EPA/620/R-94/022
                                       March 1994
      Environmental Monitoring
      and Assessment Program
   Indicator  Development Strategy
                 Edited By

              M. Craig Barber
      Environmental Research  Laboratory
     U.S. Environmental Protection Agency
              Athens, Georgia
               EMAP Center
Environmental Monitoring and Assessment Program
      Office of Research and development
     U.S. Environmental Protection Agency
      Research Triangle Park, NC 27711
                                         Printed on Recycled Paper

-------
                                           Notice
       The research described  in this document was funded by the U.S. Environmental Protection
Agency.  The document was prepared at the EPA Environmental Research Laboratory in Athens,
Georgia with technical contributions through contract #68-CO-0021  with Technical Resources, Inc.,
purchase order OBO 387 NTTA to Arizona State University, cooperative  agreement CR817489 with
Indiana University, and contract #68-08-0006 with ManTech Environmental Technology, Inc.
Mention of trade names or commercial products does not constitute endorsement or recommendation
for use.
                                          Abstract
       This document outlines the strategy of the Environmental Monitoring and Assessment
Program (EMAP) for indicator development. This strategy consists of general  approaches,  criteria,
and procedures for selecting and evaluating indicators in the context of a long-term monitoring
program.  In practice, EMAP's indicator development is a multiphase process to identify and then
evaluate indicators that can estimate the condition of ecological resources using synoptic survey
monitoring methods defined over large geographic areas.  The strategy described  in this document
outlines this process from the identification of potentially useful indicators through  the adoption of a
set of core indicators for use in EMAP.  This strategy is  intended to promote internal consistency of
indicator development across ecological resources and to provide a basis for internal and external
review of EMAP indicators.  The implementation and review of this process will continue throughout
the existence of EMAP.


Key words:  indicators (biology), environmental indicators, ecological monitoring,  environmental
monitoring, research,  United States, environmental condition,  U.S.EPA-EMAP.


Preferred citation:
Barber, M.C., ed. 1994. Environmental Monitoring and Assessment Program:  Indicator Development
        Strategy.  EPA/620/R-94/XXX. Athens,  GA: U.S.  Environmental Protection  Agency, Office of
        Research and Development, Environmental  Research Laboratory.
                                     Acknowledgments
       Many individuals contributed directly and indirectly to the development of this document.
Foremost among these are Mel Knapp, Dave Marmorek,  Joan Baker, Kent Thornton, Jeff Klopatek,
and Don Charles who co-authored  EMAP's original Indicator Development Strategy.  That technical
report was revised by Tony Olsen and was used as the basis for the current document.  Many of the
initial elements of EMAP's Indicator Development Strategy were developed during  the EMAP
Indicator Strategy Development Workshop that was held  in Las Vegas,  Nevada, in June 1990.  In
addition to the original authors,  key people who participated in that workshop included: John Baker,
Steve Bromberg, Dean Carpenter,  John Eaton, Chris Elvidge, Jerry Filbin, Sue Franson,  Luis
Hernandez,  Bruce Jones, Bill Kepner,  Bev Law, Nancy Leibowitz,  Chris Maser, Dan McKenzie,  Jay
Messer, Tom Moser, Dave Mouat,  Tony Olsen, Steve Paulsen, Jim Pollard, John Scott, WooIIcott
Smith, Louisa Squires, Renee Stang, Gary Turner, Steve Weisberg, and Jamie Wyant. In revising
the report, the editor benefitted  from numerous technical  discussions with many EMAP scientists
including Hal Kibby, Dan McKenzie, Tony Olsen,  Eric Hyatt, Rick  Linthurst, Dave Bradford, Walt
Heck, Lee Campbell,  Bill Kepner, Sam Alexander, Dick Latimer, Kevin Summers, Steve Hedtke,
Steve Lozano,  Phil Larsen,  Steve Paulsen, Spence Peterson, and Bruce Jones. Technical editing
was provided by Cynthia B. Chapman  ELS,  ManTech Environmental Technology, Inc.
                             EMAP Indicator Development Strategy
                                              ii

-------
                                          Contents

1.  Introduction  	,	 ... .	  1

2.  Background	 . ,		  3
       2.1 EMAP Overview	.'....'.-	  3
       2.2 Program Objectives		  3
       2.3 Classification of Ecological Resources and Assessment Regions  . . .	  4
       2.4 Values and Assessment Questions as Program Foundations	  5
       2.5 Quantifying Resource Condition	'.	 .   11
              2.5.1  Ecological Indicators and Risk Assessment Endpoints	   11
              2.5.2  Developing Indicators for Ecological Monitoring	   12
              2.5.3  Resource Condition and Monitoring Design	   15

3.  Framework for Indicator Development	.;...,.	   17
       3.1 Overview of EMAP's Indicator Development Strategy	•..,....	   17
       3.2 Coordinating  the Development  Process	   19
              3.2.1  Internal  Integration		.	   19
              3.2.2.  External Integration	   20

4.  Indicator Selection	   22
       4.1 Identifying Environmental  Values and Assessment Questions .	   22
       4.2 Identifying Expected Stressors	   23
       4.3 Identifying Potential Indicators	   26
              4.3.1  Developing Conceptual Models of Resources' Structure and Function  ...   27
              4.3.2  Selecting  Research Indicators	   34

5.  Indicator Evaluation	'.	   36
       5.1 Indicator Performance Criteria	   36
       5.2 Phases of Indicator Evaluation		   38
              5.2.1  Conceptual Considerations	;	   42
              5.2.2  Operational Monitoring  Considerations	• .   43
              5.2.3  Statistical Evaluation  of Indicators	   43
              5.2.4  Evaluation of Indicators for Resource Assessments .  .	 .   44
                     5.2.4.1  Identification of Nominal-Subnominal Criteria	 .   46
                     5.2.4.2  Condition-Stressor Associations	   49
                     5.2.4.3  Example Assessments	   50
       5.3 Evaluation Methods	 . . .	   51
              5.3.1  Existing Data and Desk-Top Studies	   51
              5.3.2  Pilot Research Projects	   53
              5.3.3  Regional Demonstration Projects  .	   56
       5.4 Identification  of Indicators for Implementation	   57
         ...'...'	 . . .	. . . ;	'..:. ..:...:		   59

6.  Indicator Implementation	   60

7.  Indicator Reevaluation	   61
       7.1 Reevaluation Procedures	 . ,	   62

8.  Integration Among Resource Groups	 :	   64
       8.1 Framework for Indicator Integration	:.'•..	 .	 .   64
       8.2 Types of Indicators that Integrate Across Ecological Resources	 .   65
              8.2.1  Indicators Linking Resource Groups	   66
                             EMAP Indicator Development Strategy
                                              iii

-------
               8.2.2  Indicators Shared by Resource Groups	   66
        8.3  Use of Conceptual Models to Facilitate  Integration	   67
        8.4  Coordination of Indicator Development Among Resource Groups	   67
        8.5  Problems Associated with  Differences in Spatial and Temporal Scales	   68

9. Concluding  Remarks	 .	   70
        9.1     Planned Reviews	   70
        9.2     Evolving Process	   70

10. References	   72


                                        List of Figures

Figure 2-1.     Standard Federal Regions as defined by OMB Circular A-105	'.	  6
Figure 2-2.     Cumulative distribution  of an indicator's score over the resource population.
               The proportion of the population less than or equal to a particular score can
               be determined with the associated confidence interval.  	  9
Rgure 2-3.     Conceptual model of the estuarine ecosystem.  Solid lines indicate material
               flows and dashed lines indicate interaction. Redrawn from Holland (1990)	   14
Rgure 4-1.     Conceptual model for the trophic structure of arid ecosystems. Redrawn from
               Evenari et at. (1986)	   29
Figure 4-2.     Conceptual models for the trophic structure of the soil community in
               agroecosystems.  Adapted from  Moore and de Ruiter (1991).   	   30
Figure 4-3.     Conceptual model for the vegetative succession  in arid ecosystems.  Adapted
               from Graver and Musick (1990)	   31
Figure 4-4.     Conceptual model for functional biodiversity  of macroinvertebrates  in stream
               as a function of stream order and associated landscape.  Redrawn from
               Minshali et al. (1985)	   32
Figure 5-1.     Illustration of the relationship between power and magnitude of trends
               detectable and years of monitoring.  From left to right, curves are for trends of
               approximately 2%/yr, 1.5%/yr,  1%/yr, and 0.5%/yr	   52
Figure 5-2.     Example of a pilot research project design for evaluating indicator
               responsiveness to major stressors in Virginian Province estuaries (Holland
               1990)	   55
Figure 5-3.     Cumulative distribution for Index of Biotic Integrity in  streams in Ohio during
               four months of 1986 (after Paulsen et al. 1991)	   58

                                         List of Tables

Table 4-1.      Environmental Values  Selected by Different  EMAP Resource Groups	   24
Table 4-2.      Definitions of Selected Environmental Values by EMAP  Resource Monitoring
               and Research Groups.  These values reflect Resource Group thinking as of
               early 1993.  Definitions have evolved since that time	   25
Table 4-3.      Conceptual model for the relationships between fish  assemblages and trophic
               conditions in the Great Lakes.  Adapted from Ryder and Kerr (1978)	   33
Table 5-1.      Indicator Evaluation Criteria	   37
Table 5-2.      Example of the Early Stages of an Indicator's Evaluation	   39
Table 5-3.      Example of the Later Stages of an Indicator's Evaluation	   40
Table 5-4.      Indicator Status Sheet that Tracks Moving Condition  Indicators from Research
               to Implementation Mode	   41
                              EMAP Indicator Development Strategy
                                              iv

-------
Table 5-5.     Summary of Estimates of Components of Variance for Secchi Disk
              Transparency (SD), Chlorophyll-a (Chl-a), and Total Phosphorus (TP) Derived
              from the Vermont Lake Monitoring Database	  45
Table 5-6.     Great Lakes Resource Group examples of nominal - subnominal criteria
              (ranges) established for the Laurentian Great Lakes	  48
                             EMAP Indicator Development  Strategy
                                              v

-------

-------
                                        1. Introduction
In 1989, the U.S. Environmental Protection Agency's  (EPA) Office of Research and Development
initiated the Environmental Monitoring and Assessment Program (EMAP), an interagency research,
monitoring,  and assessment program.  When fully implemented, EMAP will be an integrated,
multi-resource  program that will estimate the condition of the Nation's ecological resources at various
geographic  scales over long periods of time and will provide possible explanations for regional and
national trends in these estimates  (Messer et al. 1991).  EMAP has been designed to provide
information  needed to conduct "top-down" or effects-driven risk assessments (Messer 1990).  In
"top-down" risk assessments,  the observation of an effect stimulates efforts to identify plausible
stressors that might have caused the effect by focusing on trends in resource condition and seeking
associations between indicators of resource condition and stress.  This approach is designed to
detect cumulative impacts of natural and anthropogenic influences on the condition of ecological
resources rather than effects caused by a limited set of individual stressors.

EMAP's success depends on  its ability to characterize the condition of ecological resources.
Because of the intrinsic difficulty in defining what is meant by ecological "condition" of any ecological
resource at large,  EMAP will  monitor suites of environmental indicators that collectively describe the
condition of the resource of concern.  Within EMAP, an indicator is defined to be any environmental
measurement  that can be used to  quantitatively estimate the condition of ecological resources, the
magnitude  of stress, the exposure  of biological components to stress, or the amount of change in
condition.                                              I

This document outlines EMAP's strategy for indicator  development.  This strategy presents general
approaches, criteria, and procedures for selecting and evaluating  indicators in the context of a long-
term monitoring program. EMAP's indicator development is a multiphase process to identify and     :
evaluate indicators  that can estimate the condition of ecological resources defined over large
geographic  areas using synoptic survey monitoring  methods. The procedures  described  in this
document outline this process from the identification of potentially useful indicators (research
indicators) through the adoption of a set of core indicators for use in a national  monitoring program.
This strategy is intended to promote a consistent  approach to indicator development across
ecological resources and to provide a basis for internal and external  review of EMAP indicators.
The implementation of the strategy and refinement of the process will continue  throughout the
existence of EMAP.
                              EMAP Indicator Development Strategy
                                               1

-------
Although this document focuses on the development and  use of indicators within EMAP, many of the
issues discussed and the general strategy have broader applications.  Detailed, process-oriented
research on indicators will be conducted by other research programs;  ongoing and planned
monitoring  programs will be collecting information on indicators, and adding new indicators in their
monitoring  networks.  EMAP's objectives willbe best served through a close linking of the EMAP
indicator development process to these ongoing and planned research and monitoring programs. As
these programs  mature, EMAP's indicator development process likewise will evolve and improve, and
this strategy will be revised and updated.

