.
21. The scientific applications of the EMAP-Estuaries data are not yet
clear. The present-day open availability of raw EMAP-Estuaries data must be
maintained in order to foster the scientific use of the information.
22. Greater emphasis must be placed on the timely production of
annual statistical summaries and reports on demonstration projects.
Availability of an up-to-date assessment of how EMAP-Estuaries is doing and
what it is finding is important to
managers and to the scientific community.
Scientific Review
23. The quality of the scientific review panel organized by the Estuarine
Research Federation was high. EMAP-Estuaries needs to establish a regular,
working, review panel of similar quality for each component of the program.
Coordination with Agencies
24. Coordination between EMAP-Estuaries and state, regional, and
national monitoring programs should continue to be aggressively pursued.
These connections may be the most useful outcome of the program.
-------
Forest Health
Monitoring Program
BACKGROUND
The EPA has joined the U.S. Forest Service (USFS) and other govern-
ment agencies in an effort to monitor and assess the condition of the nation^
forested ecosystems and their response to natural and human-induced
stresses. This long-term and multitiered strategy for monitoring and assess-
ment includes elements for detecting, evaluating, and explaining changes
in forest condition. Previous attempts at a comprehensive forest inventory,
such as the USFS Inventory and Analysis plots, concentrated on traditional
forestry measurements.
In contrast, EMAP-Forests and the FHM (the Forest Health Monitoring;
EMAP-Forests' counterpart in the USFS) program propose to collect data on
environmental factors that influence forest growth as well as additional
response variables of the trees, such as foliar chemistry and canopy struc-
ture. If this is implemented, the resulting data sets will be valuable for
ecologists and foresters seeking to understand basic ecological patterns and
for policy makers who require information for the evaluation of future environ-
mental impacts on the nation's forests.
In addition to this high degree of coordination between EPA and the
USFS, ER% EMAP-Forests program has had major contributions from other
agencies, including state forestry agencies, the National Park Service, the
Fish and Wildlife Service, the Tennessee Valley Authority, and the Bureau of
Land Management. The collection of all these efforts is now referred to as
EMAP's Forest Health Monitoring Program.
This multiagency partnership is exactly the type of cooperation that EPA
and this committee see as vital to EMAP's national monitoring effort.
However, it should be emphasized that many of the good features of this
program derive from the previously established USFS FHM program and not
42
-------
Foresf Health Monitoring Program 43
from EMAP. The EMAP-Forests program has had substantive and intensive
scientific review. The committee endorses this type of scientific review.
Furthermore, it recommends that if a standing review committee is estab-
lished, as was suggested at the EPA meeting held September 1992 at Re-
search Triangle Park, North Carolina, appointments should be staggered so
that both old and new perspectives will be available to EPA.
In this chapter, the NRC committee has reviewed various draft docu-
ments related to the EMAP-Forests program. In addition, several committee
members met with the technical directors and other field personnel involved
in the forest-monitoring program for EMAP and received briefings at com-
mittee meetings on the status of the forest-health-monitoring program and
demonstration-project results. Following is a summary of the committee's
review of the EMAP Forest Health Monitoring Program as of August 1993.
THEORETICAL BASIS
EMAP-Forests has a number of conceptual flaws that could be remedied
given sufficient time and personnel. The major flaw is that the collection and
analysis of this data set, as well as the setting of priorities, were not guided
by any theory or model. In the absence of an accepted predictive theory or
model, the vast data sets to be collected in this program cannot be used to
predict future forest health or evaluate trends, both of which are prime
objectives of EMAP.
This ambiguity of theory/model is implied in the following assumption
taken from the draft documents of the USFS FHM program and echoed in
many EMAP-Forests documents as well: "Because forests are dominated by
long-lived species, they change gradually. Therefore, forest ecosystem pro-
cesses can only be understood by repeated, long-term observations". This
premise is simply not true. Detailed simulation models suggest that changes
can be rapid and catastrophic if thresholds are crossed and feedbacks
between system components, particularly plants and soil, are strong
(Shugart, 1984; Gatto and Rinaldi 1987; Pastor and Post, 1988). Under-
standing complex systems needs more than long-term observations. Such
systems are best understood by the interaction of theory/models and long-
term observations. Otherwise, one runs the risk of having a huge and expen-
sive data set that detected changes only after the fact, if at all, and has no
real predictive value.
Current theory and models (e.g., Shugart, 1984) indicate that because
forest biomass is high and dominated by a few species, changes are gradual
as long as those species that are dominant remain dominant. Furthermore,
feedback between species and resources stabilizes forests. However, once
-------
44 Review of EPA's EMAP
this feedback is disrupted by agents such as disease, air pollution, or some
other agent, productivity and species composition may change rapidly.
Therefore, gradual change should not be expected, and stasis followed by
rapid changes is more likely (e.g., dieback). This was confirmed recently
with detailed pollen and isotope analyses (MacDonald et a!., 1993).
Therefore, those variables that reflect or integrate feedbacks, particularly
those resources that limit production (e.g., mineralizable soil nitrogen) should
receive highest priority in designing sampling schemes. Serious attention
also needs to be paid to measurements of diversity that have not been
traditionally incorporated into standard inventories. The current plan does
not set priorities on any variables, because the data (e.g., photosynthetically
active radiation or Normal Differential Vegetation Index) apparently have
been collected independent of any model or theory (and involve more
traditional USFS measures that are related largely to timber production.
The EMAP-Forests management team and key scientists seem to
recognize the need for theories or models to help guide selection of
indicators and their interpretation. However, the development of a theoretical
basis for EMAP-Forests appears to be evolving concurrently with selection
and testing of indicators. In September 1992, the EMAP-Forests panel of this
committee discussed the theoretical basis of EMAP-Forests with several key
EPA scientists. Elements of a theoretical basis for EMAP-Forests included
hierarchy theory, sampling theory, epidemiological theory, and the stand-
development theory of Oliver and Larson (1990). However, the logical basis
by which these theories relate to one another and explain the responses of
forests to stress, explain the subsequent responses of surface waters to
changes in forests, and guide indicator development and sampling protocols
is not yet clear. Heavy reliance appears to be placed on a purely epidemio-
logical model. This may be due to the fact that the original purpose of the
USFS FHM program was to detect the effects and extent of diseases and in-
sects. Epidemiological models describe how diseases spread through pop-
ulations. Such models appear to have little utility in predicting how nutrient
cycles, nutrient losses, or biodiversity of ecosystems change in response to
stress. The committee strongly recommends that EMAP personnel continue
development of a theoretical basis for EMAP-Forests from which predictions
can be made of general types of forest response to different types of stress.
The theory should encompass at least productivity and diversity. While this
seems difficult, it is not out of the question. Elements of such a theory and
their relationship to policy are currently available in the literature (see review
by Lubchenco et al., 1991). With such a theory, EMAP will be better posi-
tioned to detect effects of stresses when they begin rather than after they
become obvious.
-------
Forest Health Monitoring Program 45
SELECTION OF INDICATORS
The issue of selection of indicators for forests has been dealt with
extensively in several documents that deal both with the philosophy under-
lying the choice of indicators and the actual selection of specific indicators.
As far as forests are concerned, these studies were summarized in the re-
ferred article by Riitters et al. (1992) dealing with selection of forest-condition
indicators for monitoring. Riitters et al. laid out some of the philosophy
behind the choice of indicators, pointing out that to meet the objectives of
EMAP a multidimensional set of indicators is needed. Measurements are
needed to characterize ecosystem structure, function, and processes and to
diagnose and interpret known cause-and-effect relationships. Indicators also
must be relevant "to environmental values that people care about." Again, it
should be emphasized that the Riitters et al. paper presents a "philosophy"
rather than a quantitative, predictive theory.
Data collected in 1990 and 1991 were evaluated to develop indicators for
use during 1992 according to the following six criteria developed by Knapp
et al. (1991) for all resource group indicators:
unambiguously interpretable;
simple quantification;
high signal-to-noise ratio;
regionally responsive;
index period stability; and
environmental impact.
Ultimately, five indicators were selected for use in detecting changes in
forest condition: landscape pattern, visual symptoms, tree foliar nutrients,
soil nutrients, and stand-growth efficiency. However, these are not mea-
surable indicators but rather classes that contain specific indicators.
Within each of these major categories, a number of specific indicators
were selected for further testing and consideration. Proposed indicators
must be critically evaluated against the criteria of Knapp et al. (1991) given
above. It appears that several of the proposed indicators fail to meet one or
more of these criteria. For example, foliar nutrients do not have a "high
signal-to-noise ratio," nor do they exhibit "index stability." Stand-growth effi-
ciency is a theoretical construct that is not "unambiguously interpretable,"
and it remains to be seen whether or not it is "regionally responsive." None-
the less, these indicators remain attractive to those who wish to monitor
forest conditions. Also, some indicators, even though they may have a
compelling theoretical basis, are extremely difficult or prohibitively expensive
-------
46 Review of EPA's EMAP
to measure. This poses a paradox that must be resolved, although the solu-
tions are not readily evident. A large part of the answer may lie in limiting the
number of indicators to be measured or in concentrating on the simpler
ones.
The EMAP-Forests group and the EMAP director have said that indicators
should provide assessment standards for three "societal values: biotic integ-
rity, productivity, and aesthetics" (material presented at Research Triangle
and at the panel's October 1992 meeting). To date, the indicators proposed
by EMAP-Forests clearly address productivity only. If indeed biotic integrity
and aesthetics are important, additional indicators relevant to assessing
status and trends with regard to these values must be incorporated into the
program.
Furthermore, the specific indicators have been different in each region,
having been adapted to the region and to the specific circumstances of the
type of study conducted. There is apparently no single set of standard
indicators that is to be sampled nationwide. Comments in field-study reports
(e.g., the Western Pilot Study; EPA, 1992b) suggest that a single or core set
of standard indicators is essential to the future success of EMAP-Forests.
Many parameters measured in the field are clearly important facets of
tree and stand growth, and the process by which the indicators have been
tested to date met the criteria outlined above. One can argue at length about
precisely what indicators are the most important. Nonetheless, the NRC
committee believes that the five classes of indicators listed above are sound
choices from which to proceed. However, it would be best if the final choice
of indicators proceeded logically from a theoretical basis of EMAP- Forests.
Otherwise the theory behind EMAP runs the risk of being ad hoc justification
for the indicators already chosen.
EMAP-Forests should choose a set of indicators as soon as possible,
clearly establish sampling methods, convey these to field crews, and
develop the process for interpreting the results derived from field studies.
The committee strongly recommends that priority be given to evaluating
measurements that integrate limiting factors over the growing season and
that can be performed quickly using standardized procedures. The Western
Pilot Study (EPA, 1992b) showed that trying to make many measurements
on each plot is logistically infeasible. Therefore, the list of indicators needs
to be narrowed to a set that can be measured with at most a few hours of
time spent at each plot but at the same time measure important processes
that indicate the overall status of the forest. The current management of
EMAP-Forests seems well aware of this need and has set high priority on
finding a successful solution.
-------
Forest Health Monitoring Program 47
SAMPLING DESIGN
The EMAP-Forests statistical design dovetails with the USFS Forest
Inventory and Analysis (FIA) program and produces a probability sample of
field plots in each region. The plots are selected by laying FIA photo point
grids over the EMAP hexagon grid and selecting the photo point closest to
the center of the landscape characterization hexagon. When the FIA point
already exists at the selected photo point, the study plot is deliberately offset
to avoid disturbing the FIA plot. At each location, a 1-hectare circle repre-
sents the experimental unit of interest. Within each such unit, a cluster of 4
fixed-area subplots (24-ft. radius, 1/24 acre) is designated. The^subplot
centers are 120-ft. apart; destructive and extractive sampling will be limited
to a 36-ft. circular band surrounding each subplot to ensure that, as far as
possible, all will be considering the same experimental units. This approach
appears to be sound and well thought out and there is no reason for EMAP-
Forests to reinvent the wheel regarding a sample design when the USFS
FHM program, for example, is so readily available, accessible, and appro-
priate for the kinds of measurements EMAP-Forests will be considering.
The choice of sampling design is constrained by the basic premise
around which EMAP is built, namely the unbiased estimation of current
ecological status using cumulative distribution functions. If this constraint is
maintained, it forces the program to use a grid (in this case the hexagonal
grid) in which each plot is revisited once every four years, because repeated
sampling on a grid is the only practical way to construct cumulative distri-
bution functions. This hexagonal grid system has several implications for the
ability of EMAP-Forests to assess current status and trends, as discussed
below. Furthermore, this has led to the rejection of alternative sampling
schemes (the use of specific selected sites with long-term records such as
Long-Term Ecological Research sites, watershed approaches, and stratified
random sampling such as the USFS FIA plots), because they did not meet
these programmatic criteria, regardless of whether or not they were viable
alternatives. The committee is concerned that alternative sampling schemes
(as mentioned above) not be overlooked as appropriate ways to test future
hypotheses simply because of the constraints of the EMAP-imposed grid
system.
The proposed statistical procedures for indicator development and
indicator linkages include a variety of standard methods. Linear and approxi-
mately linear relationships will be evaluated using correlations, multivariate
regression, analysis of variance and covariance, principal components analy-
sis, factor analysis, or canonical correlation. Monotonic relationships will be
evaluated using nonparametric correlations; non- parametric regression; and
-------
48 Review of EPA's EMAP
nonparametric analogues to analysis of variance, such as Kruskal-Wallis
tests. Nonlinear relationships will also be investigated. Statistical findings
will need to be closely scrutinized by the indicator research teams for bio-
logical meaning.