This document is a revision of two previous technical reports (Knapp et al. 1991; Olsen 1992). Much
of the text is taken directly from these reports, portions of the text are slight modifications, and new
text has been  added to reflect EMAP's current stage in indicator development.  Some of the
background material presented in this report relies on Hunsaker and Carpenter (1990) and Knapp et
al. (1991).  In  some cases,  text from  these sources have been slightly modified to reflect recent
changes in EMAP or to alter its emphasis to address indicator development issues. The document is
organized as follows

              Section 2 provides background on EMAP and the role of indicators in ecological
              monitoring.
              Section 3 presents an overview of the indicator development strategy.
       •      Section 4 discusses procedures for identifying potential EMAP indicators.
       •      Section 5 discusses procedures for evaluating potential EMAP indicators.
       •      Section 6 discusses procedures for implementing core EMAP indicators.
       •      Section 7 discusses procedures for periodic review of core  EMAP  indicators.
              Section 8 discusses procedures for coordination and integration of indicator
              development across resource categories.
       •      Section 9 provides concluding remarks on the strategy.
              Section 10 lists the references cited.

For additional  information regarding the use of indicators in EMAP resource groups, readers  should
refer to the reports and documentation from each EMAP resource group.
                             EMAP Indicator Development Strategy
                                              2

-------
                                       2. Background
2.1  EMAP Overview
In 1988, EPA's Science Advisory Board recommended implementing a program to monitor ecological
status and trends for identifying emerging environmental  problems before they reach crisis
proportions.  In the following year, the Administrator of EPA established an Agency priority for the
1990s to confirm that the Nation's annual expenditure on environmental  issues is producing
significant results toward maintaining and improving environmental quality  (Reilly 1989).  In an  effort
to identify environmental  problems  before they become irreversible and to allow the evaluation  of
regional and national progress toward achieving environmental  goals, EPA's Office of Research and
Development (ORD) began planning the Environmental  Monitoring and Assessment Program
(EMAP).  Initiated in 1989, EMAP was created to monitor and assess the condition of the Nation's
ecological resources, thereby contributing to decisions on environmental protection and management.
Consequently, EMAP represents a critical element in  the overall development of EPA's Ecological
Risk Assessment methodologies.

2.2  Program  Objectives                            ,

EMAP is designed to provide information  needed to quantify  the current status and trends in the
condition of the Nation's  ecological resources  and information needed to diagnose why those
conditions might exist.  Several key questions  that have guided  the development  of EMAP's specific
programmatic  objectives  include:

               What is the current extent of our ecological resources and how are
               they distributed geographically?
               What proportions of the resources are currently in acceptable
               ecological condition?                  ;
               What proportions are degrading or improving, in what regions, and at
               what rates?
               Are these changes correlated  with patterns and trends in
               environmental  stresses?
               Are adversely  affected resources improving in response to control and
               mitigation programs?
                              EMAP Indicator Development Strategy
                                               3

-------
Based on these questions, EMAP designed an interdisciplinary  research, monitoring, and
assessment program with the following objectives

               Estimate 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.
               Estimate the geographical coverage and extent  of the Nation's
               ecological resources with known confidence.
               Seek associations between selected  indicators of natural and
               anthropogenic stresses and indicators of condition of ecological
               resources.
               Provide annual statistical summaries  and  periodic assessments of the
               Nation's ecological resources.

To meet these objectives, EMAP must develop indicators of the condition of ecological  resources and
of natural and anthropogenic stressors influencing this condition.  The indicators must be compatible
with the monitoring design to permit quantitative and unbiased estimates of ecological status and
trends, analysis of associations between  indicators of ecological condition and environmental stress,
and sufficient flexibility to accommodate sampling of multiple resource types.

2.3 Classification of Ecological Resources  and Assessment Regions

EMAP's programmatic goal  is to monitor and assess the condition of the Nation's natural resources,
thereby contributing to decisions on environmental  protection and management.  To do so, EMAP will
provide regional and national assessments of the condition  and extent of ecological  resources that
are defined hierarchically  based on specific assessment needs.   In most cases, the condition of
specific ecosystems or landscapes will be the object of concern. In some cases, however, the
condition of specific populations or communities also could  be the focus.  At the lowest  level of
resolution, EMAP partitions all of the Nation's ecological resources into seven broad ecological
resource categories, i.e., agroecosystems,  arid  ecosystems, estuaries, forests, the Great Lakes,
surface waters (lakes and streams), and  wetlands. These resource categories are, in essence,
aggregations  of similar ecosystem types.  EMAP also recognizes landscapes, i.e., heterogeneous
land areas composed of clusters  of interacting  ecosystems  that are repeated in similar form
throughout the area, as a resource.  Each resource category, in turn, can be subdivided into distinct
resource classes  (e.g., oak-hickory forest). Resource classes can be further subdivided, as needed,
                              EMAP Indicator Development  Strategy
                                               4

-------
into resource subclasses that may correspond to specific community assemblages or to specific
structural components  of the ecosystem of concern  (e.g.; bird assemblages of oak-hickory forest).

The development,  monitoring,  assessment, and reporting of indicators  for each of these resource
categories is the responsibility  of an EMAP resource group  that is named after the resource
categories.  For example,  indicator development, monitoring, and assessment for agroecosystems is
the responsibility of the EMAP-Agroecosystems resource group.

The regions for which EMAP will estimate the condition and extent of ecological  resources are
likewise hierarchically defined  based on specific assessment needs.  At a minimum,  however, EMAP
is committed to report on the condition and extent of ecological resources within each of the 10
standard Federal regions defined by Circular A-105  (OMB 1974; see Figure 2-1). EMAP resource
groups, however,  are not constrained  to report only  on these regional resources.  In  many instances,
EMAP data may convey more  meaningful  information to decision makers and environmental
professionals if they  are reported for specific biogeographical regions within the standard Federal
regions.  Decisions to report at such finer scales of resolution, and therefore to satisfy EMAP's data
quality objectives (DQOs), are the responsibility of each  EMAP resource  group.

2.4 Values and Assessment  Questions  as Program  Foundations

In order for EMAP  data to be useful to decision makers,  EMAP's indicator development  and
implementation  must be designed to answer specific assessment questions that are  related to values
that society bestows  on the resources of concern.  These values and assessment questions are
therefore the basis for EMAP's  indicator development, monitoring,  and  assessment.

Identification of the values that society places on the Nation's ecological  resources is the initial step
for developing EMAP indicators. The  word, value, has multiple definitions including the  monetary
worth  of something (i.e., its marketable price), the desired use of something or its aesthetic quality
(e.g.,  beauty). Value, as defined in EMAP, are those characteristics  of the environment desired  from
an ecological resource.  These  ecological resource  values are those  characteristics of the
environment that contribute to society's quality of life including the ability  of the resource to provide
food,  fiber, clean water and air, recreation, desired plant and animal communities, and an aesthetic
experience.  Values  for the ecological  resource by one segment of society  can conflict with values for
the resource desired  by another societal segment.  However, indicators selected for  monitoring
should permit both segments to determine if the resource values are being achieved.  In general,
                             EMAP Indicator Development Strategy
                                              5

-------
Figure 2-1.    Standard Federal Regions as defined by OMB Circular A-105.
                            EMAP Indicator Development Strategy
                                            6

-------
EMAP recognizes three broad categories of environmental values,  i.e., ecological structure and
function,  consumptive uses, and non-consumptive uses.

After values of a resource have been  identified, then specific assessment questions related to those
values must be formulated.  This is an iterative process that requires an understanding  of the
relationships  among policy and science, assessment practicality and utility, and monitoring logistics.
Formulating the assessment questions that guide EMAP's programmatic endeavors is not a trivial
exercise.  These questions must be of long-term interest  to the regulatory and decision-making
community and to environmental scientists and ecologists.  For many values pertaining to a
resource's  consumptive and non-consumptive  uses, assessment questions often can be formulated
to meet the needs of specific legislative mandates, agency policies, or management regulations.  For
values related to ecological structure and function, however, it often will be necessary or desirable to
develop conceptual models of the  resource of concern in order to formulate the most critical and
encompassing assessment questions  associated with those values and the resource of interest.

EMAP is concerned with three basic types of assessment questions:

              What is the current status of the resource?
              What are the trends in the resource's status over time?
              What are the associations  between the resource's current condition
              and the occurrence of selected stressors?

Assessment  questions regarding the resource's status are the most basic. Such assessment
questions identify the assessment endpoints, i.e., formal expressions of the environmental  value
that is be protected, with which EMAP will be  concerned.  These assessment endpoints must have
unambiguous operational  definitions,  be biologically and socially  relevant, be accessible to
measurement, estimation  or prediction, and  be susceptible to environmental stressors of concern
(Suter 1990). A complete operational definition of an assessment endpoint must identify both a
subject and a characteristic of that subject.  Generally  speaking,  assessment  endpoints  are the
specific ecological components (i.e., populations,  species, guilds, communities,  etc.) or processes
(e.g., primary and secondary production, nutrient assimilation  and cycling, etc.) that can be related
conceptually  to the environmental  value of concern  (EPA 1992).  General formats for EMAP's
assessment  questions regarding a resource's  current status for a single period  of time include:
               What proportion (by area, length, or number) of resource R (i.e.,
               subject) in region X is in condition C (i.e., characteristic) ?

                              EMAP Indicator Development Strategy
                                               7

-------
        •       What proportion (by area, length, or number) of resource R (i.e.,
               subject) in region X has attribute A (i.e., characteristic) ?
        •       What is the extent (i.e., characteristic  by area, length, or number) of
               resource R (i.e., subject) in region X ?
        •       What is the extent (i.e., characteristic  by area, length, or number) of
               resource R with attribute  A (i.e., subject) in region X ?

The information used to answer these questions are EMAP's condition indicators (see Section 2.5.1).

It is important to note that the first two generic  EMAP assessment questions above focus on the
cumulative distribution of condition  indicators rather than measures of central tendency and variation
per se (Figure 2-2).  This perspective provides  more information than does the latter since
it allows for analysis not only of the central tendency of the endpoint but also of its limiting behavior
in  its extremes.  In many  cases, it may be more important to know either that the proportion of a
resource in poor or subnominal  condition  is increasing or that the proportion of a resource in optimal
or minimally distributed condition is decreasing,  rather than knowing  the average condition of the
resource at large.  Such information can be assessed  with known confidence only by estimating  the
cumulative distribution of the resource's condition  indicators.

While condition indicators alone do not provide enough information to conduct a risk assessment,
summarizing the results as a cumulative  distribution is consistent with the current risk assessment
framework (EPA 1992,  EPA 1994). Suter (1990) asserted  that the product of any  risk assessment
should be either an estimated probability of a dichotomous assessment endpoint or the probability
that an assessment endpoint will be greater than or less than some scalar value.  For example, in
the first case the risk assessment might estimate the  probability for a species' regional extinction,
whereas for the latter case, the risk assessment might estimate the probability that the number of
fishless lakes within a region is greater than some value X.  An assessment endpoint is therefore
inextricably  linked  to a formal  mathematical expression describing the likelihood  of that endpoint.
Because the proportion of a regional resource that exhibits a particular attribute  is conceptually
equivalent to the probability that any individual  resource within  that region exhibits that characteristic,
it follows that answers to  EMAP's assessment questions should provide environmental decision
makers with information consistent with current ecological  risk assessment methodologies and
philosophy.
                              EMAP Indicator Development Strategy
                                               8

-------
Figure 2-2.    Cumulative distribution of an indicator's score over the resource population.
              The proportion  of the population less than or equal to a particular score can be
              determined with the associated confidence interval.
               1.0
              0.75 —
          a.
          £
          •5
          jo
          t
          S.
0.50 —
              0.25 -
                                            Upper Confidence
                                                 Interval
                                              30
                                       Indicator Score
                            EMAP Indicator Development Strategy
                                            9

-------
In most cases, the resource with which EMAP is most concerned is the entire resource category or
resource class, that is, resources  at the ecosystem or community level of organization.  Specific
examples of possible  EMAP assessment  questions at this scale of concern include:

              What proportion of agroecosystem units in the Southeast has erosion
              exceeding tolerable limits?
       •      What is the extent of riparian communities in southwestern arid
              ecosystems?
       •      What proportion of forest  land in the Pacific Northwest has visual
              crown ratings that are expected to impact their nominal growth
              potential?
       •      What proportion of the Great Lakes has forage fish populations
              capable of supporting lake trout or walleye populations?
       •      What proportion of lakes in the Northeast is hypereutrophic?
       •      What proportion of wetlands in the prairie pothole region provides
              habitat suitable for breeding or over-wintering  waterfowl?