At this point, this is merely a laundry list of proposed statistical tools.
Final evaluation of the strength and efficacy of these techniques should be
based on the completed analyses and interpretation of the demonstration
projects. Some of these techniques have been tested and the results re-
ported in the Forest Health Monitoring Plot Design and Logistics Study
(EPA, 1991 b), a well-executed piece of work that includes various elements
of plot design and tests of a number of indicators, including soils, foliar
chemistry, vegetation structure, PAR, and visual symptoms.
An important fact emerging from the 1990 New England study (Brooks
et al., 1990), which was a full test of FHM program field technique, was that
the sample derived from the FHM program plots, which were located and
analyzed using EMAP protocol, corresponded very closely to a previous
inventory of New England forests. Thus, the New England study can be
viewed as a partial test of EMAP technology in the forest environment. EMAP
appeared to pass this first test in that it produced statistical description of
New England forests similar to that earlier produced by USFS inventory
techniques.
Although EMAP-Forests appears able to assess current status of forests
in a limited sense, its ability to detect trends is still open and subject to
question. The problem is that the so-called interpenetrating design of EMAP
results in each plot being revisited once every four years. Yet, there are
natural cycles in forests, such as mast and cone production and insect
outbreaks, as well as important forcing functions, such as El Nino events,
that also occur at approximately the same frequency.
Revisiting a site once every four years prevents EMAP from making site-
specific estimates of changes in these frequencies. If the sampling scheme
rotates annually among four plots for each grid point, then it will detect
changes in frequencies only if the four plots sample a homogenous popula-
tion. Given the known spatial variation in forests, this is unlikely. This
insensitivity to detecting changes in frequency of environmental cycles may
be unfortunate in the case of known periodic events such as El Nino.
(Swetnam and Betancourt, 1990; Cohen and Pastor, 1991). Surely, EMAP-
Forests would like to be able to detect the responses of forests to such
changes. However, it appears unlikely that they will be able to do so.
The committee recommends that the current design of four-year plot ro-
tations be replaced or augmented by a design in which at least some plots
-------
Forest Health Monitoring Program 49
are revisited every year. This is also the conclusion of the EMAP statistical
design team (Urquhart et al., 1993). The EMAP-Forest team is considering
a scheme in which 10 percent of plots are revisited every year, while the four-
year rotation is maintained among all other plots. It is not yet certain whether
10 percent is adequate, given temporal and spatial variability of properties.
The number of plots revisited yearly may vary depending on the property
being measured, its variance, costs of sampling, etc. This needs further
study. Perhaps a sampling scheme that permits some plots to be revisited
every year to maximize temporal coverage with some plots sampled on a
rotating basis to maximize spatial coverage could be employed.
One element missing from the Forest Health Monitoring Plot Design and
Logistics Study is an estimate of the actual cost of the work in dollar
amounts rather than in statistical terms. Such an analysis would be most
useful, as it would yield at least a clue as to the cost of the type of sampling
envisioned for EMAP in terrestrial communities. Such data are essential
because the high dollar cost of EMAP is one of its greatest sources of
vulnerability.
DATA MANAGEMENT
information Management (IM) for EMAP-Forests has developed a set of
short-term solutions that seem inadequate for long-term information-
management needs. The work that the EMAP-Forests Information
Management team has done needs to be integrated into a more
comprehensive system that addresses the issues raised in the overview of
the report.
The goals of the Information Management program are guided by three
reasonable principles:
1. ensure that the data in the system are of the highest quality
possible;
2. ensure that the EMAP FHM program scientists have access to the
data as quickly as possible; and
3. make the data available to users both within the proj ect and outside
the FHM program.
A set of 8 objectives have been laid out for the Information Management
program. These set the stage for the development of a comprehensive infor-
mation management system, but they are vague in terms of implementation.
-------
50 Review of EPA's EMAP
Not enough thought has been given to the best way to organize the data-
base structures. EMAP-Forests will use the Statistical Analysis System as its
data-base system and statistical programs package. The data collected in
the field study (except continuous photosynthetically active radiation
readings) will be stored in the system data sets on a mainframe computer in
Las Vegas. The Statistical Analysis System is not a data-management and
information system. Quality assurance/quality control is well taken care of,
starting with program-supported data entry in the field. Data are checked
and checked again by the time they end up in Statistical Analysis System
data bases. However, system files are an inappropriate format for long-term
management.
The Information Management team has identified four distinct data user
groups with significantly different access privileges to the data bases based
on immediacy of need. They have identified requirements for data access
(e.g., timing) ranging from real time, one month from collection,
approximately 6 months from collection, to one year from collection. Data
confidentiality continues to be a real issue and has not been resolved by
EPA. For example, private landowners do not want data collected on their
land to be used to prosecute any environmental violations and related
problems) continues to be a real issue and has not been resolved by EPA.
Guidelines have been prepared that include various mechanisms of getting
around this problem by not providing exact plot locations in certain instances
(i.e., "fuzzing" Geographical Information System representations), but final
guidelines have not been accepted. The value of national standards has
been recognized by EMAP-FHM. Categories for which standards are being
discussed by an interagency workgroup include the following: codes,
computational algorithms, portable data recorders, portable data recorders
software, measurement units, and word processing software. The outcome
from this interagency workgroup is not known at this time.
If the EMAP-FHM program is to function efficiently interagency computer
links are necessary. This will require establishing computer finks among all
participating agencies. EPA proposes that the link begin with the EPA/USFS
connection and branch out from there. The interagency link will provide
e-mail, file transfer, and data base access to all participants across the
EMAP-FHM program. Additional links with Bitnet, Internet, and LJERnet
would allow easy access to university cooperators and are listed by EPA as
needing to be explored. Such linkages should be encouraged.
LANDSCAPE CHARACTERIZATION
Although the EMAP-Forests landscape program has clear landscape
ecology overtones, it is still unclear to what degree theoretical as opposed
-------
Forest Health Monitoring Program 51
to logistical considerations drive the program. The connection to EMAPs
overall landscape characterization effort is even less clear. Documents on
EMAP's landscape Characterization plan have not yet been released by EPA
for review by this committee. Any information on this program component
has been verbally presented to date. It would appear that the overall
program is largely geographic in nature and will be carried out at a small
map scale. Conversely, the EMAP-FHM program appears to have a more
ecological and predictive character and is to be carried out at a large map
scale. These differences obviously need to be reconciled if information flow
between resource groups is to occur.
This is also an excellent example of how important it is for the evolution
of indicators and the establishment of monitoring to go hand in hand.
EMAP-Forests hopes to use landscape indicators to address such questions
as "What is the association between forest health and monoculture
management practices?" and "What is the association between forest health
and forest fragmentation?" These are supposed cause-effect relationships
that remain to be substantiated, and appropriate indicators to address such
questions have yet to be developed. Landscape characterization should be
clearly subdivided into landscape ecology and landscape-sampling sectors.
It appears that this is an area where monitoring could easily outstrip science;
the committee recommends that great care be taken in the development of
this potentially powerful but poorly understood area.
STAFFING AND PERSONNEL TRAINING
The best information about staffing and training is contained in EPA
reports such as the Forest Health Monitoring and Plot Design (Brooks et al.,
1990) and the 1991 Forest Health Monitoring Western Pilot Operations
Report (EPA, 1992b). Both of these reports contain detailed descriptions of
the backgrounds of the field crew members, their agency affiliations, and the
training they were given before field work was begun. Most personnel are
from the USFS. A soil scientist from the Soil Conservation Service was
detailed to each field team through an interagency agreement.
It is clear from studying these reports that careful thought was given to
the structuring and training of the field crews. Both reports describe a
thorough training program for the field crews and both contain extensive
feedback from the crews. In both reports, this information is synthesized into
a set of recommendations for modification of the protocols used in the field.
Because further summaries of field studies are not available, it is not
possible to know the extent to which these recommendations have been
-------
52 Review of EPA's EMAP
followed, it is reasonable to assume, however, given the thoroughness with
which personnel were briefed before they were sent in the field, that these
suggestions have been used. The conclusion that the committee draws from
the reports is that the EMAP-Forests program is using competent,
well-trained professionals to gather data in the field. Again, however, there
is no financial analysis of the dollar costs of bringing together and training
the field teams or of actually gathering the field data. It should be rather easy
to determine such costs. Their presentation and analysis is an important
facet of the EMAP program about which the committee has no data. One
can speculate that the cost of the field-sampling component of EMAP is very
high and that economies in this area may be necessary if EMAP is ever to be
fully implemented.
RESULTS OF DEMONSTRATION PROJECTS
The New England Forest Health Monitoring pilot project of 1990 was to
establish baseline conditions for assessing attributes of forest health. Field
visits were made to 263 sample plots across the six New England states, and
measurements were taken on 206 plots determined to be forested. The pilot
project showed several important things:
1. Sampling the FIA plots with the less intensive EMAP grid did not
distort means and variances of most forest attributes. This shows that the
controversial EMAP grid works in some way at least for New England forests.
2. Spatial autocorrelation did not seem to be a problem in evaluation
of forest health. However, EMAP-Forests New England project did not
consider changes in spatial patterns to be an indicator. Mean values of
environmental variables may not be as sensitive indicators of changes in
forest health as changes in their spatial distribution. More work needs to be
done along these lines, perhaps in conjunction with the landscape-
characterization port of EMAP.
3. Data were collected on many important variables, resulting in a
comprehensive data set that probably has no parallel anywhere in the forest-
ecology literature. This data set needs further analysis, especially with
regard to which variables are correlated and which are expected to change
with respect to which stresses. Also, the project needs to be repeated in a
different type of forest (Pacific Northwest, etc.) to determine if the EMAP grid
is sensitive to terrain, stand type, etc.
4. The data gathered showed that any "results indicating a health
issue" could be related to a known natural rather than anthropogenic cause.
-------
Forest Health Monitoring Program 53
The data are clearly useful, largely as baseline data, and as the report
indicates, "the full value of these data . . . will be realized with plot
remeasurements in succeeding years" (Brooks et a!., 1990). Thus, it can be
argued that the New England study is the first test of EMAP technology and
is valuable in this light alone.
The 1991 FHM Western Pilot operations study, located in California and
Colorado, was more modest in scope than the New England study in that
only seven sites were actually sampled in each state. This pilot, which was
carried out in August 1991, was an intensive analysis of staffing and training
methodology, utility of various indicators, and handling of data. In that
regard, it is an excellent synopsis of some of the problems that will be
encountered as EMAP-FHM goes nationwide.
The indicators used in the western pilot study were selected based on
those used in the eastern surveys. At a series of meetings, others were
added to this list. In order to use a particular indicator, a lead person was
identified to devise field methodology, provide crew training and debriefing,
and write a summary of the usefulness of the indicator for study objectives.
A final list of 15 indicators was chosen: mensuration (data such as diameter
at breast height (dbh), used for trees, height, etc.), crown assessments,
branch evaluation, bioindicator plants, root sampling, soils, foliar nutrients,
tree cores, vertical vegetation, photosynthetically active radiation, height,
branch extraction, lichens and mosses, soil microbiology, and understory
vegetation.
Several issues were identified in the western pilot study that provide clues
as to problems that are yet to be resolved. No structure or criteria for
selecting indicators have been accepted by both the USFS and EPA, a
conclusion that supports the committees previously expressed concerns
that the theoretical foundation of EMAP-Forests is vague. Although the
indicators used in the western pilot study proved appropriate, some of them
may not be useful for full implementation. It also proved difficult to find a
lead person to provide the technical background necessary to permitthe use
of a given indicator. All of this suggests that the issues of choice of
indicators and identification of persons who can provide the technical
justification for the indicator, training in its analysis in the field, and
interpretation of field data are crucial issues yet to be resolved. Although the
expertise the field crew used in the west was appropriate for most tasks and
indicators, a number of issues arose with respect to training and staffing.
None of these appears insurmountable. It was interesting to note that the
field crew itself feit that it needed more information about USFS-FHM and
EMAP-FHM. In other words, they did not seem to understand fully why they
were doing what they were doing in the way they were doing it. This lack of
understanding ranged from programmatic philosophy to field protocols.
-------
54 Review of EPA's EMAP
BUDGET AND COSTS
The committee has serious concerns about the costs of the proposed
EMAP-Forests program. It is clear that the field sampling described in the
pilot studies will be extremely expensive.
The committee believes that a serious effort must be made to reduce the
number of indicators and streamline the field-data collecting process so that
more than one plot can be done in a day. The committee believes that the
costs of the field work carried out so far may be prohibitively expensive when
expanded to a national basis.
The EMAP-Forests staff estimates that the annual cost to EPA of full
implementation of the proposed forest program alone is $12.5 million. This
assumes that the USFS will provide an additional $30 million annually when
the program is fully implemented. The USFS, in turn, assumes either that
states will provide some of these funds or that they can be made available by
redirecting funds currently appropriated to the state and private forestry
program. The states' responses to these alternatives are uncertain, but it is
reasonable to suppose that they will not be favorable. It is also useful to put
this in some overall context. The estimated cost of $42.5 million for EMAP-
FHM is more than double what the National Science Foundation and the U.S.
Department of Agriculture Mel ntire-Stennis fund spends annually on research
to understand and predict forest productivity and functioning. It is not clear
that, by comparison, EMAP-Forests will provide better tools or data with
which to make decisions.
RECOMMENDATIONS
The committee endorses the intensive use of scientific review in the
EMAP-Forests program and recommends that a standing review committee
be created with staggered appointments to provide new perspectives while
maintaining continuity.
Theoretical Basis
The EMAP Forest Monitoring Program does not have an adequate
theoretical basis or model. Therefore, EMAP personnel should continue
development of a theoretical basis for EMAP-Forests from which predictions
can be made of forest responses to different types of stress.
-------
Foresf Health Monitoring Program 55
Indicators
The final choice of indicators for EMAP-Forests should proceed logically
from a theoretical basis. A single, small set of standard indicators should be
developed and applied nationwide. EMAP Forests should select indicators
that adequately address societal values other than productivity. The present
list of indicators needs to be narrowed to a set that can be measured in a few
hours time at each plot. These indicators should also directly assess impor-
tant processes that indicate the overall status of the forest. In turn, these
indicators, and the rationale for them, should be clearly communicated to
field personnel.
Sampling
1. Variables that reflect or integrate feedbacks, particularly those
resources that limit production, should receive highest priority in designing
sampling schemes.
2. The current design of four-year plot rotations should be replaced or
augmented by a design in which some plots are revisited every year.
3. The EMAP Landscape Characterization Program should be clearly
subdivided into landscape ecology and landscape sampling sectors.
Cost Estimate
An estimate of the cost of gathering field data must be made as soon as
possible. Such an analysis would be useful as it would yield a clue as to the
cost of the type of sampling envisioned for EMAP in terrestrial communities.
The dollar cost of EMAP is one of its greatest sources of vulnerability.
Information Management
EMAP-Forests should develop a comprehensive information manage-
ment plan that outlines user requirements examines, long-term implemen-
tation (of hardware and software), and fits in with the overall plan for the
information management system.
Finally, for all the above reasons, the EMAP-Forests program should not
be fully implemented until the results of demonstration projects have been
thoroughly evaluated and a realistic estimate of its cost to EPA and other
agencies is available.
-------
References
Becker, R. A., W S. Cleveland and A. R. Wilks. 1987. Dynamic graphics for
data analysis. Statistical Science 2(4):355-395. See also 5 "Comment"
articles plus rejoinder following in the same issue.
Brooks, R. T., D. Dickson, W. Burkman, I. Millers, E. Cooler, and L Smith.
1990. Forest Health Monitoring: A Partnership Between U.S. Environ-
mental Protection Agency, U.S. Forest Service, and National Association
of State Foresters. Annual Report-Forest Health Monitoring, New
England. Washington, D.C.rU.S. Environmental Protection Agency.
Cleveland, W. S., and M. E. McGill. 1988. Dynamic Graphics for Statistics.
Belmont, California:Wadsworth and Brooks/Cole.
Cressie, N. 1991. Statistics for Spatial Data. New York, New York:John
Wiley and Sons, Inc. global warming. Ecology 72:1161 -1165.
Cohen, Y, and J. Pastor. 1991. The responses of a forest model to serial
correlations of global warming. Ecology 72:1161 -1165.
Delaware Department of Natural Resources and Environmental Control and
Maryland Department of the Environment. 1992. Proposal, Use of EMAP
Protocols for the Assessment of Inland Bays and Tidal Rivers of the
Virginian Province.
EPA. 1990. Near Coastal Program Plan for 1990: ESTUARIES. EPA 600/4-
90/033. Narragansett, Rhode lsland:U.S. Environmental Protection
Agency, Office of Research and Development.
56
-------
References 57
EPA. 1991 a. Monitoring Guidance for the National Estuary Program. Interim
Final. EPA 503/8-91-002. Washington, D.C.:U.S. Environmental Protec-
tion Agency, Office of Water.
EPA. 1991b. Forest Health Monitoring Plot Design and Logistics Study. K.
Riitters, M. Papp, D. Cassell, and J. Hazard, eds. EPA/600/XX-XXX/XXX.
U.S. Environmental Protection Office, Office of Research and Develop-
ment, Research Triangle Park, NC.
EPA. 1992a. Information Management Strategic Plan: 1993-1997, Version
1.6, draft. EMAP Information Management Coordinator, Washington,
D.C.:U.S. Environmental Protection Agency.
EPA. 1992b. FY91 Forest Health Monitoring Western Pilot Operations
Report. EPA 6000/X-92/009. Washington, D.C.:U.S. Environmental
Protection Agency, Office of Research and Development.
EPA. 1993a. EMAP Response to National Research Council Interim Report.
Washington, D.C.:U.S. Environmental Protection Agency.
EPA. 1993b. Environmental Monitoring and Assessment Program Guide,
EPA/600/XX-93/XXX. Research Triangle Park, NCrU.S. Environmental
Protection Agency, Atmospheric Research and Exposure Assessment
Laboratory.
EPA. 1993c. Summary of ORD Workshop to Address and Resolve Issues
and Concerns Regarding EMAP, Research Triangle Park, North Carolina,
September 1992. U.S. Environmental Protection Agency:Washington,
D.C.
Estuarine Research Federation. 1992. Review of the U.S. Environmental
Protection Agencyfe Environmental Monitoring and Assessment Pro-
gram-Estuaries. Gloucester Point, Virginia: Estuarine Research Federa-
tion.
Feinberg, S. E. 1980. The Analysis of Cross Classified Data. Cambridge,
Massachusetts:Massachusetts Institute of Technology Press.
Frithsen, J. B., M. C. Fabrizio, J. Gerritsen, A. F. Holland, G. E. Saul, and
S. B. Weisberg. 1991. Example Environmental Assessment Report for
Estuaries. EPA/600/4-91/026. Washington, D.C.:U.S. Environmental
Protection Agency, Office of Research and Development.
-------
58 Review of EPA's EMAP
Gilbert, R., and J. Simpson. 1985. Kriging for estimating spatial patterns of
contaminants: Potential and problems. Environmental Monitoring and
Assessment 5:113-135.
Gatto, M., and S. Rinaldi. 1987. Some models of catastrophic behavior in
exploited forests. Vegetation 69:213-222.
Knapp, C. M., D. R. Marmorek, J. P. Baker, K. W. Thornton, J. M. Klopatek,
and D. P. Charles. 1991. The indicator development strategy for the
Environmental Monitoring and Assessment Program. Corvallis, Oregon:
U.S. Environmental Protection Agency, Environmental Research Labora-
tory.
Lubchenco, J. 1991. The sustainable biosphere initiative: An ecological
research agenda. Ecology 72:371 -412.
MacDonald, G. M., T. W D. Edwards, K. A. Moser, R. Pienitz, and J. P. Smol.
1993. Rapid response of treeline vegetation and lakes to past climate
warming. Nature 361: 243-246.
Menzie Cura and Associates, Inc. 1990. Evaluation of Historical Data Sets
for EMAP. Prepared for U.S. Environmental Protection Agency, Environ-
mental Research laboratory, Narragansett, Rhode Island. Westford,
Massachusetts:Menzie Cura and Associates, Inc.
NRC. 1992. Environmental Monitoring and Assessment Program (EMAP)
Interim Report. June 1992. Water Science and Technology Board and
Board on Environmental Science and Toxicology, National Research
Council. Washington, D.C.:National Academy Press.
Oliver, C. D., and B. C. Larson. 1990. Forest Stand Dynamics. New York:
McGraw Hill.
Omernik, J.M. 1987. Aquatic Ecoregions of the Conterminous United States.
Annals Association of American Geographers 77:118-125.
Pastor, J.,and W M. Post. 1988. Response of northern forests to CO-
induced climatic change. Nature 334:55-58.
Riitters, K. H., B. E. Law, R. C. Kucera, A. L Gallant, R. L Develice, and C. J.
Palmer. 1992. A selection of forest condition indicators for monitoring.
Environmental Monitoring and Assessment 20(1 ):21-33.
-------
References 59
Risser, P. 1992. Personal communication from Paul Risser, President, Miami
University, to Erich Bretthauer, Assistant Administrator for Research and
Development, U.S. Environmental Protection Agency, October 6,1992.
Shugart, H. H. 1984. A Theory of Forest Dynamics. New York, New York:
Springer-Veriag.
Stanley, D. W 1993. Long-term trends in Pamiico river estuary nutrients,
chlorophyll, dissolved oxygen, and watershed nutrient production. Water
Resources Research 29:2651-2662.
Swetnam, I W, and J. L Betancourt. 1990. Fire—Southern oscillation
relations in the southwestern United States. Science 249:1017-1020.
Urquhart, N. S., W. S. Overton, and D. S. Birkes. 1993. Comparing Sampling
Designs for Monitoring Ecological Status and Trends: Impact of
Temporal Patterns. Pp. 71 -85 in V. Barnett and K. F. Turkman, eds.
Statistics for the Environment. London: John Wiley & Sons, Ltd.
Weisberg, S. B., J. B. Frfthsen, A. F. Holland, J. F. Paul, K. J. Scott, J. K.
Summers, H. T. Wilson, R. Valente, D. G. Heimbuch.J. Gerritsen, S. C.
Schimmel and R. W. Latimer. 1992. EMAP-Estuaries Virginian Province
1990 Demonstration Project Report. EPA600/R-92/100. Narragansett,
Rhode IslandrU.S. Environmental Protection Agency, Environmental
Research Laboratory
-------
Appendix A
Review of EPA's Environmental
Monitoring and Assessment Program
Interim Report
Issued: June 1992
SUMMARY
The Environmental Monitoring and Assessment Program (EMAP) of the
U.S. Environmental Protection Agency (EPA) is an ambitious program
designed to assess the nationwide distribution of ecological resources in the
United States and to assess trends in their condition. A unique aspect of the
program is its reliance on probability-based selection of sampling locations
for both of those major goals.
In response to a request from EPA, the National Research Council (NRC)
established the Committee to Review EPAs Environmental Monitoring and
Assessment Program. Both EPA and the NRC committee recognize that
EMAP is an evolving program, and that an interim report has the potential to
guide that evolution, even though firm conclusions and recommendations
are difficult at such an early stage. Therefore, the committee has used the
opportunity in this report to raise substantive questions about the design and
implementation of EMAP. The committee hopes that the questions can be
used to improve the program as it evolves. These questions should also be
used as criteria for evaluating the results of the demonstration (pilot) projects
that EMAP is currently performing. The committee has based this report on
documents received by January 10,1992, most of which were still drafts (see
Appendix A), and on briefings provided by EPA officials on or before January
21,1992.
EMAP is a large and controversial program whose purpose is laudable
and endorsed by the committee. The statistically based sampling grid
covering the entire continental United States—the basis of EMAP—is novel
and is the cause of much of the controversy surrounding EMAP. Initiation
and development of EMAP is a positive indication of an evolving recognition
by EPA that long-term ecological monitoring and research are essential and
that EPAs statutory responsibilities are not limited to human health.
60
-------
Appendix A 61
The committee concludes that probability-based sampling on a grid is an
appropriate method of delineating the extent and providing a periodically
updated survey of the nation's ecological resources. However, the
committee has many questions concerning the ability of EMAP to achieve its
other major goals, i.e., the assessment of status, the detection and
characterization of changes or trends in the status of the ecological
resources of the country, and the establishment of associations between
ecological condition and human-induced stresses. The development of
ecological indicators is a critical part of EMAP and deserves a very high
priority. The committee is concerned that EMAP might be emphasizing
probability-based spatial sampling of ecological resources at the expense of
probability-based sampling of temporal variation in indicators of ecological
status, i.e., trends. In addition, the committee has questions about EMAP's
financial and human resources, design, data management, documentation,
organization, and the involvement of the broader scientific community in all
aspects of EMAP, including its leadership.
New and complex programs are sometimes misunderstood or under-
appreciated— perhaps even canceled—during their development stages, but
others turn out to be large wastes of resources. The committee concludes
that it is too soon to decide whether EMAP is likely to be successful or not,
although the potential value of a program that could achieve EMAP's goals
is obviously large. Because of the many substantive questions concerning
the design and implementation of EMAP, the committee recommends that
the program not be fully implemented until the results of the demonstration
projects and other reviews can be evaluated and incorporated into the plans.
This recommendation should not be interpreted as a veiled condemnation
of EMAP. The committee believes that a condemnation would be premature
at this stage, as would a blanket approval. No large, complex project-
including EMAP—should be rushed into implementation before it has been
thoroughly tested and evaluated.
INTRODUCTION
In 1988, the Science Advisory Board (SAB) of the U.S. Environmental
Protection Agency (EPA) recommended that EPA "undertake research on
techniques that can be used to help anticipate environmental problems" and
that a staff office "be created within EPA for the purpose of evaluating
environmental trends and assessing other predictors of potential environ-
mental problems before they become acute" (EPA Science Advisory Board,
1988a). EPA subsequently established the Environmental Monitoring and
-------
62 Review of EPA's EMAP
Assessment Program (EMAP) "to monitor ecological status and trends, as
well as to develop innovative methods for anticipating emerging environ-
mental problems before they reach crisis proportions" (EPA, 1991 a). EMAP
is an ambitious program; understanding the complexities and nuances of its
goals and design is a challenge. EMAP's goals (EPA, 1991 a) are to
"• Estimate the current status, extent, changes, and trends in indicators
of the condition of the nation's ecological resources on a regional basis with
known confidence.
"• Monitor indicators of pollutant exposure and habitat condition and
seek associations between human-induced stresses and ecological
condition.
"• Provide periodic statistical summaries and interpretive reports on
ecological status and trends to resource managers and the public."