Many resources, however, are valued not only as a component of a larger ecological entity but also
as an entity in their own right. Consequently, there may be cases where the resource of concern is  a
specific species assemblage  or population rather than the ecosystem itself.  Two specific examples
of potential EMAP assessment questions  at this scale of concern might be

       •      What proportion of forest  bird communities in  the Southeast has its
              expected  biodiversity?
       •      What proportion of salmonid smolts in the Pacific Northwest is
              maintaining expected growth rates?

Both of these assessment questions have their counterparts at the ecosystem or landscape scale of
resolution, for example:

              What proportion (by area) of forests in the Southeast has its  expected
              avian biodiversity?
       •      What proportion (by length) of rivers and streams in the Pacific
              Northwest supports salmonid smolts with expected growth rates?
                             EMAP Indicator Development Strategy
                                              10

-------
It is extremely important to recognize, however, that although these assessment questions are
related, they are not identical. These questions specify very different statistical populations and  may
require different indicators or design considerations.  Thus, when confronted with such dichotomies,
EMAP resource groups must carefully consider the implications and  ramifications of their decision to
deal with one question or the other.

2.5 Quantifying  Resource Condition

2.5.1  Ecological Indicators and Risk Assessment Endpoints

EMAP defines an indicator to be any expression of the environment that quantitatively estimates the
condition of ecological resources, the magnitude of stress, the exposure of biological components to
stress, or the amount of change in condition.' In the past, EMAP has recognized many types of
indicators (Hunsaker and Carpenter  1990; Knapp et al. 1991; Olsen  1992).  However, in a effort to
clarify and simplify its conceptual foundations and terminology, EMAP now  considers only two
generic types of indicators:                            ',

  •    condition indicator— any characteristic of the environment  that provides quantitative
       information on the state  of ecological resources  and is conceptually tied to a value.   If
       necessary, condition indicators can be subdivided into
               biotic  indicator—any characteristic of the environment that  estimates
               the condition of a biological component of the resource,  and
               abiotic indicator—any characteristic  of the environment that estimates
               the condition of physical or chemical components  of the  resource.
               Some indicators  might be both condition and stress indicators (e.g.,
               dissolved oxygen).
       stressor indicator—any characteristic of the environment that is suspected to elicit  a
       change in the condition  of an ecological resource.

EMAP's definition of a condition  indicator is conceptually equivalent to what the ecological risk
assessment literature calls a measurement endpoint,  i.e., "a measurable  environmental
characteristic that is related to the valued  characteristic  chosen as the assessment endpoint" (Suter
1990, 10).  For additional discussion, see RAF (1992) and Norton  et al.  (1992).
Using the term "indicator" rather than "endpoint,"  however,  reenforces EMAP's program objective to
estimate the condition of the Nation's ecological resource's at multiple scales.  No monitoring

                              EMAP Indicator Development Strategy
                                               11

-------
program could possibly monitor and assess all aspects of a resource's condition.  However, a
program could monitor and assess selected indicators of the resource's condition that address
specific environmental  values and assessment questions that are also indicative of the condition of
other non-monitored resource components.

An indicator may be a  single field measurement, an index based on multiple  measurements,  or the
output of a mathematical simulation model that has been parameterized  using field measurements.
Karr et al. (1986) developed the Index of Biotic Integrity (IBI) to describe conditions in freshwater
streams.  Properly developed,  indices of resource condition can be compared more easily across
regions than can the individual measurements  from which they are derived (Hughes  1989;  Karr
1991). The aggregation process, however, can be highly controversial and mathematically  complex.
Although results tend to be extremely dependent on the data and the aggregation  procedures used
(Westman 1985), indices afford a valuable methodology for assessing  resource condition.
Consequently, the development of indices for assessing resource condition will be pursued as an
important concept of EMAP's indicator development program.  Simulation results from mathematical
models also can be used as condition indicators of the resource of concern.  Barnthouse (1992)
discusses the utility of  mathematical models to assess environmental concerns on local,  regional and
global scales.  Models  that simulate important ecological  processes  such as primary or secondary
productivity, nutrient cycling, or inputs to other resource categories may be particularly useful in this
regard. Although presently EMAP has developed few indicators of this type,  such  indicators  might
provide the most effective means to address the interactions between  resource categories  (e.g.,
sediment transport from agroecosystems to surface waters).

2.5.2 Developing Indicators for Ecological Monitoring

Having identified  environmental values and assessment questions  of concern, potential indicators are
identified using  a variety, of conceptual models.  These models may be based either on current
understanding of the structure and function of ecological resources in good condition or on the
responses and recuperative capacities of stressed  resources.   The resulting set of research
indicators then are evaluated systematically to confirm their relationships to EMAP assessment
questions and their utility toward meeting EMAP's programmatic objectives.  The product of this
evaluation process is a set of core indicators that will be used for EMAP's routine monitoring  and
assessment activities.
Developing and documenting conceptual models that describe  a resource's  structure and function is
an essential part of EMAP's indicator development process.  Not only do such models provide a

                             EMAP Indicator Development Strategy
                                              12

-------
focus to select indicators for previously  defined assessment questions, but they also can be used to
identify assessment questions that might be overlooked if considering only the assessment needs of
current environmental regulation and management.  Conceptual models should identify, at some level
of detail, all significant components and processes of a resource that contribute to its ecological
organization and operation.  These models then can provide strategic frameworks to identify and
develop indicators.  In particular, conceptual models can be used to

               link indicators to their identified value(s),
               identify gaps and redundancies in the indicators needed to address
               assessment questions of concern,       ',
               identify indicators that can be used as surrogates  for resource
               elements of primary concern but which cannot be addressed
               effectively due to data collection and measurement  constraints, and
               suggest schema to construct indices or other quantitative models  for
               evaluating a resource's condition (see Section 4.3.1).
                                                     i
Conceptual  models also can provide  frameworks for resource assessments and reporting by

               identifying anticipated stressed  and unstressed behaviors,
               identifying anticipated relationships between stressor and condition
               indicators, thus providing a data analysis strategy for diagnosing
               plausible causes of subnominal conditions,  and
               identifying linkages to other ecological resources

Conceptual  models are important  representations of EMAP's scientific understanding of the
ecological resource for monitoring  purposes.  They should  be descriptive and clearly demonstrate  the
relationships between indicators and  among condition indicators,  assessment endpoints,  and
resource stressors. An example of such a model developed  for the estuarine environment is
presented in Figure 2-3.
In addition to estimating  resource condition, it is also important to suggest or associate plausible
causes of poor or degrading conditions.  Therefore, condition indicators must be selected because of
their likely responses to the major anthropogenic and natural stressors that are likely to impact the
resource.  If information about these stressors is not being compiled by other programs,  EMAP will
not necessarily develop stressor indicators for routine monitoring.  EMAP will monitor selected
stressor indicators when a relationship between  a specific condition and  stressor is known or if a

                              EMAP  Indicator Development Strategy
                                               13

-------
I
p
1

1
•O
CD



I
ro
£
ro


I

I
o
CO
I

i~

|i
oC.

§1
"5 2

f!


II
Mo _
o g
"S g

1*
ED:
ra c
3 p
o g
C» flj
  a*-*
  =
1
q>
5T

                                                                    o>
                                                                    Q-
                                                                    ^
                                                                    0)

                                                                    I
                                                                    LLJ
                                      14

-------
testable hypothesis can be formulated.  EMAP also may develop stressor indicators to confirm and
evaluate the responsiveness of research condition indicators with no intention of long-term
implementation.  When stressor indicators  are developed, EMAP will use them primarily  to examine
their statistical association, on a regional scale, with indicators of resource condition.  Although these
correlative  analyses  cannot establish causality, they can serve to narrow the range of probable
causes for observed regional patterns and trends in resource status through process of elimination
procedures. More detailed monitoring and research efforts to determine cause-and-effect
relationships can then be focused on those geographical areas,  stressors,  and resource  classes of
greatest concern.                                     ;

Indicator development and program  implementation in each major resource category are the
responsibility of an EMAP resource group.  The indicator development program of each  EMAP
resource group will be tracked through the use of research plans, peer reviews, and interaction with
other resource groups.  Each resource group will produce research and monitoring  plans that are
subject to written peer review.  These plans will describe the evidence and  rationale for selecting
each  research indicator,  identify the specific types of information needed to evaluate these indicators,
and discuss important implications of design and assessment associated with each indicator.  The
current status  of all indicators and rationale for all decisions made during the development process
will be documented annually in an indicator data base.  It is envisioned that this data base will be
available as part of the EMAP Information  Management: system, and thus could be used to facilitate
rapid  review of both  the current state and the evolution of all indicators used by each of  the resource
groups (see section  3.2.1).                            i

2.5.3  Resource Condition and  Monitoring Design   ,

EMAP's monitoring design focuses on implementing sampling designs that  enable status and trend
estimates to be  made for the resource classes or subclasses specified by an EMAP resource group.
Each resource class or subclass  is a statistical population for which estimates of ecological condition
are made.  Indicators of resource condition are monitored on sampling units that constitute a
probability  sample of the  population  (i.e., resource class or subclass).  These indicators  are
developed  in such a way that they will apply to entire resource classes or subclasses and will
produce  population estimates of their ecological condition.

In order to provide reliable estimates of the status and trends in the condition of our Nation's
ecological  resources, indicators must be monitored regionally for long  periods of time (i.e., years  to
decades) using  a statistically designed network for probability-based sampling  of explicitly  defined

                              EMAP Indicator Development Strategy
                                               15

-------
ecological resources (or populations in statistical sampling terminology).  Moreover,  indicators must
be monitored on sampling units that are consistent with their level of ecological  organization.  These
indicators also should provide meaningful information across the entire resource.  EMAP resources
are sampled during index periods.  An index period is a period  of the year when measurement of an
indicator provides meaningful information about the condition of the resource. The index periods are
generally resource specific.  For example, the lake index period is during late summer while the
stream index periods during the spring.

EMAP's sampling design builds on experience  gained from previous surveys, incorporates their key
features, and uses a systematic grid (the EMAP grid) to ensure random selection and appropriate
sampling distribution (White et al. 1992).  Other agencies, including  the U.S. Departments of
Agriculture, Commerce, Energy,  and Interior, have active, ongoing monitoring programs that address
some of EMAP's needs for broader monitoring  data.  EMAP will develop procedures for directly
integrating data and components from these monitoring programs into  the EMAP grid, where the form
and nature of the data are appropriate.  In cooperation with other agencies,  EMAP will supplement
existing networks to fill critical data gaps. For additional  detail  on the EMAP design, see Overton et
al. (1990), Messer et al. (1991), White et al. (1992), and  Urquhart et al. (1993).
                              EMAP Indicator Development Strategy
                                              16

-------
                           3.  Framework for Indicator Development

3.1 Overview of EMAP's  Indicator Development Strategy

Although EMAP's original indicator development strategy recognized six distinct phases in the
development of indicators,  activities outlined for these phases are usually iterative and do not
represent a step-by-step schedule for indicator development.  Conceptually, all indicator development
activities are intended to

              define the purpose, focus, and scope of EMAP's indicators,
       •      evaluate the utility of proposed research indicators,
       •      implement national monitoring of selected core indicators that will
              satisfy EMAP's programmatic objectives, and
       •      reevaluate core indicators periodically and develop new indicators  as
              needed.

These four general  categories of efforts simply are designated  as  indicator selection, evaluation,
implementation, and reevaluation.

During indicator selection,  EMAP  resource groups

              identify the  environmental values of their resource,
              document or formulate assessment questions that motivated or
              followed from identified values,
              identify major stressors that are expected to be most important  for
              interpreting  the status and trends of condition indicators in tentative
              diagnostic assessments,                '
       •      develop conceptual models  that depict the resource's structure  and
              function and likely responses to stressors of concern,  and
              select indicators for research and evaluation.

Because EMAP's indicator  development process is an assessment-driven activity, environmental
values and long-term assessment questions are the catalysts  for the development of EMAP
indicators.  EMAP's assessment questions  must be of long term interest (e.g., 10- to 20-year time
frame) and be related to well defined environmental values.
                              EMAP Indicator Development Strategy
                                              17

-------
During indicator evaluation,  EMAP resource groups use best professional judgement,  existing data,
and results of pilot research and regional demonstration projects to

               evaluate the logistics of monitoring each indicator on regional and
               national scales,
       •       characterize each indicator relative to its temporal and spatial
               variability, ecological responsiveness,  and overall  interpretability,.  '              -
       •       identify or develop  nominal-subnominal criteria for condition indicators,
       •       prepare example statistical  summaries and resource assessments,
       •       determine sampling density needed to estimate regional status and
               trends and assess  associations between'condition  indicators and
               selected  stressors, and                     r
       •       select core indicators for program implementation.
                           '            .      •             .     •           :;> .
In EMAP's original Indicator Development Strategy (Knapp et al. 1991),  the primary focus of
indicator evaluation was the characterization of the indicator's  logistic feasibility and statistical
behavior.  Identification  of nominal-marginal-subnominal  condition thresholds was a task implicitly
relegated  to the assessment  process.  If EMAP indicators are  to add value to the decision-making
process,  they must be able to address "so what" questions.  For example, is the majority of the
resource in nominal, marginal or subnominal condition?  Consequently, indicators must be evaluated
explicitly  on their ability to address this fundamental assessment need.

After receiving approval through formal  programmatic or peer review, core indicators  are moved
forward to the indicator  implementation  stage of the development  process.  During this stage, EMAP
resource groups

               monitor  core indicators nationally,                  '
       •       prepare annual statistical summaries,  and             -  '
       •       prepare periodic resource assessments in'conjunction with the'EMAP-
               Assessment and Reporting  coordination  group:

During the final stage of the indicator development process (indicator reevaluation), EMAP  resource
groups                              "               •                          -
               periodically reevaluate core indicator performance,
               identify emerging assessment questions, and

                              EMAP-. Indicator Development Strategy
                                               18

-------
               conduct research on new indicators.

Although this strategy is written to address the development of individual indicators, it should  be used
by each EMAP resource group to assess its full suite of indicators.  Often multiple  indicators may be
under development at the same time.  The objective of this process is to develop a comprehensive
suite of indicators  that complement each other and  provide a clear picture of the status and trends in
the ecological condition of resources of interest through  time.  It is anticipated that, due to limited
financial or human resources, time, or scientific knowledge, EMAP resource groups will be
developing indicators at different rates.

3.2 Coordinating the  Development Process

Integration among EMAP resource groups, State  governments, Federal  agencies, and
nongovernmental  organizations  is critical to the success of EMAP.  Some of the primary concerns
related to internal  and external integration are discussed below.

3.2.1  Internal Integration

Although  implementation  of the indicator development process is the primary responsibility of the
individual  EMAP resource groups, to fully achieve EMAP's goals these activities must be integrated
and coordinated.  Important  issues that must be addressed include: 1) consistency in terminology
regarding environmental values,  assessment questions, and conceptual models, 2) consistency  in
collecting and applying stressor information, 3) inclusion of indicators that link or integrate resource
categories (e.g., landscape  pattern, soil carbon, mobile  wildlife, etc.), 4) use of  common indicators,
compatible sampling, and analytical methods, and 5) co-locating sampling units for special studies.
Thus, it is essential  that all  EMAP  resource groups formally communicate on a regular basis  to
facilitate intergroup  information exchange,  to avoid redundant data collection efforts, and to improve
the amount of information available for each EMAP resource group to use in assessing  status and
trends in  ecological  resource condition.  Because program .implementation  for each EMAP resource
group will proceed at different rates, intergroup integration will foster and improve  the efficiency and
effectiveness of the overall  program.

 Each EMAP resource group will compile and add timely information about its indicator development
 activities  to the EMAP  indicator database. The indicator database will store up-to-date  information
 about each indicator being  evaluated  or considered by  each  EMAP resource group. This data base
 should  contain  retrievable records of the indicator status sheets (Table  5-4) that contains all  pertinent

                               EMAP Indicator Development Strategy
                                               19

-------
information about each indicator ever considered by the resource group, including at least the level of
detail presented in the appendices to EMAP's original report on ecological indicators (Hunsaker and
Carpenter 1990). Once an indicator is listed, it should never be deleted from the data base even
though it can be deleted  from further consideration (status:  rejected) or revised and refined during
later stages of indicator development.

As information is developed during the process of evaluating indicators in each of the four phases of
indicator development, it should be added promptly to the EMAP indicator data base.  Data base
listings for each indicator should be initiated during the identification of research indicators and
updated during each of the subsequent phases of the indicator development process.  Indicator
information entered into the data base should be kept simple and short to make the data base easy
to update.  It is extremely important that each resource group's indicator data base be updated
frequently.  Not all information categories accommodated by the data base will  be evaluated during
the first phases of indicator development; therefore,  not all data base entries will be completed while
updating records for each indicator. The appropriate results from each phase of the development
process should be entered into the data base following completion  of that phase of development.

Although the design of the data base  has not been completed, the  structure and format of the
indicator data base will be consistent  across all EMAP resource groups. This consistent structure will
allow for information exchange  and synthesis.

3.2.2.  External Integration

The Science Advisory  Board's Ecological Monitoring Subcommittee has stressed  the importance of
interagency coordination  and integration to  the success of EMAP (SAB 1990).  Although integrating
results from other monitoring  efforts into EMAP  is both efficient and essential, interagency
cooperation  should also include sharing information and expertise.  For example,  in addition to
valuable data that can be obtained from the USDA Forest Service's Forest Inventory and Analysis
program, Forest Service  personnel can be active participants in the indicator development process
for the EMAP-Forests  resource group. Similarly, the U.S. Fish and Wildlife Service, National
Biological  Survey, the  Bureau of Land Management, the Soil Conservation Service, the National
Oceanic and Atmospheric Administration, and many other Federal and State agencies can contribute
to the indicator development  efforts of EMAP resource groups.

Formal arrangements,  such as  Memorandums of Understanding,  should be established by  EMAP
resource groups as necessary to assist the indicator development process and to ensure that EMAP

                              EMAP Indicator  Development Strategy
                                              20

-------
develops appropriate tools to monitor the condition of ecological resources.  For example, it may be
appropriate for EMAP resource groups to obtain indicator information from sources that could include
State agencies (e.g., for regional resource management actions), other Federal agencies (e.g., USDA
for soil erosion rates and crop production data), and other EPA programs (e.g., Office of Water for
pollutant discharge  information).

Interest in the use of indicators and indices extends beyond EMAP.  The National Academy of
Sciences (1975) discussed the need for ecological indicators to monitor the environment (ecosystem
condition) and to judge the effectiveness of environmental  protection programs.  Indices of
ecosystem condition also can  be used  to

               prioritize funding for dealing with environmental problems,
               rank locations  (regional comparisons),
               conduct environmental  trend analysis,
               provide public  information,
               condense and  focus scientific research, and
               enforce standards (Ott  1978).                                                  ,
                              EMAP Indicator Development Strategy
                                              21

-------
                                     4.  Indicator Selection

The first phase of the indicator development process establishes a framework for indicator
interpretation by identifying the environmental values,  assessment questions, primary environmental
stressors, and critical ecosystem components and processes for each resource of concern.  This
phase defines the context of EMAP's programmatic concerns and establishes the conceptual and
functional relationships among  environmental values, assessment questions, condition indicators, and
selected indicators of stress.

4.1 Identifying Environmental Values and Assessment  Questions

Ecological resources have both intrinsic and extrinsic worth ranging from the societal value placed  on
the protection of pristine ecosystems and their inherent ecological structure and function (e.g.,
biodiversity and nutrient cycling) to the direct economic benefits derived from resource harvests (e.g.,
agricultural and timber production  and commercial fisheries).  The first step toward developing
indicators that allow meaningful assessments of the condition of the Nation's ecological resources is
to identify the major environmental values associated with each EMAP  resource category.

Each  of the resource groups, depending on their needs and experience, initially identified
environmental values from a variety of sources, which include:

        •       legislative mandates  (e.g., Endangered Species Act, Wetlands
               Protection Act,  Clean Water Act, etc.),
        •       agency policies,
        •       management regulations,
        •       literature review - including  articles in newspapers and popular
               magazines as well as articles published in scientific journals,
        •       government reports,
        •       conceptual models,
        •       workshops to determine expert opinion,
               EMAP  management and guidance, and
        •       contributions of peer reviewers.

The process of identifying environmental values and associated assessment questions requires a
broad perspective regarding  both the desired structure and function of resources of concern (as
expressed by resource managers, scientists, private industry, legislators, and the general public) and

                              EMAP Indicator Development Strategy
                                               22

-------
stressors that are likely to impact the resource (which may occur on local to global spatial scales,
and over short- to long-term temporal scales).  In some cases, environmental  values can be identified
                                                                 \
from existing regulatory  assessment questions.  In other cases, the values themselves motivate the
assessment questions on which EMAP resource groups will focus their indicator development.   In
general, these values concern structural or functional aspects of the resource that contribute to
society's quality  of life by providing food,  fiber, clean water and air, aesthetic experience, recreation,
and desired animal  and plant communities.  Forest ecosystems, for example,  are valued for timber
production, wildlife habitat,  food, water storage, erosion control, and aesthetics.  Wetland ecosystems
moderate downstream flooding,  improve water quality,  control erosion, and  provide breeding, shelter,
and feeding habitat for aquatic  and terrestrial wildlife.

Assessment questions that will drive a resource group's indicator development usually should be
formulated in terms of the resource's status since questions about associated trends follow
immediately.  Specific examples of potential EMAP  status questions already have been discussed in
Section 2.4.  It is important that these questions are the actual assessment questions that are driving
the resource group's  indicator development, and not simply questions that can be addressed using
EMAP data.  For example,  many fundamental  EMAP assessment questions may involve multivariate
indicators of resource condition.  Although assessment questions could be formulated for each of the
component variables  of these indices, such  questions per se are not driving the resource group's
indicator development process.  Depending  on the sophistication of the resource group's intended
clients,  initial assessment questions may be reasonably generic. Such assessment questions,
however,  eventually must be refined to identify assessment endpoints clearly.  Table 4-1 summarizes
the environmental values that have been identified by each EMAP resource group  to date, and
definitions for these values are provided  in Table  4-2.  These values show a strong overlap, reflecting
commonality in the perceptions  of key issues among these groups.  This commonality  highlights the
need for integration and coordination  among the EMAP resource groups  to ensure that all important
information is collected  efficiently and in  a form that will facilitate completion of multiresource
assessments, as discussed in Section 8.  This initial list of societal values will change  over time and
be refined as EMAP evolves.  In addition, values common to multiple resource groups will be
identified, which will facilitate the integration of information across resource groups.

4.2  Identifying Expected Stressors
 In addition to monitoring status and trends of ecological condition,  EMAP's data also will be used in a
 diagnostic mode to identify plausible causes of adverse or subnominal conditions.  When adverse
 conditions are detected, the question that will inevitably follow is why? Although EMAP is not

                              EMAP Indicator Development Strategy
                                               23   ,

-------
Table 4-1.      Environmental Values Selected  by Different EMAP Resource Groups.
EMAP Resource Group
Environmental  Values3
Agroecosystems


Arid Ecosystems


Estuaries


Forests


Great Lakes

Surface Waters

Wetlands


Landscapes
Sustainability  of agricultural commodities,  quality of air, water
and soil, biological integrity

Biological integrity, harvestable productivity, sustainability,
aesthetics

Ecological  integrity,  harvestable productivity, fishability,  trophic
condition, aesthetics

Biological integrity, harvestable productivity, sustainability,
aesthetics

Biological integrity, fishability, trophic condition, aesthetics

Biological integrity, fishability, trophic condition

Biological integrity, harvestable productivity, hydrologic
function, water quality improvement

Distribution of patterns of communities  and ecosystems,
sustainability,  biological integrity
* These values direct the selection of condition indicators.
                               EMAP Indicator Development Strategy
                                                24

-------
Table 4-2.    Definitions of Selected Environmental  Values by EMAP Resource Monitoring  and
              Research Groups.  These values reflect Resource  Group thinking as of early
              1993.  Definitions  have evolved since that time.
 Environmental Values3   Definition (EMAP Resource Group)
 Biological Integrity
  Ecological Integrity
  Sustainability
  Air, Soil, and
  Water Quality

  Trophic Condition
  Water Quality
  Improvement

  Hydrologic Function
  Supply of
  Commodities

  Fishability
  Harvestable
  Productivity


  Aesthetics
The composition, structure, and functional organization of the biotic components
of the ecosystem at the genetic, species, community, ecosystem, and landscape
levels of organization that constitutes an integrated and adaptive community of
organisms.  (Agroecosystems, Arid Ecosystems)

The ability to support and maintain a balanced, integrated, adaptive community of
organisms having a species composition, diversity, and functional organization
comparable to that of natural habitats in the region.   (Great Lakes)

"A balanced, integrated,  adaptive community of organisms having a species
composition, diversity and functional organization  comparable to that of natural
habitat in  the region" (Karr and Dudley  1981). (Surface Waters, Wetlands)

A balanced, integrated, adaptive,  ecosystem of communities, populations, and
organisms having a composition, diversity, functional organization,  functional
attributes, and physical environments comparable  to that  of natural ecosystems in
the region.  (Estuaries)

The ability to maintain the desired biological integrity. Components of
sustainability include the extent of the resource (at multiple levels of resolution),
successional dynamics  in biological integrity, and  maintenance of forcing
functions.  (Arid Ecosystems)

The physical and chemical condition of these natural  resources.
(Agroecosystems)                                                          .