In response to a request from EPA, the National Research Council's
(NRC) Board on Environmental Studies and Toxicology and Water Science
and Technology Board appointed the Committee to Review the EPA£
Environmental Monitoring and Assessment Program. The committee first
met in March 1991 and has met four times since then. The committee's
charge is to consider ways to increase EMAPfc effectiveness in monitoring
conditions and trends in six representative types of ecosystems (later
increased to seven): near-coastal, arid lands, inland surface waters, wet-
lands, agroecosystems, Great Lakes, and forests. The committee will also
review the overall design objectives of the program; the strategies for
developing indicators of ecological condition, pollutant exposure, stress, and
response; methods of data collection; data analysis and interpretation, and
communication plans (statement of task, November, 1990).
The committee is scheduled to provide a review of EMAP's near-coastal,
forest monitoring, surface-water monitoring, and indicator development plans
by March 1993, providing that the products in question are available in time.
A final, comprehensive report is scheduled for December 1993. The present
interim report was tentatively scheduled "if the committee had been ade-
quately briefed" (contract with EPA, 1990). Despite the enormous complexity
of the program, the committee decided to provide the present interim report
as a summary of its evaluation to date and as early advice. The committee
has reached few conclusions at this early stage, but many substantive
questions have arisen. Those questions need to be answered conclusively
before the committee or EPA can proceed with greater confidence to eval-
uate EMAP's potential usefulness. In this report, the committee focuses on
an overall review of the general concept and implementation of a national
-------
Appendix A 63
environmental monitoring and assessment program. Analysis of various
details of EMAP will be provided, as appropriate, in later reports.
EMAP is a much-reviewed program. In addition to the committee's
review, EPAs SAB and several professional societies are reviewing selected
aspects of the program. This committee has reviewed more than 70 EMAP
documents, including earlier reviews of EMAP and EPAs responses to them
(listed in Appendix A). Most of the documents the committee reviewed were
not yet in final form. The committee has been briefed by several EPA
officials; committee members and staff have attended meetings of the SAB,
the Association of Ecosystem Research Centers, the Ecological Society of
America, the American Institute for Biological Sciences, and the American
Statistical Association at which EMAP was discussed; and the committee has
provided two sets of written questions to EPA and has considered the written
answers to them. In addition, the committee invited representatives of eight
federal agencies to its October 1991 meeting to discuss coordination of their
programs with EMAP; the agencies are the Forest Service, Bureau of Land
Management, Army Corps of Engineers, Soil Conservation Service, National
Oceanic and Atmospheric Administration, National Aeronautics and Space
Administration, Fish and Wildlife Service, and Geological Survey.
PURPOSE AND GOALS
EPA is developing EMAP to provide credible scientific information on
emerging environmental problems before they become critical and "to
confirm that the nation's efforts to protect the environment are producing the
expected results in maintaining and improving environmental quality"
(Overton et al., 1991). The establishment of EMAP is "consistent with recent
directives within EPA aimed al managing for results" (EPA, 1990a; emphasis
in original). In recommending that EPA establish such a program, EPAs
SAB (1988b) recommended that EPA "conduct a regular and systematic
monitoring assessment of the status of the American environment, applying
this knowledge to determine the status of representative ecological systems,
as well as reaching such conclusions on regional and local scales." The SAB
recommended that the "mapping and inventory projects designed to provide
[a firm foundation for extrapolation] should include ecosystem status
indicators that can be correlated with ecosystem function and value." The
SAB also pointed out that, given the regulatory mission of EPA, it was
important to know the relationships between environmental stresses and the
effects observed.
-------
64 Review of EPA's EMAP
The purpose of EMAP, described above, is translated into specific goals
through its program design. The goals have been set forth in various EMAP
documents and are listed on page 2 of this report. In its analysis of them, the
committee recognizes four separate scientific objectives of EMAP in the first
two goals (p. 2), which are important to distinguish. The first is a mapping
of the nature and extent of the country's ecological resources, the second is
the assessment of the condition (status) of those resources, and the third is
the detection of changes or trends in the condition of those resources. The
fourth is seeking "associations between human-induced stresses and
ecological condition" (EPA, 1991 a) or, more strongly, determination of "the
likely causes of adverse effects" (EPA, 1991 b).
No other national program is attempting to assess the status and trends
of the full range of the nation's ecological resources in a statistically rigorous
fashion. The committee views this national overview as the main potential
contribution of EMAP.
PROGRAM DESIGN
Description
The EMAP program design is first and foremost a probability-based
sampling scheme. EPA% commitment to such a scheme is based on its
conviction that probability-based sampling is required for a nationwide,
unbiased assessment of the distribution of ecological resources and trends
in their condition. That is what makes EMAP unique and that is what raises
most of the questions voiced by its reviewers and by participants in every
briefing or panel discussion of EMAP. The sampling approach includes four
levels or "tiers."
Tier 1. EMAP's basic Tier 1 sampling design to map the distribution and
extent of the nation's ecological resources includes 12,600 grid points evenly
spaced across the United States. Landscape descriptions of 40-km2
hexagons centered on each grid point provide the basic Tier 1 data set, and
these hexagon ("40-hex") descriptions "constitute a probability area sample
of the United States. From these descriptions will be generated estimates of
the areal extent of all ecological and landscape classifications and regional
estimates of numbers of entities of all discrete ecological objects of interest,
such as lakes, stream reaches, or prairie potholes" (Overton et al., 1991).
-------
Appendix A 65
Tier 2. Tier 2 is the basis for reporting on regional status and trends in
ecological resources. Tier 2 sampling locations will be identified from Tier 1
grid points and selected "explicitly for field visitation in a manner to best
respond to questions regarding status and trends of the specific resources"
(Overton et al., 1991). The sample sites will be probabilistically selected; if
a resource is rare, the grid will be enhanced "by one of several available
enhancement factors, solely for this resource, and solely in the necessary
area to cover this resource. The enhanced grid will also be regular and
triangular (Overton et al., 1991). On the basis of information on the spatial
distribution of resources, including that developed by Tier 1 sampling, the
designers of the plan for each resource category are free to stratify and
modify the Tier 2 design as necessary to provide sufficient information on
important resource classes; the only restriction is that the selection of
sampling sites remain probabilistic and systematic.
For reasons of cost, approximately one-fourth of the Tier 2 sites (spatially
distributed over each resource type) will be sampled annually during a period
when the effects of human activities on each resource are expected to be
most easily detected. According to the "interpenetrating" sampling design,
all Tier 2 sampling sites will be visited in a 4-year cycle. Each resource will
be sampled in Tier 2 independently of other resources (Overton et al., 1991).
indicators. EMAP also has an indicator-development strategy (EPA,
1991c) that provides a framework for deciding what should be measured.
The development of these indicators is clearly a crucial part of EMAP—as
important as the development of a sampling scheme. EMAP will produce
four types of indicators (EPA, 1991 c):
"1. Response indicators represent characteristics of the environment
measured to provide evidence of the biological condition of a resource at the
organism, population, community, ecosystem, or landscape level of orga-
ganization.
"2. Exposure indicators provide evidence of the occurrence or magni-
tude of contact of an ecological resource with a physical, chemical, or bio-
logical stressor.
"3. Habitat indicators are physical, chemical, or biological attributes
measured to characterize conditions necessary to support an organism,
population, community, or ecosystem (e.g., availability of snags; substrate
of stream bottom; vegetation type, extent, and spatial pattern).
"4. Stressor indicators are natural processes, environmental hazards,
or management actions that effect changes in exposure and habitat (e.g.,
-------
66 Review of EPA's EMAP
climate fluctuations, pollutant releases, species introductions). Information
on stressors will often be measured and monitored by programs other than
EMAP.
"Potential indicators are identified using conceptual models of ecosystems,
followed by systematic evaluation and testing to ensure their linkages to the
assessment endpoints and their applicability within EMAP.... The models
used may be based either on current understanding of the effects of stress
on ecosystems or on the structural, functional, and recuperative features of
healthy ecosystems."
Tiers 3 and 4. Tiers 3 and 4 are intended to focus more closely on high-
resolution trend detection and process studies. They are not yet clearly
defined (Overton et al., 1991). However, in its description of indicators, EPA
(1991c) goes into some detail: 'Tier 3 activities provide for increased
sampling intensity, either spatially or temporally, to ensure adequate informa-
tion about status and trends for subpopulations of interest (e.g., redwood
forests, low alkalinity lakes) or to provide diagnostic information beyond that
available from Tier 2 efforts. Tier 4 includes process-level research that may
be conducted at specific, nonrandomly selected sites." Most EMAP activ-
ities are currently confined to Tiers 1 and 2. It is not clear how much EMAP
effort will be devoted to Tiers 3 and 4 or how and when this effort will be
undertaken.
Evaluation
Resource Mapping. The committee supports the idea of probability-
based sampling as a means of obtaining information on the nationwide
distribution and abundance of ecological resources for EMAP; the grid is a
powerful way of ensuring that data on different resource types are integrated,
because different resource types have to be sampled near to each other.
The committee also supports the use of a systematic grid design to
characterize the spatial extent of ecological resources and changes in spatial
distribution; although a statistical average or sum might remain unchanged,
spatial patterns can change in ways that indicate environmental change (e.g.,
climate change).
However, alternative sampling schemes that might meet EMAPs goals
more parsimoniously have not been fully considered. Efficacy studies com-
paring alternatives and the proposed scheme are lacking; the EMAP grid
needs to be tested as thoroughly as any other major demonstration project.
-------
Appendix A 67
Furthermore, the sensitivity of Tier 1 to various scales of environmental
variation has not been fully tested. The grid could be ecologically biased in
some environments, even if it is statistically unbiased. The means and
confidence intervals of the sparser Tier 1 grid compared favorably with the
denser U.S. Forest Service Inventory plots in New England (Riitters et al.,
1991). But, until the grid is fully tested, it cannot be known whether ft is
terrain-dependent or resource-dependent, i.e., whether comparisons in less
or more rugged terrain or with finer-grained or coarser-grained resources
might be less favorable. Until more tests are performed, the assertion that
EMAP's grid design is unbiased must be considered only as a working
hypothesis. In addition, the geometry and other properties of resources or
ecosystems could affect the results of a sampling scheme. For example,
rivers and coasts have one dominant dimension at the scale of EMAP's grid
(they are long and narrow), while many other resources, like forests, typically
have two.
The committee also emphasizes that Tier 1 must be done completely to
be useful. Enough money must be budgeted to ensure complete coverage
of Tier 1, because absence of data will compromise Tier 1's usefulness. If
Tier 1 values are missing, they might be missing at random, but they also
might be missing for systematic, or bias-inducing, reasons (e.g., if EMAP
cannot gain site access in particular kinds of environments or during partic-
ular conditions).
For maximal usefulness, the design must be adopted by as many federal
and other agencies as possible, as would be true of any design. The hexa-
gon structure is novel, and some agencies might be reluctant to change from
a different format, especially if change requires large investments in mapping,
software, hardware, etc.
Assessment of Status and Trend Detection. Among the most impor-
tant objectives of EMAP are the assessment of the condition or status of
ecological resources and the detection of trends and new ecological and
environmental phenomena that might result from human activities. Clearly,
rf those objectives could be achieved, management and policy-making could
be improved. But the committee is unable at this stage to reach a firm
conclusion about the ability of EMAP as designed to achieve them. The
committee has a number of troubling questions, all of which have been
communicated to EPA, but none of which has been resolved—not because
EPA is unresponsive or inattentive, but because the questions are inherently
difficult.
One important question concerns the meaning and usage of ecological
"condition" and "health." Partly because ecosystems are not tightly inte-
-------
68 Review of EPA's EMAP
grated, those concepts are extremely difficult to define. It will be extremely
difficult to avoid implicit value judgements about normative assessments
about ecosystems and resources, such as "health," "nominal" and "sub-
nominal," and "condition." In addition, EMAP documents repeatedly refer to
ecosystems and their properties. Yet it remains unclear to the committee
whether EMAP really is focusing on ecosystems and, indeed, whether it
should.
Much hard, clear thinking about the crucial topic of EMAP indicators is
needed. The indicators chosen will be critical to EMAP's ability to detect
changes in properties of ecosystems and in species populations. Clearly
defined questions must be used to choose indicators and types of system
(resource) change that are important to monitor.
EPA has argued that probability-based sampling is needed (among other
reasons) so that decision-makers will receive information on status and
trends accompanied by confidence limits and hence will be able to allocate
scarce financial resources to the most important problems. But, because of
the difficulty of providing a sound scientific basis for implicit value judge-
ments about condition and because policy decisions will be made on the
basis of those nonscientific judgments, EMAP should make value judgments
explicit wherever possible.
A related question concerns the selection of what to measure and knowl-
edge of what the measurements mean. Those require an understanding and
evaluation of ecological processes and identification of the ecological
function or functions of interest. The indicators of resource condition
represent complex ecological problems that are at the heart of EMAP-
problems that are the subject of active basic ecological research and that
constitute a difficult political and social aspect of resource management. As
described in the Indicator Development Strategy document (EMAP, 1991 c),
what and how to measure are being determined through a series of work-
shops and peer reviews and are being tested in pilot demonstrations. But
peer review—even if it includes the possibility of rejection of a plan—is
probably not enough. Because indicator development is crucial to the entire
program and because the ecological issues involved are complex and
incompletely understood, the success of the program will require the
involvement of the best available scientific talent in all stages of research and
review. What to measure probably cannot be finally decided without
rigorous peer review of research proposals and research results.