A classification based on actual or potential primary production, dependent on the
ambient levels and turnover rates of nutrient salts. (Great Lakes)

Algal and  macrophyte abundance and water clarity comparable to natural systems
of the region. (Surface Waters)

Ability to  assimilate nutrients, trap sediments, or otherwise reduce downstream
pollutant loads.  (Wetlands)

The natural  water flow pattern necessary for the sustainability of the wetland and
its function, e.g., flood conveyance, water storage capacity, and shoreline
protection.  (Wetlands)

The ability to produce adequate amounts of food  and fiber for human needs of
sufficient  quality to satisfy market standards.  (Agroecosystems)

The presence of catchable fish that are safe to eat.   (Estuaries, Great Lakes,
Surface Waters)

Quantity or quality of any service or product that  the resource provides society
(e.g., commercial timber, wildlife, recreation, and  food production).  (Arid
Ecosystems, Estuaries, Forests, Wetlands)

Attributes that affect the public's perception  of their environment.  (Arid
Ecosystems)                    :
  "These values direct the selection of condition indicators.
  bSome resource groups have not yet adopted the definitions provided in this table.
                                 EMAP Indicator Development Strategy
                                                   25

-------
designed  as a cause and effect research program, it must anticipate how major stressors are likely to
affect the indicators that it will monitor (e.g., Rapport et al. 1985).  The starting point for assembling
stressor information is the listing of the major problems and stressors currently impacting or
threatening  the resource.  Major environmental  stressors that should be considered by EMAP
resource groups when developing  indicators include:

              reduction, loss, or fragmentation  of critical habitat,
      •        reduced food resources due to pest control programs or other causes,
      •        introduction of exotic species (including domesticated livestock and
              recreational wildlife, e.g., fishery stocking  programs),
      •        global warming,
      •        alterations to regional hydrologic cycles.
      •        chemical pollution of air, soil/sediment, and water,
      •        increased sediment and nutrients loadings to surface waters (including
              wetlands and estuaries), and
      •        over-harvesting regional flora or fauna.

Certain indicators  reflect external stresses or pressures that affect ecological resource condition.
Generally, these indicators are measured or estimated by other programs or agencies rather than by
the EMAP resource groups.  Collating and analyzing these auxiliary data  also will be used to develop
indicators for EMAP; however, considerable  effort may be needed to assemble these data in the
proper format. These stressor indicators are most often anthropogenic (e.g., pesticide  applications,
human  population densities, livestock grazing pressures, atmospheric deposition,  emissions of
atmospheric pollutants, applications of fertilizers or other nutrients,  numbers of fishing and hunting
permits, and numbers of discharge permits), but they also include natural forcing functions, such as
precipitation or solar radiation, that in turn may be affected by anthropogenic factors (e.g., global
climate change) or indirectly by the ecological  resources themselves.

Explicitly defining  potential  stressors serves  to increase the  relevance of the selected condition
indicators to current and future environmental concerns.

4.3  Identifying Potential Indicators

The process of identifying potential condition indicators that can be conceptually  and functionally
linked to the resource's environmental values requires  a broad scientific perspective obtained through
both detailed literature reviews and interactions with scientists conducting relevant research.  This is

                              EMAP Indicator Development Strategy
                                               26

-------
a continuing process, however, since scientific and technological advances will generate new
research indicators or improve the feasibility of previously rejected indicators .(see Section 5.).

For consumptive and non-consumptive use values, the very process of formulating explicit
assessment questions often identifies  the potential indicators with which the  resource group should
be concerned.  In terms of the risk assessment paradigm, this situation is equivalent to the
assessment endpoint and the measurement endpoint being the same.  Consider, for example, the
following fisheries assessment questions

             What proportion of lakes in the Northeast United, States have game fish
             present (i.e.,  >  0)?
             What proportion of streams and rivers in Pacific Northwest currently
             support reproducing  salmonid  populations?
             What proportion of lakes in the Northeast United States show evidence of
             degraded fish habitat?
             What proportion of the nearshore Great Lakes do not support game fish
             populations whose median body size is greater than or equal to legal size
             limits?
             What proportion of the nearshore Great Lakes do not support forage fish
             populations of acceptable quality to desired game fish?
             What proportion of estuarine area in the Virginian Province demonstrate
             high incidence  (e.g., > 1%) of external fish pathologies?
             What proportion of estuarine area in the Gulf  of Mexico has fish with body
             burdens of contaminant  X that exceed FDA action limits?

For these particular  questions,  it is fairly obvious what appropriate condition  indicators might be.

When identifying  potential  indicators for values related  to ecological  structure and function, however,
objective identification of potential condition indicators generally  requires well-developed conceptual
models of the resource  of  concern.

4.3.1  Developing Conceptual Models of Resources' Structure and Function
 Each resource group should identify or develop a suite of conceptual resource models to guide,
 corroborate,  and document their indicator selection for values pertaining to the" resource's ecological
 structure and function; in some  cases, these models even guide and document the formulation of

                              EMAP Indicator Development Strategy
                                              27

-------
 assessment questions. These models should identify the structural components of the resource, the
 interactions between these components, inputs from and outputs to surrounding  resources, and the
 important factors and stressors that determine the resource's ecological operation and sustainability.
 Conceptual resource models can be constructed  at many scales, ranging from simple population or
 community models to complex ecosystem models identifying full complements of ecosystem
 structural and functional attributes.  Each of these approaches may be useful, and should be
 developed or reviewed  as appropriate.  Conceptual resource models can be formulated  and  .
 documented in a variety of different formats including: "box and arrow" diagrams (Figures 4-1 and 4-
 2), kinematic graphs (Figure 4-3), and graphic and matrix representations (Figure 4-4 and Table 4-3).
 Components or processes that fulfill-key or central functions in these models  obviously might be
 better indicators of the  resource's overall condition than those components  or processes  that occupy
 peripheral  positions.  Figure 4-1 presents one possible conceptual model for identifying key indicators
 (and assessment questions) regarding the ecological organization of arid ecosystems.  Figure 4-2
 displays one of the conceptual  models that EMAP-Agroecosystems  considered in deciding that
 nematodes were an appropriate indicator of the biological integrity of agroecosystem  soil community.
 The predominance  of nematodes in the structure  and function of this model corroborates the decision
 made  by EMAP-Agroecosystems to' pursue evaluation of this indicator.

 Conceptual resource models also should consider the temporal and spatial  dynamics  of the resource
 at multiple scales because information from  different scales can result in different conclusions about
 resource condition (Wiens 1989). Developing such models is an extremely  important  exercise that is
 required to substantiate the choice  of a particular  indicator.  For example, annual wood increment
 can be linked directly to forest productivity and can be incorporated  into a conceptual  model.  Data
 can be readily obtained at the temporal  scale appropriate to EMAP. Soil microbial  respiration, on the
 other hand, is more difficult to link to forest productivity,  and is fraught with  interpretation  problems at
 differing temporal and spatial scales of interest in  EMAP.  Figure 4-3 displays a conceptual model of
 successional dynamics  that EMAP-Arid  Ecosystems has used to organize its  process  for selecting
 indicators of ecological  sustainability and for expected trends in the resource's vegetative biodiversity.
 On the other hand,  Figure 4-4 presents  a graphical conceptual model  of the functional biodiversity of
 aquatic macroinvertebrates as function of stream  order and associated landscapes.  Clearly, this
 model would have significant ramifications to potential sample designs, data interpretation, and index
 development.       '•"•••-                                           •

Conceptual resource models also should identify how major stressors of the resource  are expected to
Impact its structure  and function.  For example, Table 4.3 presents a conceptual  model for the Great
 Lakes' fish communities in matrix format. The EMAP-Great Lakes group believes this matrix shows

                             EMAP Indicator Development Strategy
                                              28

-------
Figure 4-1.      Conceptual model for the trophic structure of arid  ecosystems.  Redrawn  from
                   Evenari et al. (1986).
          POOR RELIABLE RESOURCES
                                                                       RICH EPISODIC RESOURCES
      arido-acttv8 arthropod
      predators (scorpions,
      ants & spiders)
     detrithrorous Invertebrates
     (isapods, beetles, termitos,
     snails)
                        detritus
                                              large vertebrate predator
                                              (raptors, snakes &
                                              carnivorous mammals)
                                                    t
     small vertebrate
     omntvores (rodents
     & lizards)
ephemeral & migratory
iruecthrores (birds etc.)
                                                  (ants)
      lichens & algae
                 DEW & FOG
arido-acttve plants
(xerophytes & succulents)
                                             ephemeral & migratory
                                             folivorous insecte
                                             (caterpillars & locust)
                                                                                       ephemaral plants
                                                                                       (annuals & geophytes)
                                                                                       TOPSOIL MOISTURE
                                                                        RAIN
                                     EMAP  Indicator Development Strategy
                                                           29

-------
Figure 4-2.    Conceptual models for the trophic structure of the soil community  in
              agroecosystems. Adapted from Moore and de Ruiter (1991).
                            EMAP Indicator Development Strategy
                                          '; 30

-------
Figure 4-3.     Conceptual  model  for the vegetative  succession in arid  ecosystems.   Adapted
                 from Grover and Musick (1990).
Perennial Grass
i

t
Perennial
Grass/
Mesqulto
1


Mesqutta/
Perennial
Grasses
i


Coppice Dune
Formation with
Deflation zones
                                                  Aridlfloatton
                                                 Livestock Grazing
                                                  Seed source
                                                Shrub Competition
                                               Select Livestock Grazing
                                                 Fire Suppression
                                               Rodent/Rabbit Activity
                                                     ecus Plant
                                                Nutrient Distribution
                                                 Wind Erosion
                                              Local/Regional Climate
                           Deep Coarse Soils No CaCO
                                                                 Shallow Coarse Soils
                                                                 CaCO near surface
                                 EMAP Indicator Development  Strategy
                                                     31

-------
I
ra
                                                 .CM  O>
i
o
 o
1
a
 ra
 $
£
 to
I
 o
                                                         CO
 o —^
 « «
   S«-i
   0
1
 g
I
 O  ra
 3'
 §
 CD
8
§     §     5     OJ

(%) indui aiBinoRjBd
                                                 .CM O>
                                                     co
                                                 -CM
                                                                                                 o>
                                                                                                 1
                                                                                                 w
                                                                                                 o>

                                                                                                 a.
                                                                                                 jo

                                                                                                             LJU
                                                           32

-------
 (0
 0)
 ra

O

I
 c

 w
 o
S
•o
 o
 o
_o

 Q.
 2
4^
T3
 re
 (0
 CO


 CD

 re
£
 (0
I
 a  .
 22

 E
 O  0.
 C  (Q
 O  T3
O  <
co
•<*
JQ
 re

"o

Q)

Q.
CO CD
CL >,
"0 CD
•i 5!
Q- >
'•§
3
•a
2
a.

CD
CD"

1
CO
CO
CO
E
0
ha
CD
S?-
1


*£•
•a i
O c
o fc
CD O
CO 0
I
E
o
O
CO
CL
c:
.O
'•*-•
•a
o
0
o
0.
2
1-






























'c



"O C
*C "^
0 §
12

.0
r"
a.
2
0
.£?
2
-i—*
3

„
§
- ., **5

I 1 1 ^
o>.*: S3 o .co
CD O S -Q CO









1








1




T3
'CJ
CD
Q.



'E
0
"CO
CO




O
a.
2
0
O)
O
"3
o ™
-b 2
CD
to en -t;
.8 1 el
flllftl
a. o 73 O Q. Ill CD





0)
2
CD
T3
0








JZ
.D)


^
12^ "o
11
1 8




T3
2
CD
a.




.0
Q.
2
0
CO
CD
5



„
CO
centrarchid:
cyprinids









D)



,

D)
ic
2
1
T3
'c
'o.
o
'2
0)
Q.
!


2
1
8




.0
,0.
2
i3






























"c
8
D)

T3
]c
Q.


O

Q.
2
^-i
2
CD
CL
>>
X
 TO
.2
 2
•4-*
CO
.»-«
 CD

 Q.
JO

 I
Q


1
 o
T3
                                                                                                                                               LLJ
                                                                            33

-------
how the resource's biodiversity and fishability is related to eutrophication, which is one of the major
stressors of Great Lakes ecosystems.

Developing conceptual models that describe a resources's  ecological components is one of the most
important aspects of the indicator development  process; it also may be one of the most difficult.  The
process of developing  models and of determining the degree of complexity and the model focus (e.g.,
ecological structure or function) is the responsibility of each EMAP resource group. Conceptual
models should serve as reference points both for identifying indicators needed to assess the
condition of ecological  resources  and for guiding data analyses and construction of multivariate
indicator indices.
4.3.2 Selecting  Research Indicators

Indicators selected for development may be either indicators that were previously suspended pending
new findings or completely new indicators that will augment or substitute for existing indicators.  New
indicators  should consider recent advances in environmental sciences and monitoring technologies
(e.g., new methods of  remote sensing) and consolidate insights gained through analysis of data
collected by EMAP and other research programs.  Several  approaches may be used to identify
research indicators for further evaluation and development.  These include workshops, systematic
literature reviews to identify potential improvements to the current suite of indicators, attendance at
major conferences, solicitation of involvement of the scientific  community through presentations and
published  articles, and continued  personal contact with leading scientists researching  relevant topics.
Whatever the mechanism, the key to continual  replenishment  of indicators used  in EMAP is active
and effective communication  with the scientific community.

Periodic technical workshops  should be conducted  by each EMAP resource group to update their
suite of research indicators and to assess existing indicators for completeness with respect to the
environmental values,  assessment questions, and conceptual  resource models that have been
previously  identified for their  resource.  Particular emphasis should be devoted to identifying gaps in
their suite of indicators and to generating ideas for new indicators. To facilitate effective workshops,
the Technical Director of each EMAP resource  group should prepare information  summarizing the
status of indicator development for distribution prior to the workshop.  At a minimum, this information
should include: 1) EMAP's current indicator development strategy, 2) the resource group's current
research plan, 3) examples of results to date, and 4) the current research plans  of other EMAP
resource groups  working in similar media.
                              EMAP Indicator Development Strategy
                                             .34

-------
Although some discrimination at this stage may be appropriate, it is better to identify too many
research indicators than to overlook potentially useful ones.  It is not necessary to provide substantial
amounts of evidence about indicator behavior nor is it necessary to conduct peer reviews of the
selection process and documentation.  Critical evaluation of selected indicators during the second
phase of the development process will eliminate inappropriate  indicators or suspend  their evaluation.
The main issue is whether each indicator appears to address environmental values and assessment
questions that have been  identified  by the resource groups and whether the indicator appears to be
quantifiable and applicable for a national monitoring program.
                              EMAP Indicator Development Strategy
                                               35

-------
                                     5.  Indicator Evaluation

EMAP's indicator evaluation  is a multistage process designed to compile and analyze a body of data
and information  needed to select a set of core indicators for EMAP's national implementation  from a
host of research indicators identified by the indicator formulation  process.  Each step in this process
involves two stages:  1) an assessment of the sufficiency of the available data to support the
indicator's evaluation and 2) screening the indicator based on the selection criteria.  This process
results in one of three outcomes: 1) acceptance for consideration at the next stage of evaluation,  2)
temporary suspension of consideration due to insufficient data, technology, time or resources for
proper evaluation, or 3) rejection for failure to satisfy one or more of the selection  criteria.  The latter
two outcomes may lead to new approaches for collecting, synthesizing, and analyzing data.  Many
research indicators are expected to be placed in a state of suspended evaluation,  to be revived at
some future date when evidence, time, and resources are sufficient to thoroughly  evaluate them.
During the initial stages of evaluation,  an indicator will be rejected only if it fails certain critical  criteria,
and the indicator's  performance can not  be expected to improve  over the next decade.  In contrast,
during the final stages of evaluation, it is likely that a much  higher proportion of indicators will  either
advance or be rejected, rather than suspended, because  sufficient data  regarding  their overall
performance will be available to make a firm decision.

5.1 Indicator Performance  Criteria

For EMAP's indicator development  process to be scientifically defensible, each resource group must
use a consistent procedure and set of criteria to judge each  potential indicator.  The use of clearly
defined criteria not only increases the  objectivity  and  consistency of indicator evaluations, but  also
provides a framework for documenting EMAP's indicator evaluation process.  Although certain
decisions made  in the evaluation process will be subjective,  the goal of the evaluation procedure  is to
provide an appropriate amount of information  so that an independent evaluation by peer reviewers
could follow the logic of the original decisions.

EMAP's indicator evaluation  is guided  by a set of criteria that initially were identified by Messer
(1990). Later the same year, these  criteria were modified based on an EMAP Indicator Strategy
Workshop (Hunsaker and Carpenter 1990). These modifications were incorporated  into EMAP's
initial indicator development  strategy (Knapp et al. 1991; Olsen 1992).  These  criteria  address a
variety of issues and are identified as  being either critical or desirable with -respect to EMAP's  overall
programmatic objectives (see Table 5-1).  Kelly and Harwell (1990) proposed similar criteria for
indicators of ecosystem recovery. The type and  amount of data  needed to evaluate any specific

                              EMAP  Indicator Development Strategy
                                               36

-------
Table 5-1.
Indicator Evaluation  Criteria
 Criteria
                 Definition of Criteria
 Essential Criteria
 Unambiguously  Interpretable


 Ecologically responsive

 Index period stability

 Amendable to synoptic survey

 High signal-to-noise ratio


 Nominal-subnominal criteria

 Minimal environmental impact
 Desirable Criteria
 Available method

 Historical record

 Retrospective
 Anticipatory

 Cost effective
 New information
                 Relates unambiguously to a recognized environmental value or
                 assessment question and quantitatively conveys the same information for
                 most resource sampling units within a regional resource class.
                 Responds to stressors and to changes  in the condition of the resource
                 across most pertinent habitats within a regional resource class.
                 Exhibits low measurement error and temporal variation during an index
                 period.
                 Can be quantified by synoptic monitoring or by cost-effective automated
                 monitoring.
                 Possesses sufficiently high signal strength (when compared to natural
                 annual or seasonal variation) to allow detection of ecologically significant
                 changes within a reasonable time frame.
                 Possesses documented or identifiable thresholds or  patterns of trends
                 that identifies the nominal or subnominal condition of the resource.
                 Sampling produces minimal environmental impact.
                 Possesses a generally accepted and standardized measurement method
                 that can be applied on a regional scale.
                 Has existing historical data base or one can be generated from accessible
                 data sources.
                 Relates to past conditions by way of retrospective analyses.
                 Provides an early warning of widespread  changes in ecological condition
                 or processes.
                 Has low incremental cost relative to its information.
                 Provides new  information; does not merely duplicate data already
                 collected by cooperating agencies.
                                 EMAP Indicator Development1 Strategy
                                                   37

-------
indicator with respect to these criteria vary greatly.  In some cases, best professional judgement
based on theoretical and logistical considerations can be used to evaluate an indicator's
programmatic utility.  In other cases, existing  data from the literature and other sources can be used.
If such data are not available,  EMAP resource groups will conduct pilot research and  regional
demonstration projects to generate the information needed to evaluate the indicators in question.
In general, the critical  criteria are those considered essential for satisfying EMAP's programmatic
objectives. The first two critical criteria (unambiguous interpretability and ecological responsiveness)
should be the initial focus of all indicator evaluations.  Although these two criteria continue to be
important in the evaluation at subsequent phases, other critical criteria relating  to the  utility and
feasibility of sampling  the indicator (index period stability, amenable to synoptic survey, low year-to-
year variation, minimal environmental  impact) increase in importance.  Tables  5-2 and 5-3 exemplify
the use of these critical criteria during the early and later stages of an indicator's evaluation.  During
this evaluation process,'it is not crucial that each indicator satisfies each criterion completely; rather,
there should  be a reason to believe that each criterion can be satisfied when the appropriate data
and models for detailed analyses are assembled  in  later stages of the evaluation process.
Indicators that fulfill some or all of the desirable criteria have obvious advantages over those that do
not.  These advantages may include an improved assessment of associations  between stresses and
ecological conditions (the third objective of EMAP), an increased  timespan over which the indicator
can be quantified, higher information value per  unit cost, greater ease of implementation,  or special  .
value for early detection of widespread ecological changes.  Desirable criteria  should  be applied to
assist in distinguishing  among alternative indicators  for the same assessment question.

5.2 Phases  of Indicator Evaluation

There are four types of evaluations that must be performed to evaluate  a research indicator:

        •      evaluation  of the indicator's conceptual soundness,
        •      evaluation  of the feasibility of implementing the indicator in a routine monitoring and
               assessment  program (current and future),
        •      evaluation  of the indicator's statistical  behavior, and
               evaluation  of the indicator's utility in resource assessments.

Although all evaluations of EMAP indicators  begin with a thorough evaluation of their  conceptual
soundness (Section 5.2.1), the remaining phases of the process may be conducted either in the
sequence indicated above or with overlap  among phases.  For example, in pilot research projects
discussed below (Section 5.3.2), resource groups may simultaneously  evaluate both the feasibility of

                              EMAP Indicator Development Strategy
                                                38

-------
            .C  CD T3 .£  ^
            Zl  '-a S '..  to

                       (0
                       o
      te




             co).>^-.0c2   °


O

1

I
UI
to
ca
re
      •s.
      0)

      1
          "wESwwSScc^
          sllgs&lgsil

period.

Small to
                                    (O
                                         E
                                         Q.
                                         C1
                                         O
                                         'co

                                         •e
                                                                         I
                                                                         ^—•
                                                                         CO
                                                                         0)


                                                                         D.
                                                                        _O
, a>
 I?
£
re
ill
               ii

                    1

                       li
                       g e


                                                                         O
                                                                        111
          o m
          g.co
 CD
 "5.

 CO

 2
                'S   en
                '-5   .c

       5T
       o:

                                •
                               g

                               I
                            >. .
                            a)T3o
s
CD
              CM
                   CO
                            1O
                               tO
                                      39

-------
on
 ra
&


t
*
 o
(0
 03


?




i
2
0)
         01

         I
        CO
        f
               -s
             II
             Q -Q
                 a
                      I  8
                      8g£
                            •

^  °
•g  £3
2  «» ~°

                                          o

                                          *=
                                          <1)
              • -  ai
             If l
                   llll

                              co"
             Of
             S -
             w
                : -o

              I-Sl
                      O CO
                      £ to

                      .£ "o
                              j   s
                      U}
                      c
                  c   ^

                  is   i -|
                  s   as


                                 .
              .2ro^'jr!S^ta)(B:5o—


                            '

                                 8
                                 e
                                 Q.
                                 C
                                 O
                                 'eo

                                 "8
                                       U)
                                           
                                 73
                                                       3
                                                                                                  O)
                                                                                                  0)

                                                                                                  2
                                                                                                 •~


                                                                                                 0)


                                                                                                 Q.
                                                                                                  
-------
Table 5-4.     Indicator Status Sheet that Tracks Moving Condition Indicators from Research
               to Implementation Mode.
Name of indicator:
Developed by: EMAP resource group •
Status: Core, Research, Suspended, or Rejected
Date:
Formulation Criteria '
Associated environmental value(s)
Associated assessment questions
• Principal stressors
Conceptual link(s) to value(s); for values related to
ecological structure and function, conceptual
resource models are available
References: "•"'*.'••••
Research Evaluation Criteria - Pilot Level ' • •"
Me.thods available and fairly standardized
Measurable on all sample units
• Responsive, to stressors or to changes in the
condition of the resource at large
• Literature reviewed for data appropriate for
characterizing variance components
Analysis of intra-annual variety sufficient to define
index period
Analysis of within and among site variability
sufficient to define the number of sample units.
References:
Research Evaluation Criteria - Demonstration Level
• ' Statistical power analyses for trend detection
• Correlates with regional patterns of stressors
• Analysis of logistical and economical feasibility for
regional monitoring
Analysis within the context of annual statistical
report and regional assessment
References:
Core Implementation Criteria
• Quantitative relationship to assessment endpoints
or values established
•• Correlates with regional patterns of stressors and
can be used for diagnostic assessments
Responds quantitatively to changes in the condition
of the resource category at large
• Nominal-subnominal criteria established
EMAP DQO's satisfied
Standardized methods established
References:
' References may be direct documentation, internal EMAP reports, or published papers that indicate how and when each
criterion is satisfied.
                              EMAP Indicator' Development  Strategy
                                               41      !