A third major question is whether Tier 2 will be able to detect any
significant trends. The ability of Tier 2 to detect trends can be questioned on
various grounds. First, there is a great deal of natural variability in most
ecological variables and processes in space and time (e.g., NRC, 1986,
-------
Appendix A 69
1990). Some trends can show up within months; others require decades to
be seen. Much experience suggests that monitoring programs are most
successful when they focus on specific places and on specific questions
(e.g., NRC, 1986,1990), but that clearly is antithetical to the basic motivation
of EMAP. A second, related concern is that different ecological processes
have different characteristic time and space scales; a comprehensive
sampling scheme would have to take those scales into account. Given the
many time and space scales, will the need to distinguish signal from noise
overwhelm EMAP? A practical concern is that EMAP might be unable to
obtain data from private sources or—even worse—might be denied access
to sites. EMAP is too often discussed without adequate consideration of
possible problems and failures that could affect trend detection. The
detection of trends cannot proceed at an adequate rate until much testing
and basic research have been performed. The roles and importance of
research envisioned by EPA in its plans for Tiers 2, 3, and 4 remain some-
what obscure to the committee. In particular, EPA needs to make clear its
intentions for Tiers 3 and 4 through concise and consistent statements of its
intended objectives and techniques for achieving them.
In its discussions with EPA, the committee has mentioned alternatives to
EMAP's reliance on its grid for detecting trends, such as intensive monitoring
at sites selected on the basis of geological or ecological criteria. EMAP
officials have responded that such selection can take place within the basic
sampling plan by augmenting the grid, but it remains unclear to the
committee whether EMAP is relying too much on the grid for trend detection.
The committee is concerned lest sole reliance on the grid preclude the full
use of existing data and programs. Furthermore, the committee has not
seen evidence that EPA has adequately considered alternative approaches.
For example, in its written response to a question from the committee, EMAP
officials said that they had "not systematically documented alternative
approaches or thoroughly analyzed their strengths and weaknesses for
EMAP" (EPA, 1991e).
The committee concludes that the above questions must be answered
before EMAP is adopted as a full-scale program. The results of the demon-
stration projects must provide the answers.
Finding and Explaining Associations. One of EMAP's goals is to
"monitor indicators of pollutant exposure and habitat condition and seek
associations between human-induced stresses and ecological condition"
(goal statement of EMAP, EPA 1991 a). This goal has two unclear aspects.
The committee notes that the first part of the statement implies only that
indicators of pollutant exposure are to be monitored, and not the pollutants
-------
70 Review of EPA's EMAP
themselves. Some pollutants are being monitored by EMAP as well as other
EPA programs and other agencies, although it is not clear from EMAP
documents or presentations exactly who will conduct the field-monitoring
programs or how the data from them will be integrated into EMAP. Indeed,
identifying suitable indicators of pollutant exposure is itself challenging, as
a recent EPA-sponsored National Research Council report concluded (NRC,
1989a). The committee also notes that the cost to EMAP of monitoring all
stressors would be enormous.
The second unclear aspect is related to the establishment of causality
In recommending that EPA establish a nationwide ecological mapping and
monitoring program, the SAB pointed out that, given the regulatory
framework of EPA, it was important to know the relationships between
environmental stresses and the effects observed (EPA SAB, I988b). In one
statement (EPA, 1991 a), EMAP's goal stops short of establishing causality,
but it comes closer in another (EPA, 1991 b): determining "the likely causes
of adverse effects." Despite the SAB's recommendation, it is not clear to the
committee to what degree EMAP plans to focus on establishing cause-effect
relationships. It is clearly not planned in Tiers 1 and 2. Associations that
appear in the results will be identified for followup research, but EMAP Tier
2 projects are not cause-effect studies. Obviously, identification of an
association is not scientific proof of causality, but the committee cannot tell
whether or to what degree the establishment of causality is planned in Tiers
3 and 4. The committee emphasizes that the establishment of causal
relationships would need a carefully designed and adequately funded
program. EMAP must be clearer with respect to its plans for establishment
of cause-effect relationships and its plans for the regulatory use of the
information it produces.
SCIENTIFIC CREDIBILITY
EMAP is a science-based program that, in many respects, plans to break
scientific ground. It needs participation from the broad scientific community
at all stages of its design, testing, and implementation. Its scientific credibility
is less than it could be in two respects. The first concerns inadequate
responsiveness to peer review and the lack of involvement of the best
available scientific expertise in design, development, and implementation of
the program. The second concerns EMAP's use of the scientific literature.
-------
Appendix A 77
Involvement of the Scientific Community
and Peer Review
Peer review and the involvement of the scientific community are
frequently cited by EPA as integral to its development of EMAP. EPA has
voluntarily had its EMAP proposals peer-reviewed by a wide range of pro-
fessional organizations. However, the actual development process of EMAP
has often not included many of the best environmental scientists in academe
and elsewhere, even though they have been involved in peer review. Thus,
it is not clear to what extent the many panels that have reviewed EMAP have
had an influence on its scientific soundness. Peer review in EMAP has
consisted of advice, but it has not necessarily required consent. No panels
have been convened to choose among proposals, as is usually the case in
peer review of proposals. Committees of experts review documents and pilot
programs, meet with EMAP personnel or contractors, and give advice about
parts of EMAP, but they do not often question the general outline and goals
of EMAP. In addition, the review committees often must base their review on
incomplete documents and have little time to devote to review.
To achieve its ambitious goals, EMAP will require the involvement of the
best environmental scientists in all aspects of the program. EMAP should
make use of thematic advisory panels (i.e., panels focusing on resource
types and on such topics as indicator development) of scientists to guide the
development of requests for proposals and to assist in the evaluation of the
proposals submitted. Proposals should be considered from both inside and
outside federal agencies. To ensure that the best work is ultimately funded,
it is essential that requests for proposals be broadly circulated to the
scientific community (e.g., through newsletters and bulletins of scientific
societies).
Scientific Knowledge and Literature
One troubling aspect of the program is the general lack of evidence that
the EMAP team has adequately reviewed and evaluated the available
scientific literature. Many of the EMAP documents do not cite pertinent
literature, except EPA literature, and even relevant papers published by EPA
are overlooked sometimes. A good example of the pattern appears in EPAs
document that describes the surface-waters pilot project (EPA, I991d, in
section 3.5) in the discussion of how zooplankton data will be collected and
analyzed. Phase II of the National Surface Waters Survey included the
collection of zooplankton data from 146 randomly selected lakes in north-
-------
72 Review of EPA's EMAP
eastern United States. Analyses of those data have appeared in at least two
documents (Tessierand Horwitz, 1988,1990). The data analyzed appear to
be of exactly the kind to be collected by the pilot project, and the analyses
seem to be very similar to those planned for the pilot project. They provide
a good prediction of the kinds of patterns that will be evident in data
collected by the pilot project. Yet it appears that EMAP intends to perform
new studies instead of taking advantage of a previous EPA project. The need
to get the program going is understandable, but it must be balanced against
the need to incorporate the full set of knowledge that is currently available.
As another example, little or no use is made of state-level information.
Many states have detailed resource maps and active programs that monitor
resource conditions. It appears that the design of EMAP, a national program,
will ignore state-generated information and state-supported expertise.
Finally, this committee notes that the American Statistical Association
(ASA) EMAP Design Review Committee recommended that EMAP pay more
careful attention to the refereed statistical literature (ASA Committee on
EMAP, 1992). The ASA committee provided several statistical topics and
specific journal articles for EMAP scientists to consider (ASA Committee on
EMAP, 1992). Other examples will be documented in reviews of program
elements in future reports of the present committee.
DATA MANAGEMENT
The committee has not seen evidence that EMAP's planning has been
adequate for the entry, storage, retrieval, and processing of the large vol-
umes of data that it will generate. Software support and development, in
particular, are important to EMAP's success. Spatial data-base support for
scientific environmental applications is a subject of active research (e.g., the
Earth Observing Satellite program, EOS), but EMAP planners show little
familiarity with this research. Furthermore, they appear not to have appre-
ciated the degree to which computational support is critical to EMAP's
success or that planning for such support must be integrated into the whole
design phase. The problem must not be left to be solved as an afterthought.
Particular issues that require attention, including attention to associated
costs, are the following:
• Realistic estimates of the size and growth rate of the data base.
• Data entry and validation.
• Assuring confidentiality of data sources (i.e., assuring that providing
access to sampling sites on private land does not lead to landowner pen-
-------
Appendix A 73
alties for violating pollution standards) and providing other researchers with
access to data.
• Plans for quality assurance and quality control.
• Archiving and storage of data.
• Software systems, in particular data-base management systems,
that permit (among other functions) the integration of data generated by the
various resource groups and relevant data from other sources and permit
adequate, timely access to data by widely distributed users.
• Hardware systems.
• Data-processing capabilities that permit spatial, statistical, and
scientific analysis of the data and generation of reports for decision-makers.
• The methods of integration of data-base development and support
with the other research projects conducted by the various resource groups.
For storing and processing large amounts of scientific environmental
data, current commercial geographic information systems (GIS) are not yet
adequate, and they might never be. Many ecologists are frustrated in their
attempts to use such systems for the scientific analysis of relatively small
amounts of data. It follows that EMAP planners must give immediate
attention to the computational support of their sampling and analysis pro-
gram.
The committee has concerns as to whether EMAP data can be turned
into information that is useful to scientists and decision-makers in a timely
way. Can the information be digested and interpreted soon enough for some
management options to be implemented in time to be useful, or will EMAP
have only historical value?
INTERNAL ORGANIZATION AND LEADERSHIP
Internal Organization
The committee is concerned that EMAP is not being implemented with
a clear set of priorities and attention to the eventual integration of assess-
ments across resources. For example, the landscape-characterization
(EM AP-LC) element was to have provided information on the distribution and
extent of resource classes necessary for the selection of Tier 2 samples.
However, EMAP-UC is the least developed program element, and current
pilot projects have had to develop their own sampling frames independently.
The selection of ecological indicators has also proceeded independently in
each resource category; only recently has a temporary indicator coordinator
been named to oversee the integration of indicators among resource groups.
-------
74 Review of EPA's EMAP
Preliminary information suggests that the quality of the pilot projects is not
uniform (more detailed reviews of the forest, near-coastal, and surface-water
pilot projects are planned for the committee's next report).
The committee agrees with a recent NRC report {NRC, 1990) that stated
that "regional and national trends monitoring programs should include
research elements . . . designed to provide information critical to the
interpretation of monitoring results and to improve the design of monitoring
programs and the collection and interpretation of monitoring data." How-
ever, as now organized, EMAP has no mechanism or structure to be mod-
ified by the results of basic, investigator-originated, peer-reviewed science.
EMAP must involve the scientific community, as described above, to conduct
peer-reviewed research on the development of ecological indicators for the
assessment of conditions and on the detection of trends. It is logical to place
both research efforts in the office of the indicator coordinator, rather than to
distribute them among the ecological-resource groups, as suggested in the
Indicator Development Strategy (EPA, 1991c).
Leadership
The committee believes that some of the questions it has about the
EMAP program are a result of a lack of continuity in the program leadership
during the development process so far. Changes in management personnel
at the highest levels of EMAP lead the committee to be concerned about the
effects of instability on such a large, complex, and important program. The
extensive use of acting directors and temporary appointments such as those
under the Intergovernmental Personnel Act ("IPAs") is not satisfactory for
filling leadership positions in any organization, much less one that is in the
process of development. EPAs SAB argued for stability when it recom-
mended that "EPAs Assistant Administrator for Research and Development
(ORD) should be changed from a political to a career position" (EPA Science
Advisory Board, 1988a). Similarly, EMAP must have continuous, strong
intellectual and administrative leadership, if it is to succeed. Such leadership
is important for two reasons. First, the program must be presented to
Congress, the scientific community, other federal agencies, and the public
in a clear and coordinated way. The program, at a projected $100-300
million annually, represents a substantial investment in environmental
science. If it fails, the loss to environmental science in credibility will have
been even more serious than the unproductive expenditure of $100-300
million for a few years. Second, the program needs clear and strong
intellectual and administrative leadership within EPA to ensure that it is exe-
-------
Appendix A 75
cuted according to sound scientific principles and that it will not become a
fractured, disorganized, ineffective program or a general source of funds for
other parts of EPA to use toward their own goals, which might not be EMAP's
goals.
The committee believes that EMAP must be led by scientists of the
highest caliber and standing, who must be supported by superior organiza-
tional and management resources. This is an issue of the utmost impor-
tance; it must be resolved soon rf EMAP is to have the credibility and sound
foundation that it requires.
INTERAGENCY COOPERATION
Interagency cooperation is critical to EMAP's success; its importance
was stressed by EPAs SAB (1988b). The committee is impressed by the
degree to which EMAP is attempting to coordinate its activities with other
federal agencies. However, the degree of cooperation vanes from agency
to agency. It has been good with the Forest Service (Riitters et al., 1991;
Brooks et al., no date) and with one NOAA program. The nationwide
probabilistic sampling of ecological resources makes EMAP data attractive
to some federal and many state agencies.
The committee urges that there not be duplication of effort, especially if
data potentially useful to EMAP are being collected by some other federal or
state agency or EPA program, but not on EMAP's grid. The U.S. Geological
Survey (USGS) National Water Quality Assessment (NAWQA) program—a
program similar in motivation to EMAP—comes to mind; there is potential for
wasteful duplication of effort or for efficient cooperation and pooling of
resources. The committee understands that the NAWQA program will focus
on rivers, streams, and groundwater, and EMAP will focus on lakes and
wetlands. Although that separation of water-resource interests between the
programs is not absolute, it might be enough to prevent unnecessary dupli-
cation of effort. However, the committee is concerned that lack of coopera-
tion between the two programs could preclude or hinder a comprehensive
evaluation of conditions and trends of the nation's surface waters. At this
writing (early 1992), there is no formal cooperative agreement between
USGS and EPA concerning integration of the water-quality monitoring
aspects of EMAP and the NAWQA program.
-------
76 Review of EPA's EMAP
EMAP'S PLACE IN THE FEDERAL GOVERNMENT
The proper home for EMAP is an important question that has not
received much attention from those who could decide (Congress and the
Administration). It seems reasonable to decide whether to support the whole
EMAP package before deciding where it should be housed, but perhaps
some consideration of EMAP's proper place would clarify the assessment of
its potential value. It might turn out that EPA is indeed where EMAP should
be, but that is not yet clear. For example, an NRC committee that reviewed
NAWQA (NRC, 1989b) commented that 'there are certain inherent
advantages to having a nonregulatory agency collect, analyze, and interpret
NAWQA-type data; e.g., it may be unrealistic to expect a regulatory agency
such as EPA to objectively evaluate the effectiveness of its own water
regulatory programs and expenditures." Although that committee focused
only on water quality, the comment is valid for other media as well. One
difficulty that has been considered by EMAP is the problem of acquiring data
on private land. Will EPA abdicate its enforcement responsibility if it
discovers violations while collecting data on private land? Even if appro-
priate legislation is passed or other arrangements are made, will private
landowners willingly allow EPA access to their land? In addition, EPA is not
primarily a research or data-gathering agency, as USGS and NOAA are. The
present committee notes that Congress has recently considered the estab-
lishment of a national institute for environmental research and has mandated
a study by the National Academy of Sciences of federal environmental re-
search. This committee makes no recommendation at this time with respect
to the best place for EMAP.
BUDGET CONSIDERATIONS
EMAP is an expensive program. The committee has heard budget esti-
mates ranging from $100 million to $300 million annually for the program
when it is fully deployed (D. McKenzie, U.S. EPA, personal communication,
September 1991). Serious efforts must be made to prepare realistic budgets
and plans for the program for as long a period as possible. Commitment of
enough funds to implement EMAP over the long term will probably require
support for EMAP from the Administration and Congress as well as from the
administrator of EPA.
EMAP cannot succeed without a truly long-term commitment to its
funding. A program that seeks to monitor the state of the natrons ecological
resources and to determine trends in their condition cannot be turned on and
-------
Appendix A 77
off, or the data will become meaningless. There is a direct analogy between
EMAP and the National Weather Service. It would be impossible to have any
understanding of climate change without long-term ciimatic records, such as
those collected by the Weather Service and its predecessors. If EMAP is to
be successful, it must have the same kind of continuing financial commit-
ment.
DOCUMENTATION
Effective and timely communication has been a problem for EMAP. The
program has generated an enormous number of documents (see Appendix
for a partial list). Most are drafts and in constant revision. Some are so
technical as to be almost impossible for a nonspecialist to understand;
others are very general. Some—apparently authoritative, succinct descrip-
tions of EMAP—contradict others; see, for example, Overton et al. (1991) and
EPA (1991c) for contradictory descriptions of Tiers 3 and 4. Few publica-
tions on the scientific aspects of EMAP have appeared in peer-reviewed
scientific journals. (See Appendix B.)
As a result, although many ecologists have heard of EMAP by early 1992,
it is hard to find one who understands it and can explain it succinctly.
Indeed, after four meetings, the committee members have difficulty in
grasping the whole program or in seeing the relationship of one part to
others. Although some of this confusion may result from the committee
having access only to draft reports and to some of the issues not being fully
thought out by the EPA staff, we must also recognize that EMAP is a large,
complex, and conceptually difficult program. But it must be understood by
a larger group than those who currently understand it, if it is to get the
scientific and financial support that it must have to succeed. Therefore, a
serious, coordinated effort must be made to produce a series of documents
within a comprehensible organizational structure that are clearly written and
well edited and that clearly and understandably cross-refer to each other. It
is likely that such an effort would also help to clarify some of the other
organizational uncertainties discussed here. It would have made this com-
mittee's review much easier, especially with respect to understanding the
goals, design, and impiementation of the program.
CONCLUSIONS
EMAP is a large, ambitious, and controversial program whose purpose
is laudable and endorsed by this committee. The basis of EMAP—the statis-
-------
78 Review of EPA's EMAP
tically based sampling grid covering the entire continental United States—is
novel and is the cause of much of the controversy surrounding EMAP. Initia-
tion and development of EMAP indicates an evolving recognition by EPA that
long-term ecological monitoring and research are essential and that EPAs
statutory responsibilities to protect environmental resources are not limited
to human health.
On the basis of its review, the committee concludes the following:
• Probability-based sampling on the EMAP grid is appropriate for
delineating the extent of and periodically updating the nation's ecological
resources, if funding is sufficient to ensure complete coverage of Tier 1. The
committee emphasizes the need for thorough evaluation of the sampling
grid.
• EMAP must ensure that it has planned adequately for the entry,
storage, retrieval, and processing of the large amounts of data it will gen-
erate. Those plans must be integrated into the whole EMAP design.
• As discussed in the evaluation section, the committee has questions
about EMAP's ability to detect temporal trends.
• The ability of EMAP to achieve its other major goals—i.e., the
assessment of status of the ecological resources of the country, the
detection and characterization of changes or trends in status, and the estab-
lishment of associations between ecological condition and human-induced
stresses—is not yet clear.
• The development of indicators is a crucial prerequisite for successful
implementation of EMAP and deserves very high priority.
• EMAP might be emphasizing probability-based spatial sampling of
ecological resources at the expense of probability-based sampling of
temporal variation in indicators of ecological status, which is a major purpose
of EMAP. EMAP must ensure that adequate resources are devoted to
developing and testing the design components focused on detecting tem-
poral variation in ecological resources.
• The committee has many questions about financial and human
resources, design, data management, documentation, organization, and the
involvement of the broader scientific community in all aspects of EMAP.
Therefore, the highest priority should be given to focusing the available
resources on establishing sound leadership for EMAP, on integrating scien-
tists who are doing basic research in ecology and environmental biology into
the program, on completing the pilot (demonstration) projects and evaluating
their results, and on developing indicator strategies and plans for data
management. Those efforts must be accompanied by critical testing and
evaluation of EMAP's major components, to help answer the many sub-
stantive questions about EMAP's ability to achieve its purposes.
-------
Appendix A 79
The committee recognizes that some new and complex programs might
be of high quality but be misunderstood or underappreciated—perhaps even
canceled—during their development stages, while others turn out to be
wastes of large amounts of money and time. It is too soon to decide whether
EMAP will be successful or not without answers to the questions that the
committee has discussed in this report. The answers are needed for an
informed decision about the proper future of EMAP. It is clear that a properly
designed and well-executed program with EMAP's broad goals has great
potential value. It would be a tragedy if a lack of understanding led to a
premature decision to cancel a potentially most valuable program, but it
would equally be a tragedy if the fear of such a decision led to the deploy-
ment of a hugely expensive and wasteful program before it had been pro-
perly planned and assessed. The committee believes it is very important to
obtain complete answers to the questions raised in this report in order to
chart a solid future for the EMAP program.
REFERENCES
ASA Committee on EMAP. 1992. Review of EMAP Statistics and Design.
Report of a meeting held November 4-6,1991, San Francisco, California.
American Statistical Association, Alexandria, Virginia.
Brooks, R. T, D. Dickson, W Burkman, I. Millers, E. Cooler, and L Smith.
No date. Annual Report: Forest Health Monitoring, New England, 1990.
U.S. Department of Agriculture, Forest Service, Northeastern Forest
Experiment Station.
EPA. 1990a. Ecological Indicator Report for the Environmental Monitoring
and Assessment Program. February 1990. U.S. EPA, Off ice of Research
and Development, Research Triangle Park, N.C.
EPA. 1990b. Near-Coastal Program Plan forl 990: Estuaries. November
1990. EPA 600/4-90/033. U.S. EPA, Office of Research and Develop-
ment, Narragansett, R.I.
EPA. 1991 a. An overview of the Environmental Monitoring and Assessment
Program. EMAP Monitor, January 1991. EPA-600/M-90/022. U.S. EPA,
Office of Research and Development, Washington, D.C.
-------
80 Review of EPA's EMAP
EPA. 1991b. EEI-1. Mission Needs Statement. EMAP. Environmental
Monitoring and Assessment Program. July 1991. U.S. EPA, Office of
Research and Development, Washington, D.C.
EPA. 1991c. The Indicator Development Strategy for the Environmental
Monitoring and Assessment Program. April 1991. U.S. EPA,
Environmental Research Laboratory, Corvallis, OR.
EPA. 1991d. Environmental Monitoring and Assessment Program: Surface
Waters Implementation Plan - Northeast Lakes Pilot Survey, Summer,
1991. June 1991. Environmental Monitoring Systems Laboratory, Office
of Research and Development, U.S. EPA, Las Vegas, NV
EPA. 1991e. The Environmental Monitoring and Assessment Program.
Answers to National Research Council Questions. 2. October 1991.
U.S. EPA, Office of Research and Development, Washington, D.C.
EPA Science Advisory Board. 1988a. Future Risk: Research Strategies of
the 1990s. SAB-EC-88-040. Science Advisory Board, U.S. EPA, Wash-
ington, D.C.
EPA Science Advisory Board. 1988b. Future Risk: Research Strategies of
the 1990s. Appendix C. Strategies for Ecological Effects Research.
SAB-EC-88-040C. Science Advisory Board, U.S. EPA, Washington, D.C.
NRC. 1986. Ecological Knowledge and Environmental Problem-Solving:
Concepts and Case Studies. National Academy Press, Washington, D.C.
NRC. 1989a. Biologic Markers of Air-Pollutant Stress and Damage in For-
ests. National Academy Press, Washington, D.C.
NRC. 1989b. Review of the USGS National Water Quality Assessment Pilot
Program Interim Report of the National Research Council. Water
Science and Technology Board, National Research Council, Washington,
D.C.
NRC. 1990. Managing Troubled Waters: The Role of Marine Environmental
Monitoring. National Academy Press, Washington, D.C.
Overton, W S., D. White, and D. L Stevens. 1991. Design Report for EMAP:
Environmental Monitoring and Assessment Program. Draft, August,
1991.
-------
Appendix A 81
Riltters, K., M. Papp, D. Cassell, and J. Hazard, eds. 1991. Forest Health
Monitoring Plot Design and Logistics Study. EPA/600/ / U.S. EPA,
Office of Research and Development, Research Triangle Park, N.C.
Tessier, A.J., and R. J. Horwitz. 1988. Analysis and interpretation of zoo-
plankton samples collected during phase li of the National Lakes Survey.
Report No. 88-18, submitted to USEPA.
Tessier, A. J., and R.J. Horwitz. 1990. Influence of water chemistry on size
structure of zooplankton assemblages. Can. J. Fish. Aquat. Sci. 47:1937-
1943.
-------
Appendix B
EMAP Documents Reviewed
by NRC Committee
1. The EMAP Design Perspective - September 30,1990.
2. EMAP Landscape Characterization Research and Implementation
Plan-May 1990.
3. Indicator Development Strategy for EMAP - December 1990.
4. Design Report for EMAP Part I, May 14,1990.
5. Near Coastal Program Plan for 1990: Estuaries - November 1990.
6. A Review of the U.S.G.S. National Water Quality Assessment
(NAWQA)-Pilot Program by National Research Council.
7. The EMAP Design Perspective presentation by Scott Overton,
Oregon State University March 26,1991.
8. The EMAP Landscape Characterization 1991 Pilot Project Sum-
maries, March 1991.
9. Binder - National Research Council Review of Environmental Moni-
toring and Assessment Program, March 26-27,1991.
10. Guide to EMAP Documents for NRC Review by EPA.
11. Summary Update for the EMAP Landscape Characterization Re-
search and Implementation Plan - by Douglas J. Norton, May 1991.
82
-------
Appendix B 83
12. List of EMAP material titles and status of drafts.
13. List of EMAP Interagency Contacts, program components, and tele-
phone numbers, April 30,1991.
14. Review of the Near Coastal Report dated March 20,1991 by a panel
convened by the Estuarine Research Federation.
15. May 31,1991, letter to Rick Linthurst regarding the panel results of
the ERF review of EPAs EMAP-NC, conducted April 16-18,1990.
16. Review of the EMAP Near Coastal Program Plan for 1990 by the
Estuarine Research Federation, April 16-18,1990.
17. EMAP-NC Responses to Estuarine Research Federation Comments.
18. Peer review comments by James Hornbeck on Monitoring and
Research Strategy for Forests - Questions and Answers.
19. Binkley's preliminary comments on EMAP documents.
20. Comments about EMAP forests by Joseph B. Yavitt
21. February 28,1991, letter to Steve Paulsen from Kenneth L Dickson,
Final Report of EMAP Surface Waters Component Peer Review Panel
February 1991.
22. February 7, 1991, memo to EMAP Surface Waters Component
Review Panel from Kenneth L Dickson, Final Draft of the Surface Waters
Component Peer Review Panel (attached).
23. February 18,1991, memo to Roger Blair from Dr. Steve Paulsen and
Dr. David Larsen, EMAP Surface Waters Reconciliation memo.