-------
indicators and important aspects of their statistical variability  (e.g., within- and between-site variance).
Resource groups should document the current status of each indicator using the indicator status
sheet presented in Table 5-4.

5.2.1  Conceptual Considerations

The most important criteria in evaluating a potential  EMAP indicator is its ability to provide
information that addresses environmental values and assessment questions  identified during the
indicator selection phase.  Each  potential indicator must be evaluated  in terms of how it contributes in
forming a complete suite of indicators  that can address all environmental  values and assessment
questions that have been identified by the resource  group.  Particular  attention must be  paid to avoid
developing multiple indicators that provide redundant information on the same environmental values
and assessment questions.  Multiple indicators for the same fundamental assessment question
should be evaluated or developed only when there is no a priori reason  to assume that  one indicator
will perform better than one or more possible alternative indicators or when a single indicator is not
expected to be  responsive to all major stressors that might impact the resource.  Resource groups
also should evaluate  potential  indicators in terms of how they link or integrate with the needs of other
EMAP resource groups.

For an indicator to be useful in regional assessments, it must convey  the same quantitative
information  concerning the condition of the resource regardless of where it was measured within the
region.   For example,  if a particular score for an indicator of forest primary productivity corresponds
to a subnominal condition in one locality, then, for the purpose of interpreting the indicator's regional
cumulative distribution, that score should correspond to a subnominal  condition elsewhere in the
region.   To ensure this characteristic,  indicators  often will need to be normalized or  otherwise
adjusted with respect  to underlying environmental variables that determine limits for their expected
values.  For example, forests that occur at different elevations and latitudes  and on  different soil
types generally would be expected to demonstrate different  levels of nominal primary productivity.  If
such an indicator is not adjusted for this expectation, its regional cumulative distribution  would be of
limited value for assessing the condition  of forest resources.  Resource  groups should identify and
evaluate these  normalizing variables as an integral  part of evaluating  an indicator's  conceptual
soundness.

These considerations can be evaluated best by critical analysis  of each resource group's  conceptual
 models  and of  those conceptual models that link several resource classes together. Decisions made
                               EMAP Indicator Development Strategy
                                                42

-------
during this phase are based on best professional judgment;  no field activities or data analyses are
needed for this stage of the evaluation process.

5.2.2 Operational  Monitoring Considerations

Having confirmed the conceptual  linkage between a potential indicator and one or more of the its
formal assessment  questions,  the resource group must consider practical implications that affect an
indicator's possible  utility. These considerations  include:

               Are available field methods appropriate to the anticipated skill level of field crews?
               What field sampling methods should be used for sample collection, and are there
               operational constraints on the use of these methods?
               What are the best analytical techniques for measuring the indicator?
               How precisely  can the indicator be measured?
               What are the beginning and ending  times for site access and for the indicator's
               expected index period?
               What time of the year should measurements be made (definition of the index period)?
               Within what location of the resource sampling  unit should the indicator be measured
               (identification of the index sampling  site/area)?
               Can sufficient measurements of the  indicator (e.g., numbers of target organisms) be
               collected, given the sampling design, operational constraints, and  the need to
               minimize the environmental impacts of the sampling process?
               What are expected sampling costs for transportation,  equipment, and personnel?
               What are the data-processing  demands in terms of sample  identification, shipping,
               and archiving?
       •       What are the data-processing  and reporting  demands  in  terms of data management
               arid statistical support?

Many operational monitoring questions regarding sampling methods and analytical techniques can be
evaluated based on the literature.  However, others can be evaluated  only by conducting pilot
research or regional demonstration  projects.

5.2.3 Statistical Evaluation of Indicators
To evaluate an indicator's statistical behavior, resource groups must first explicitly identify those
sources of variance that must be quantified to evaluate its performance. In general, these sources qf

                              EMAP Indicator  Development Strategy
                                              43

-------
variance include the indicator's 1) temporal variability within and between potential index periods, 2)
spatial variability within and between sampling units, 3) annual variability between index periods, and
4) response variability to stressors.  Specific research questions that should be answered for each
indicator include:

               What is the indicator's spatial variability among concurrent samples  collected  at
               different locations within a resource sampling unit?
               What is the indicator's spatial variability among non-concurrent samples collected
               within the index window but from different ecological  resource sampling  units  within
               the same  region?
               What is the indicator's spatial variability among samples collected within the index
               period in different regions during the same year?
               What is the indicator's temporal variability among samples collected from the  same
               ecological resource sampling unit,  during the same year and at the  same locations
               but during different  potential index periods?
               What is the indicator's temporal variability among samples collected within the same
               region, during the same index period, but spanning a number of years?
        •       What is the indicator's variability among concurrent samples collected at different
               sampling units that  are impacted to differing degrees by stressors identified during  the
               indicator  formulation phase?

Similar questions and similar information are required  for indices composed of these indicators.
These questions are intended to ascertain whether the condition of the resource of concern can be
estimated with the  precision required to meet data quality objectives using  indicators measured
during an index period and whether the indicator's responses to major stressors or to changing
ecological condition of the resource can  be detected and quantified within a reasonable time frame.
An example of partitioning the variance into its component parts for Vermont lakes  is shown  in Table
5-5.  Both spatial (population - among lakes, index - within lake) and temporal (year - among years)
variability are considered in evaluating an indicator.

5.2.4  Evaluation of Indicators for Resource Assessments
 EMAP will not implement regional or national monitoring of any indicator unless it is expected to
 convey meaningful and useful information to decision makers,  environmental scientists, or the public
 at large (see acceptance criteria in Table 5-3).  In order to evaluate an indicator's ability to convey
 information,  each resource group should  address a number of specific research questions including:

                               EMAP Indicator Development  Strategy
                                               44

-------
Table 5-5.     Summary of Estimates of Components  of Variance for Secchi Disk
               Transparency  (SD), Chlorophyll-a (Chl-a), and Total Phosphorus  (TP) Derived
               from the Vermont Lake Monitoring Database3.
VARIANCE SOURCE
a2 lake (Population Variance)
a2 year (Year Variance)
a4 ,ake.year (Lake-Year Interaction)
o* ,ndex (Index Variance)
a2 meas (Measurement Variance, a component of
Index Variance)
VARIANCE ESTIMATES
SD
4.203
0.007
0.628
0.660
0.191
Chl-a
0.229
0.002
0.178
0.230
0.111
TP
0.412
0.018
0.089"

a Grand means: SD - 4.82 m, Chl-a = 10.0 //g/L, and TP = 11.7 jug/L. Both Chl-a and TP were
 transformed before calculating variance components.
b Only the aggregate of a1 ,ake.year and a* index could be calculated because single samples were obtained each
 year.
                             EMAP Indicator Development Strategy
                                              45

-------
               How will field measurements be standardized  with respect to underlying
               environmental  constants that determine limits  for their numerical values to ensue that
               indicators convey the same quantitative information concerning the condition  of all
               resource units within a region?
               How will multiple field  measurements be combined  into indicator indices that  estimate
               the ecological condition of resources of concern in  regional assessments?
               How will the indicator contribute to defining the proportion of the ecological resource
               considered  to be degrading, improving, or impacted?
               How will the indicators  aid in identifying the likely causes for  patterns and trends  in
               ecological condition?
       •       How will the monitoring results be summarized in annual statistical summaries?
               How will the monitoring results from the indicator be used in  EMAP's annual
               statistical summaries?

5.2.4.1  Identification  of Nominal-Subnominal  Criteria

To achieve its goal, EMAP must present its results and assessments in a form that is readily
understood and meaningful to intended users. The concept of nominal-marginal-subnominal  is at the
crux of the assessment process and the use of EMAP information  in environmental decision  making
and management.  Nominal refers to the state of having  desirable  or acceptable  ecological condition.
Subnominal is its antonym  or the state of having undesirable  or unacceptable ecological condition.
Marginal condition exists when the nominal and subnominal criteria are not contiguous.  Although
many other terms could be used to describe this concept (i.e., desired-undesired,  acceptable-
unacceptable, attainment-nonattainment,  impaired-unimpaired, degraded-undegraded, stressed-
unstressed, polluted-pristine, etc.),  EMAP has adopted "nominal-subnominal" as  its preferred
terminology.

EMAP is committed to developing nominal-subnominal criteria for each of its condition indicators in
order to add  value to the decision making process and to address  the "so what question" associated
with interpreting the cumulative distribution that will be reported by EMAP resource groups in
statistical summaries.  The fact that 40 percent of the lakes in the  northeastern United States have
an Index of Biotic Integrity  (IBl) score less than 35 is not significant to decision makers.  What is
important is knowing that 40 percent of the lakes in the northeastern  United  States have an  IBl  score
less than 35 and that such a score implies subnominal biological integrity.
                              EMAP Indicator Development Strategy
                                               46

-------
There are at least two different approaches  for identifying nominal or subnominal conditions of
ecological resources.  The first approach involves  identifying  critical scores or thresholds along an
indicator's cumulative distribution that delineate different conditions of the resource.  Depending on
the amount  of information available and the assessment needs of potential clients, resource groups
may choose to identify either a single nominal-subnominal threshold or multiple thresholds.  In the
latter case,  one threshold could delineate resources in subnominal condition, whereas a second
threshold could identify resources in optimal condition.  Presumably, resources having indicator
scores in between these two values would be considered to be in marginal condition.  When
developing thresholds, however,  it is important to remember that criteria are value specific.  For
example, one would expect indicator scores for trophic condition  of inland surface waters, estuaries,
or the Great Lakes to have three distinct nominal-subnominal thresholds  depending on whether the
value of concern  is the resource's biological integrity, fishery  productivity, or aesthetic features.

For certain indicators (e.g., certain indices of biodiversity and landscape indicators), there may be
little consensus of what constitutes a nominal-subnominal | threshold, but there might be a general
consensus that a negative or positive trend  over time is indicative of change  in a resource's
condition.  In such cases, it may be sufficient to ^stablish a reasonable criteria for status estimates.
Nominal-subnominal criteria indicate an indicator can satisfy EMAP's trends DQO (i.e., detect a 20
percent change per decade) and convey relevant information to decision  makers and environmental
managers regarding the resource's nominal-subnominal condition. The establishment  of reasonable
criteria,  however, should not preclude an attempt to establish nominal-subnominal criteria for the
status estimates of EMAP condition  indicators.  Examples of nominal-subnominal criteria for various
biotic and abiotic condition indicators for the Great Lakes are listed in Table 5-6.
Expectations or definitions of nominal-subnominal  condition change as the resource itself changes
and  as more is learned about natural processes and system  functioning.   Five years ago, for
example, it would  have been easier to characterize the condition  of rangeland and rangeland
management practices than it is today because the paradigm previously used in  rangeland
management is changing.  As more information on the factors influencing  rangeland condition
becomes available, the classification of range condition indicators into  nominal-subnominal categories
will change.  This  evolution of scientific knowledge and societal expectations  must be factored into
the specification of nominal-subnominal scores.