24. Peer Review of EMAP Wetlands to Environmental Research
Laboratory Office of Research and Development, EPA, Corvallis, Oregon by
Peer Review Panel, November 28-30,1990.
25. Reconciliation Letter dated January 25,1990 to Dr. Roger Blair from
Nancy Leibowitz, "Research Plan for Monitoring Wetland Ecosystems" - Re-
viewers' concerns and suggestions.
-------
84 Review of EPA's EMAP
26. Review of the EMAP-Arid Strategic Plan - March 8, 1991, by a
technical review team.
27. Report of the Ecological Monitoring Subcommittee of The Eco-
logical Processes and Effects Committee - Evaluation of the Indicators
Report for EMAP - August 2,1990.
28. Comments on: Ecological Indicator Report for EMAP by Glenn W.
Suter, II, Environmental Sciences Division, ORNL.
29. February 1,1991, memo to EMAP AD's, IDs, TCs, and a few others
from Don Charles, Acting Technical Coordinator for Indicators, 'The Indicator
Development Strategy for the Environmental Monitoring and Assessment
Program" (cover letter).
30. January 22,1991, memo to Mel Knapp from Bob Hughes, Review
of the EMAP Indicator Development Strategy.
31. January 14,1991, letter to Charles (Mel) Knapp from David Rapport,
Research Coordinator on his review of the document, 'The Indicator
Development Strategy for EMAP".
32. January 21,1991, letter to Donald Charles, EPA from Brock Bern-
stein, Ph.D. on his review of the Indicator Development Strategy.
33. January 23, 1991, letter to Charles Knapp from Richard Latimer,
Acting Technical Director, EMAP-NC. Comments on the "Indicator Develop-
ment Strategy" document from Dr. John Scott (SAIC) and Dr. Dan Campbell.
34. December 24,1990, memo to Don Charles from Ann Fairbrother on
her comments on EMAP Indicator Strategy.
35. May 15,1990, letter to Dr. Rick Linthurst, Directorfrom Linda Young,
Chair, American Statistical Association, et al., on their concerns before
reviewing the statistical progress of EMAP. April 1990.
36. December 20,1989, memo to Rick Linthurst from ASA Review Com-
mittee for EMAP on Grid Design for Characterization and Tier 1 Sampling.
-------
Appendix B 85
37. January 25,1991, letter to Mr. Douglas Norton from Gene Thorley
on the final version of the report entitled, Report of the Peer Review Panel:
"EMAP Landscape Characterization Research and Implementation Plan",
Environmental Monitoring and Assessment Program - June 1990.
38. EMAP - Responses to National Research Council Questions
(booklet) July 1991.
39. Characterizing Dissolved Oxygen Conditions in Estuarine Environ-
ments by S. Weisberg (Versar, Inc.), J. Summers (EPA), A. Holland (Versar,
Inc.), J. Kou (Versar), V Engle (Technical Resources, Inc.), D."Breitburg
(Academy of Natural Sciences),R. Diaz (Virginia Institute of Marine Science).
40. Appendix to Question 12 from NRC committee concerning Memo-
randums of Understanding and Interagency Agreements.
41. Appendix to Question 15 from NRC committee concerning Peer
Review Comments and Reconciliations.
42. Comments on Ecological Indicator Report for the Environmental
Monitoring and Assessment Program Glenn W. Sutler II - Environmental
Sciences Division, ORNL March 14,1990.
43. June 4,1991, memo to Thomas E. Dixon, Acting Associate Director,
EMAP from William Laxton, Director of Technical Support Division,
comments on EMAP Ecological Indicator Report.
44. February 26,1991, memo to Robert Blair, Chief, Watershed Branch
from Donald Charles. Reconciliation of comments on 'The Indicator
Development Strategy for the EMAP Program*.
45. Reconciliation Memorandum EMAP, Arid Ecosystems Strategic
Monitoring Plan, 1991 - Summary of Peer Review Comments.
46. Report of the Ecological Monitoring Subcommittee of the Eco-
logical Processes and Effects Committee, August 2,1990, Evaluation of the
Indicators Report for EMAP - A Science Advisory Board Report.
47. Final Report of the Surface Waters Component Peer Review Panel -
dated February 1991.
-------
86 Review of EPA's EMAP
48. Appendix 4-B. The Interpenetrating Design for EMAP by W Scott
Overton, April 1990.
49. EMAP Responses to Comments by the Science Advisory Board,
Ecological Processes and Effects Committee, Ecological Monitoring
Subcommittee on "Evaluation of the Indicators Report for EMAP" March
1991.
50. EMAP - Plan for Converting the NAPAP Aquatic Effects Long Term
Monitoring (UM) Project to the Temporally Integrated Monitoring of Eco-
systems (TIME) Project - Internal Report. June 1990.
51. Appendix 4-A to Question 4: Changes and Trends, by Scot Overton,
May 1991.
52. Summary, Results of the First Sampling Design Workshop for
Estuaries, May 9-10,1991, Columbia, MD. May 21,1991.
53. EMAP - Example Environmental Assessment Report for Estuaries -
May 1991.
54. Implementing EMAP data management facilities - Progress Review,
January 17,1990.
55. Science Advisory Board's Review of the EMAP plan. Letter dated
July 30,1991 to The Honorable William Reillyfrom Raymond Loehr, Chair-
man, Science Advisory Board and Kenneth Dickson, Ecological Processes
and Effects Committee. Report to the Ecological Monitoring Subcommittee
of the Ecological Processes and Effects Committee - Evaluation of the
Program Plan for EMAP, July 1991.
56. Letter dated July 30,1991 to The Honorable William Reilly from Dr.
Raymond Loehr, Dr. Kenneth Dickson, and Dr. Richard A. Kimerle, Chairman,
Marine Disease/Diagnostic Task Group. Report of the Marine Disease
Diagnostic Task Group of the Ecological Processes and Effects Committee -
Evaluation of the Proposed Center for Marine and Estuarine Disease Re-
search, July 1991.
Letter dated March 21, 1991 to Dr. Robert Menzer, Director, Gulf Breeze
Environmental Research Laboratory Kenneth L Dickson, Richard A. Kimerle,
and Edward S. Bender. Response to two questions addressed as part of the
SAB review.
-------
Appendix B 87
57. Letter dated August 11,1991 to Sheila David from K. Bruce Jones,
Associate Director, Terrestrial Ecosystems (EPA), with Data Confidentiality
Report by Dr. Sue Franson. Copies of reports dealing with information
management, a draft copy of a user mission needs assessment and informa-
tion management plans for Near Coastal and Forest field activities, and a
manuscript highlighting 1991 indicator development for EMAP-Forests.
57a. Proposed Policy and Rationale: Use of Data Collected Under the
Auspices of EMAP - July 1991.
57b. Graph - Complete Software Life Cycle.
57c. Graph - Exhibit 1-3 System Category EEI Matrix.
57d. EEI-1-Mission Needs Statement - EMAP - July 1991.
57e. Appendix A. EEI-1 Mission Needs Statement EMAP Agro-
ecosystems.
57f. Appendix D. EEI-1 Mission Needs Statement EMAP - Information
Center.
57g. Appendix E. EEI-1 Mission Needs Statement EMAP Forests.
57h. Appendix F. EEI-1 Mission Needs Statement EMAP - Great Lakes.
57i. Appendix G. EEI-1 Mission Needs Statement EMAP - Integration &
Assessment.
57j. Appendix I. EEI-1 Mission Needs Statement EMAP - Near Coastal.
57k. Appendix K. EEI-1 Mission Needs Statement EMAP Wetlands.
57I. Binder: FY91 Indicator Evaluation Field Study for Environmental
Monitoring and Assessment Program - Forests (EMAP-F) June 1991.
58. Data Management for Near-Coastal Demonstration Project, August
1990.
59. Surface Waters Monitoring and Research Strategy - February 1991.
-------
88 Review of EPA's EMAP
60. Monitoring and Research Strategy for Forests, March 1991.
61. Agroecosystem Research Plan 1991 - February 1, 1991 - Peer Re-
view.
62. Arid Ecosystems Strategic Monitoring Plan, 1991.
63. EMAP Research Plan for Monitoring Wetland Ecosystems - January
1991.
64. Example Environmental Assessment Report for Estuaries, May 1991.
65. Design Report for EMAP - August 1991.
66. Surface Waters Implementation Plan - Northeast Lakes Pilot Survey,
Summer 1991 - June 1991.
67. The Indicator Development Strategy for EMAP, April 1991.
68. Forest Health Monitoring Plot Design and Logistics Study, August
1991.
69. Annual Report: Forest Health Monitoring, New England 1990.
70. Surface Waters Implementation Plan - Northeast Lakes Pilot Survey,
May 1991 -REVIEWS.
71. Environmental Monitoring and Assessment Program: Surface
Waters Implementation Plan - Northeast Lakes Pilot Survey, June 1991 -
Response to Reviews.
72. The Environmental Monitoring and Assessment Program - Re-
sponses to National Research Council Questions October 1991, Part 1 and
Part 2, October 1991.
73. Monitoring the Condition of Agroecosystems, by Julie Meyer and
George Hess, August 4-7,1991.
74. Memorandum dated March 10,1992 to William K. Reilly, Chair from
Erich Bretthauer, Executive Director of Task Force 45-Day Task Force on
Science Recommendation, Requested 1-Page Action Plans.
-------
Appendix B 89
75. Review of EMAP Statistics and Design (Review meeting held
November 4-6, 1991, San Francisco, CA), Prepared by the American
Statistical Association Committee on EMAP, November 1991.
76. Implementation Plan for Monitoring the Estuarine Waters of the
Louisianian Province, 1991, October 1990.
77. EMAP Estuaries Component: Louisianan Province 1991 Demon-
stration Field Activities Report, January 1992.
78. A Selection of Forest Condition Indicators for Monitoring'Abstract,
January 1990.
79. Booklet: Safeguarding the Future: Credible Science, Credible
Decisions, March 1992.
80. Agroecosystem Monitoring and Research Strategy, May 1991.
81. Arid Ecosystems Strategic Monitoring Plan, 1991, June 1991.
82. EMAP Estuaries Virginian Province 1990 Demonstration Project
Report, January 15,1992.
83. Estuaries Virginian Province Logistics Plan for 1991, April 1991.
84. Near Coastal Louisianian Province 1992 Sampling, Field Operations
Manual, March 1992.
85. Indicator Development Strategy (Updated Version 1.5) June 2,
1992.
86. Final Peer Review Panel Report of EMAP Great Lakes Monitoring
and Research Strategy, (Draft EPA Document February 1992), Prepared for
U.S. EPA. Submitted by Research and Evaluation Associates, Inc. April 15,
1992.
87. The Relationship Between EMAP and CASTNET, July 1992.
88. EMAP Great Lakes Response to Peer Review Panel Report, June
1992.
-------
90 Review of EPA's EMAP
89. EMAP-Estuaries Virginian Province 1990 Demonstration Project Re-
port, June 1992.
90. Technical Design Proposal, Clean Air Status and Trends Network
(CASTNET), External Review Draft, February 18,1992.
91. EMAP Estuaries: A Review Organized and Facilitated by the Estua-
rine Research Federation, March 30-31,1992. Report dated April 10,1992.
92. EMAP Estuaries: Draft -1991 Virginian Province Field Activities Re-
port, December 1991.
93. Executive Summary of the Third Review of EMAP Estuaries,
Organized and Coordinated by the Estuarine Research Federation, April 10,
1992.
94. EMAP Great Lakes Monitoring and Research Strategy, June 1992.
95. EMAP Great Lakes Response to the Peer Review Panel Report,
April 1992.
96. EMAP - Program Plan, February 1991.
97. Monitoring and Research Strategy for Forests, March 1992.
98. EMAP-IM 93 Tactical Plan Component 1: Consolidated Statement
of Work, October 10,1992.
99. Information Management Strategic Plan: 1993-1997, Version 1.6
September 30,1992.
100. FY91 Forest Health Monitoring, Western Pilot Operations Report.
101. EMAP-Estuaries Virginian Province 1990 Demonstration Project
Report, June 1992.
102. EMAP-Surface Waters Response to Peer Review Comments, Jan-
uary 1993.
103. Summary of ORD Workshop to Address and Resolve issues and
Concerns regarding EMAP, Research Triangle Park, North Carolina,
September 1992.
-------
Appendix B 91
104. EMAP Response to the National Research Council's Interim Report,
June 1992.
105. EMAP-Surface Waters 1991 Pilot Report, February 1993.
106. Report on the Ecological Risk Assessment Guidelines Strategic
Planning Workshop-Risk Assessment Forum, February 1992.
107. Hexagon Mosaic Maps for Display of Univariate and Bivariate
Geographical Data by Daniel B. Carr, Antony R. Olsen, and Dennis White,
(Abstract from Cartography and Geographic Information Systems, Vol 19,
No. 4,1992, pp. 228-236, 271).
108. Forest Health Monitoring 1991 Statistical Summary.
109. Ambient Water-Quality Monitoring in the United States, First Year
Review, Evaluation and Recommendations, December 1992.
110. EMAP-Arid Colorado Plateau Pilot Study -1992: Implementation
Plan, January 1993.
111. EMAP-Agroecosystem 1992 Pilot Plan (April 3,1992).
112. Annual Statistical Summary: Agroecosystems A Hypothetical Ex-
ample, August 1990.
113. EMAP-Response to Congress on the National Research Council
June 1992 Report, April 12,1993.
114. Implementation of a National Monitoring Program, April, 1993.
115. Reprint: Comparing Sampling Designs for Monitoring Ecological
Status and Trends: Impact of Temporal Patterns.
116. Statistical Summary: EMAP-Estuaries Louisianian Province -1991,
January 1993.
117. EMAP FY1994 Issue Planning Paper, March 15,1993.
118. Summary of ORD Workshop to Address and Resolve Issues and
Concerns Regarding EMAP (Research Triangle Park, NC, September 1992),
January 7,1993.
-------
92 Review of EPA's EMAP
119. Program Guide, May 1993 (handout May 24-25, 1993 meeting).
Updated version - June 1993.
120. Louisianian Province Demonstration Report, EMAP-Estuaries: 1991,
dated October 1993.
121. Environmental Monitoring and Assessment Program Master Glos-
sary, October 1993.
122. SAB Review of EMAP's Draft Assessment Framework, September
30,1993.
-------
Appendix C
Biographical Sketches of
Committee Members
RICHARD FISHER, chair, serves as professor of forest soils, head of the
Forest Science Department, and director of the Institute for Renewable
Natural Resources at Texas A&M University. His research interests are in
soil-plant interactions in both temperate and tropical systems and the use
of experimentally derived knowledge for managing forest productivity. He
is a fellow of the Soil Science Society of America and the Society of
American Foresters and is president-elect of the National Association of
Professional Forestry Schools and Colleges. He served on the National
Research Council Committee on Forestry Research and is currently Co-
Editor-in-Chief of Forest Ecology and Management. Dr. Fisher received his
B.S. from the University of Illinois in 1964 and his Ph.D. from Cornell
University in 1968.
PATRICK L BREZONIK received a B.S. in 1963 from Marquette Univer-
sity, an M.S. in 1965 from the University of Wisconsin, and a Ph.D. in water
chemistry in 1968 from the University of Wisconsin. Dr. Brezonik is professor
of Environmental Engineering and Director of Water Resources Research
Center at the University of Minnesota. He was chairman of National Re-
search Council Panel on Nitrates in the Environmental in 1975-1978; and the
National Research Council on Restoration of Aquatic Ecosystems. He is
currently a member of the National Research Council's Water Science and
Technology Board. Dr. Brezonik research interests are eutrophication of
lakes, nitrogen dynamics in natural waters, nutrient chemistry, acid rain, trace
metals in natural waters, and organic matter in water.
INGRID C. BURKE is a Research Associate, Natural Resource Ecology
Laboratory, Colorado State University. She is currently involved in inter-
disciplinary research programs investigating the control of plant productivity,
soil organic matter turnover, and trace gas flux in the Great Plains. Dr. Burke
93
-------
94 Review of EPA's EMAP
was a member of the EPAs EMAP Landscape Characterization Panel and
NSF's Conservation and Restoration Biology Panel, 1990. She has a B.A. in
biology from Middlebury College and a Ph.D. in botany from the University
of Wyoming. She has published several journal articles and publications and
is a member of the Association of Women in Science, the American
Association for the Advancement of Science, American Institute of Biological
Sciences, Ecological Society of America, and Soil Science Society of
America.
EDWIN H. CLARK, II received a B.S. in engineering from Yale University
in 1960, an M.S. (water resources engineering) and MA (economics) in 1966
from Princeton University and a Ph.D. in economics from Princeton in 1971.
Dr. Clark is an expert in water quality and agriculture as management issues.
He has worked as a consulting engineering for Harza and did research in
Pakistan relative to water supplies for agriculture. He taught economics at
Williams College and served as a senior staff member at the Council on
Environmental Quality through 1978. For three years (until 1981), Dr. Clark
served as deputy director of the U.S. EPAs Office of Pesticides and Toxic
Substances, where he was concurrently a special assistant to the admin-
istrator. Until recently, he was vice president of The Conservation Founda-
tion's Water resources program. He is now secretary of the Department of
Natural Resources and Environmental Control for the state of Delaware.
LOVEDAY L. CONQUEST received her BA in mathematics in 1970 from
Pomona College, her M.S. in statistics from Stanford University in 1972, and
her Ph.D. in biostatistics from University of Washington in 1975. Dr.
Conquest is a biostatistician working in the areas of environmental
monitoring (experimental design, sampling design, data analysis/inter-
pretation), natural resource management (e.g., fisheries, forestry, ecology),
and related areas. She is an associate professor in the Center for Quanti-
tative Science (COS) in Forestry, Fisheries, and Wildlife at the University of
Washington, and heads the Statistical Consulting Laboratory for COS. She
has provided consulting services to other researchers, environmental con-
sulting firms, and public agencies. Dr. Conquest has provided expert witness
testimony on environmental data and is very familiar with assessing the
"statistical quality" of information that arises from diverse monitoring
programs.
ARTHUR W. COOPER received a BA in 1953 from Colgate University, a
MA in 1955, and a Ph.D. in botany in 1958 from the University of Michigan.
His research interests are plant ecology; general plant sociology; micro-
-------
Appendix C 95
environments; resource management; and forest productivity. Dr. Cooper
is currently Professor of Forestry North Carolina State University.
SHIRLEY DREISS received a B.S. with honors in 1971 in geology from
the University of Texas, an M.A. in 1974 in geology/hydrogeology, and a
Ph.D. in geology/hydrogeology from Stanford University in 1980. She was
a member, National Academy of Sciences Committee on the Mono Basin
Ecosystem from 1987-1990 and a panelist, U.S. Geological Survey, Water
Resources Research Program, Section 105 of Water Research Act of 1984,
Public Law 98-242. Dr. Dreiss's research activities were geostatistical
characterization of heterogeneous soils, fluid flow and deformation1 in active
subduction zones, variable density groundwater circulation in arid basins,
moisture and contaminant transport through variably saturated soils, and
quantitative modeling of groundwater flow in karst aquifers.
THURMAN L. GROVE received a B.A. in 1966 from Wilkes College and
a Ph.D. from Cornell University in ecology and soil science in 1982. He is
currently Professor of Soil Science, Assistant Dean of the College of Agri-
culture and Life Sciences and Director of International Programs at North
Carolina State University. His research interests are biogeochemistry,
agroecology, and international development.
JOHN E. HOBBIE received a BA in 1957 in ecology from Dartmouth
College, an M.A. in 1959 from the University of California, and a Ph.D. in
zoology in 1962 from Indiana University. His was Assistant Professor in
zoology to Professor at North Carolina State University from 1965-1975. Dr.
Hobbie is currently Senior Scientist, Marine Biology Lab, and Director,
Ecosystems Center at the Marine Biology Lab, Woods Hole, Massachusetts.
Dr. Hobbies research interests are arctic and antarctic limnology, hetero-
trophic bacteria in aquatic ecosystems, estuarine ecology, and global carbon
cycle.
CHARLES C. JOHNSON, JR. received his M.S. in civil engineering from
Purdue University He is a retired environmental engineering consultant. He
was president of C.C. Johnson & Malhotra, P.C. from 1979-1991; vice
president of Malcolm-Pirnie, Inc., 1974-1979; and a commissioned officer in
the U.S. Public Health Service from 1947-1971 when he retired as an Assis-
tant Surgeon General. He has had a long career in environmental engi-
neering and public health. Mr. Johnson has had the responsibility of
managing and administering programs reflecting every concern for the
environment including water, pesticides, solid waste, air pollution, food and
-------
96 Review of EPA's EMAP
drugs, occupational health, and community sanitation. These experiences
have given him hands-on opportunity to develop, direct, and administer
environmental activities associated with program development, monitoring
for compliance, research-enforcement programs, establishment of informa-
tion systems, and interaction with the public and community groups. He is
a member of the National Academy of Engineering.
TIM KRATZ received a B.S. in 1975 (botany) from the University of Wis-
consin-Madison, an M.S. (ecology and behavioral biology) from the Univer-
sity of Minnesota-Twin Cities in 1977, and a Ph.D. (botany) from the
University of Wisconsin-Madison in 1981. From 1981-1985 he was Project
Associate and Site Manager, Northern Lakes Long Term Ecological Research
Project. He is currently Assistant Scientist and Site Manager for that project.
His research interests include limnology, landscape/lake interactions, wet-
land formation, and landscape ecology.
ANNE E. MCELROY received a B.S. in aquatic biology from Brown Uni-
versity in 1976 and a Ph.D. in oceanography from the Massachusetts Institute
of Technology/Woods Hole Oceanographic Institution Joint Program in
1985. From 1986-1991 she was an assistant professor in the Environmental
Sciences Program at the University of Massachusetts-Boston. Dr. McElroy
currently serves as the director of the New York Sea Grant Institute and holds
an appointment as an associate professor at the Marine Sciences Research
Center at the State University of New York at Stony Brook. Her research
interests concern how aquatic organisms interact with toxic chemicals in the
environment.
JOHN PASTOR received his B.S. in geology in 1974 from the University
of Pennsylvania, an M.S. in soil science in 1977 from the University of
Wisconsin-Madison, and a Ph.D. in forestry and soil science in 1980 from the
University of Wisconsin-Madison. Dr. Pastor is currently Research Associate,
Natural Resources Research Institute at the University of Minnesota; Adjunct
Professor, Department of Ecology and Behavioral Biology, University of
Minnesota; and Adjunct Professor, Department of Fisheries and Wildlife, also
at the University of Minnesota. His research interests are northern eco-
systems, nutrient cycling, climate change, forest productivity, timber
management, and landscape ecology.
JAMES N. PITTS received a B.S. in 1945 from the University of
California, Los Angeles, and a Ph.D. in chemistry in 1949. He has been a
member of National Academy of Sciences/National Research Council Advi-
-------
Appendix C 97
sory Board on Military Personnel Supplies for the Committee on Textile
Functional Finishing, and Chairman of the Panel of Polycyclic Organic Matter
for the Committee on Kinetics of Chemical Reactions. Dr. Pitts is Professor
Emeritus in the Department of Chemistry and past Director of the Statewide
Air Pollution Research Center at the University of California, Riverside.
Currently, he is Adjunct Professor in the Department of Chemistry and
Biochemistry, California State University, Fullerton, and Coordinating
Instructor in Air Pollution in the Extension Division of the University of
California, Irvine. His research interests include fundamental processes in
photochemistry and photooxidations and their application to the atmospheric
chemistry of photochemical smog, acid rain, airborne toxics, and mutagenic
and/or carcinogenic pollutants.
RAYMOND A. PRICE received his Ph.D. in geology from Princeton
University in 1958. He served as the Director General, Geological Survey of
Canada from 1981 to 1987. His contributions integrate many earth science
subdisciplines and have provided essential links between global-scale plate
tectonic processes characterizing the recent revolution in geoscience on one
hand and the local-scale interpretations of specific continental regions on the
other. He is very familiar with GIS and management of large data gathering
programs. He is a member of the National Academy of Sciences foreign
associates.
SAUL B. SAILA received a B.S. from University of Rhode Island in 1949,
an M.S. in fishery biology in 1950 from Cornell, and a Ph.D. in fishery biology
in 1952. He was a fishery biologist at the Rhode Island Department of
Agriculture and Conservation. He was professor of oceanography and chief
scientist, Office of Marine Programs at University of Rhode Island 1975-1988.
In 1988 he became Emeritus Professor and consultant. His research
interests are in fish population dynamics and stock assessment.
TERRENCE R. SMITH received a Ph.D. in 1971 from Johns Hopkins
University. He attended the Graduate School of Management, University of
Rochester, New York: Doctoral Program in Applied Economics in 1975-
1976. Dr. Smith is currently Chairman, Department of Computer Science,
University of California, Santa Barbara, Professor of Computer Science, and
Professor of Geography, University of California, Santa Barbara. He is also
Associate Director, National Center for Geographical Information and
Analysis, Co-Director, Remote Sensing Unit, Co-Director, Center for the
Study of Spatial Cognition and Performance, and Associate Director, Com-
munity and Organization Research Institute. Dr. Smith's research interests
-------
98 Review of EPA's EMAP
are machine intelligence, spatial databases, spatial cognition, and motion
planning.
SUSAN STAFFORD is a forest biometrician at the Department of Forest
Science, Oregon State University. Dr. Stafford consults with forest science
researchers on the design of experiments in forest ecology, forest genetics
and on the collecting, handling, and analysis of data. She is also data
manager for the H.J. Andrews Experimental Forest and director and creator
of the Forest Science Data Bank at Oregon State University Dr. Stafford
received her Ph.D. in applied statistics in 1979 from State University of New
York, College of Environmental Science and Forestry.
DONALD R. STRONG, JR. received a B.S. in zoology in 1966 from the
University of California, Santa Barbara; an M.S. in biology in 1968 from the
University of California, Irvine; and a Ph.D. in biology in 1971 from University
of Oregon. His areas of interest are community and population ecology,
insect-plant interactions, biological control, natural enemies of herbivorous
insects, statistical ecology, biogeography, tropical ecology, and salt-marsh
ecology. Dr. Strong is the recipient of the Albert W. Boyce Award in Ento-
mology, University of California, Riverside, 1990.
MICHAEL J. WILEY received a B.G.S. in 1973, an M.S. in 1976, and a
Ph.D. in 1980 from the University of Michigan. He is Assistant Professor,
School of Natural Resources, University of Michigan. Dr. Wiley's research
interests are ecology of river systems, benthic invertebrates, and fisheries
management. Prior experience includes Associate Professional Ecologist,
Illinois Natural History Survey, 1984-1987. Dr. Wiley is a member and re-
viewer of the Ecological Society of America, American Fisheries Society, and
North American Benthological Society.
------- |