Rapport (1989) lists three approaches  or criteria commonly used to assess ecosystem health:  1)
identification of systematic indicators of ecosystem functional and structural integrity, 2) measurement
of ecological sustainability  or resiliency (i.e., the ability of the system to handle stress loadings, either
natural or anthropogenic), and 3) an absence of detectable symptoms  of ecosystem disease or

                              EMAP  Indicator Development Strategy
                                              47

-------
Table 5-6.    Great Lakes Resource Group examples of nominal - subnominal criteria (ranges)
              established for the Laurentian Great Lakes.
Condition Indicator
Nominal Threshold
Source
Remarks
Fish Populations ' ' ' ~
Internal Pathology
External Pathology
Recruitment
< 2-4 % of fish
< 6-8% of fish
> 80-90%
AOC Listings, Literature
USFWS, Pers. Communication
L. Superior Objectives
Some Published
Little Info Published
100% ultimate goal
Trophic Status . ,_,--., , ,
Total Phosphorus
Chlorophyll a
TN/TP
SJO2/SRP
< 6.3 - 15 ug/L
< 2.5 - 4.3 ug/L
>29
> 93
ULRG, Literature
ULRG, Literature
Smith, Literature
Holm/Armstrong, Literature,
Well Documented
Well Documented
Lacks verification
Lacks verification
Fish Contaminants "
Aldrin/Dieldrin
DDT+metabolites
Endrin
Heptachlor+Epoxide
Lindane
Mirex
PCBs
Mercury
TCDD
0.3 ug/g1
1.0 ug/g2
0.3 ug/g1
0.3 ug/g1
0.3 ug/g3
absent4
0.1 ug/g5
0.5 ug/g6
absent3
IJC 1972-1987
IJC 1972-1987
IJC 1972-1987
IJC 1972-1987
IJC 1972-1987
IJC 1972-1987
IJC 1972-1987
IJC 1972-1987
IJC 1972-1987
edible portions
whole fish
edible portions
edible portions
edible portions
all aquatic organisms
whole fish
whole fish
all biota
1   Based on wet weight; for the protection of human consumers of fish
2   Based on wet weight; for the protection of fish-consuming aquatic birds
1   Based on wet weight; for the protection of human consumers of fish
1   Based on wet weight; for the protection of human consumers of fish
3   Based wet weight; for the protection of human consumers of fish; recommended change - total BHC
    isomers should not exceed 0.3 ug/g.
4   For the protection  of fish-consuming birds and animals
5   For the protection  of fish-consuming birds and animals
6   Based on wet weight; for the protection of aquatic life and fish-consuming birds
3   For the protection  of fish-consuming birds and animals
                                EMAP Indicator Development Strategy
                                                48

-------
 stress.  Thus, he defines ecological health as both the occurrence of certain attributes deemed to be present
in a sustainable  resource and the absence of conditions that result from known stressors or problems
affecting the resource.  Some of the approaches that can  be used effectively by EMAP resource groups to
identify  or develop nominal-subnominal  classifications include:

         Field calibration and hypothesis testing in sampling units that have been identified a priori
         as "good" and "bad" using criteria that are independent of the indicators being evaluated.
         Criteria developed using reference  sites, e.g., the long term ecological research sites
         (LTERs) supported by the National  Science Foundation.
    •     Laboratory methods such as bioassays for toxicity.
         Legislative mandates such as Food and Drug Administration  (FDA) action limits for fish
         contamination.
         External criteria such as regulatory  mandates, goals and scientifically defensible criteria.
         Common  sense and professional judgment.  Public meetings, hearings, focus groups,
         special interest groups, and other outreach approaches can be used to determine what
         the public considers to be nominal  and subnominal.  Professional  and scientific
         perspectives concerning what constitutes nominal or subnominal  conditions often can be
         obtained by Delphi methods  or Kepner-Tregoe analysis (Kepner and Tregoe 1981).
         Model based  criteria.  For example, to evaluate  the condition of soils with respect to
         credibility,  calculated soil indicators, S, could be compared to T scores of the Universal
         Soil Loss Equation (USLE) (Wischmeier and Smith 1978),  i.e.,
                                  A ^ T = Nominal
                                  A > T = Subnominal       '
       •  Similarly Habitat Suitability Indices  (HSI) (e.g., Fausch et al.  1988) could be used to
         identify nominal-subnominal  conditions for selected wildlife species as individual objects
         of concern and as surrogates for other biota.  Successional models could be used to
         identify expected trends in biodiversity under nominal  and  subnominal conditions.
         Retrospective analyses and  paleoecological analyses  can  be used to define a historical
         background or baseline condition that can compared with current  and future conditions.
         Specific experimental  research directed  at determining nominal-subnominal  scores and
          ranges.

 5.2.4.2 Condition-Stressor Associations                  :

 In addition to knowing what proportion of the resource is  in subnominal condition, it is also desirable to know
 that stressors are correlated with the same condition. Although such correlations do not prove causality,  they

                                  EMAP Indicator Development Strategy
                                                   49

-------
are an important diagnostic tool toward establishing  such relationships and for initiating detailed,  cause-and-
effect research.  Resource groups must plan these statistical analyses for each selected condition indicator
by 1) identifying non-EMAP stressor data bases that could be used for this purpose, 2) developing  selected
stressor indicators concomitant with condition indicators  based on specific testable hypotheses or
documented  conceptual  models,  and 3) using landscape indicators  as stressor indicators for certain classes
of stressors.

Although the utility of the first two approaches is intuitively obvious, the utilization  of landscape indicators for
condition-stressor associations may be less clear. The EMAP-Landscapes resource group will be developing
a number of landscape indicators to estimate the condition of spatially defined complexes of ecosystems and
other resource categories.  Initial analyses,  however, suggest that several of these landscape condition
indicators can be potential indicators of specific types of stress.  Besides the obvious stressor of habitat
reduction  and fragmentation, it appears to be possible to associate certain landscape indicators with such
stressors  as  regional hydrologic alterations, increased  sediment loadings, and  point and nonpoint chemical
pollution  sources.  The utility of landscape condition indicators  as surrogate stressor indicators for other
EMAP resource groups is a major area of research  planned  by EMAP-Landscapes.  If successful, this
approach  affords a very cost-effective way to perform certain analyses of condition-stressor associations.

5.2.4.3  Example Assessments

The annual statistical summary is a primary mechanism  for reporting on resource  condition.  Results from
each demonstration  study should be analyzed as proposed by its EMAP resource group and  reported in a
preliminary annual statistical summary to confirm the utility of the data. Although  the degree  of evaluation
possible  is extremely limited during the first year of data collection,  this evaluation will allow confirmation  of
the basic  rationale for including each indicator.  Subsequent  annual  statistical summaries will be increasingly
important for evaluating the ability of each indicator  to identify changes and trends in the status of ecological
resources.

Example assessments explore the types of data analysis, data presentation,  indicator responsiveness,  and
indicator variability that can be expected.  These assessments may  be conducted using either simulated  or
real data.  Conducting an example assessment is intended to assist in selecting among alternative indicators,
identifying redundant indicators,  identifying gaps in the suite of indicators, deciding how data  from each
indicator would  be used in EMAP for assessing status and trends, exploring  the ability of the indicator suite
to ascribe plausible  causes to observed patterns and trends  in the region's subnominal areas and refining the
conceptual model.  An EMAP resource group's  conceptual model, which  links  stressors with assessment
endpoints, should be used throughout the example assessment to guide  the analysis and interpretation  of

                                  EMAP Indicator  Development Strategy
                                                   50

-------
 both empirical and simulated data.  Example assessments should be conducted  for multiresource
 assessments to assist in the selection of indicators that aid integration of information across resources.

 5.3  Evaluation Methods

 One of the first steps  in indicator evaluation is to review and analyze existing data.  Existing data can permit
 a preliminary assessment of an indicator's  characteristics, such as stability during the index period,  and
 identify additional  data needs for further evaluation.

 In general, there are three principal sources of data used to quantitatively evaluate EMAP's indicators,  i.e.,

     1)    existing  data from the literature and compiled data bases,
    2)    data generated by limited-scale field pilot research projects, and
    3)    data generated by regional demonstration projects.

 Because the acquisition of these  data involve increasing costs, resource groups should utifize them in the
 sequence indicated. Although much of the existing data may not be appropriate  to the temporal and spatial
 scales required by EMAP, to the greatest degree possible,  an indicator's evaluation,  should be undertaken
 using such data.                                            .    ,  '•            "   <   "' =

 5.3.1 Existing Data and  Desk-Top  Studies                          •

 Ideally,  a large portion of an indicator's statistical behavior could be characterized and evaluated using
existing data (e.g., Table 5-5). Consequently,  before initiating any field research, resource groups should
determine what types of data currently exist that can be, used to address each of the research questions
defined  in Section 5.2.3.  Data sets from field studies or surveys that have sampled  resources more
intensively than is being considered for EMAP, for example, can be extremely valuable  for evaluating an
indicator's intra-annual  and spatial variability and, consequently, would be valuable for identifying potential
indicator index periods  and index locations, respectively.  Data from  long-term plot experiments  can be useful
to assess inter-annual  and local spatial variability (Franklin et al. 1990).  Although such  data would be of
obvious value for corroborating potential index periods and  index locations for an indicator, these data also
could be used  to establish the indicator's  responses to changes in the condition of the resource  (i.e., plot) at
large. Retrospective data also may be useful to characterize the natural  temporal variability of related
indicators.  For example,  sediment cores and tree ring chronologies could be useful  for evaluating research
indicators of  biological  integrity or water chemistry (using diatom assemblages) and of forest productivity,
respectively.  Finally, data on indicator scores at regional reference sites or data on  indicator responses

                                  EMAP Indicator Development Strategy
                                                  51

-------
Figure 5-1.   Illustration of the relationship  between power and magnitude of trends detectable and
            years of monitoring.  From left to right, curves are for trends of approximately 2%/yr,
            1.5%/yr, 1%/yr, and 0.5%/yr.
               POWER
               1.00


               0.80


               0.60


               0.40


               0.20


               0.00
                       2   4    6    8   10   12   14   16   18   20

                                          YEARS
                             EMAP Indicator Development Strategy
                                            52

-------
across stressor gradients can be useful in characterizing an indicator's variability under prescribed
environmental conditions.

Having obtained variance estimates for an indicator,  simulation models of its statistical behavior should be
constructed and analyzed.  These investigations should be used to estimate an indicator's preferred  index
period for sampling, the time needed for an indicator to detect changes of a specified magnitude,  and other
aspects of the indicator's cumulative distribution such as changes  in its cumulative distribution during
different index periods and  over longer time frames of several years.  Simulation results for longer time
periods can be input to statistical programs to determine minimum detectable trends in condition indicators
(e.g., 2 percent change per year in indicator score over 10 years).  For example, statistical power analysis
can be conducted to determine the time or duration required  to detect different rates of change in an
indicator for a given variance  structure (Figure 5-1).  If only limited data are available,  it may be possible to
use Monte Carlo simulation techniques or simple process models to generate hypothetical data with
reasonable  spatial  and temporal  variability that can be employed for the same purposes.

5.3.2 Pilot Research Projects

Pilot research projects are designed  to evaluate any aspect of an  indicator's overall performance  and
programmatic utility except for the indicator's  ability to provide preliminary regional estimates  of the
resource's condition. Consequently,  sampling sites for pilot research projects need not be selected
according to probability sample protocols unless pilot project objectives require it.  The objectives might be
achieved  more efficiently using alternative research designs.

One of the  primary purposes  of these investigations  is to corroborate the regional monitoring  feasibility of the
research indicators. These studies  answer the when, where, how, and what questions associated with the
indicators that appear to be tractable for a regional monitoring program.  These projects also can be used to
evaluate the relative merits of alternative sampling and  analytical methods (including  considerations  of
sampling  and analytical error).  Finally, from a logistical point of view, pilot research projects are used to
develop general  protocols for sample acquisition, handling, and archiving.

The second principal purpose of pilot research projects is to generate data needed to answer the questions
that were identified in Section 5.2.3  regarding the indicators'  statistical behavior.  In general,  data from these
studies are used to begin the characterization of any or all of an indicator's components of variance  with the
exception of its regional spatial variability  that must be addressed  via a regional demonstration project
(described  below in Section 5.3.3).   In some  cases,  questions identified in Section 5.2.3 should be refined to
                                   EMAP Indicator Development Strategy
                                                    53

-------
address specifics of the indicator's statistical behavior.  Examples of subjects to emphasize in a pilot
research project include:

         Intensive temporal sampling to define the best boundaries for the index  period (e.g., fall turn-over in
         lakes, seasonal low water in wetlands) and to quantify the within-index period sampling variability.
    •     Extensive spatial sampling within a regional resource class to determine the value of data collected
         at index sampling sites relative to more intensively or randomly located sampling sites and to
         quantify  indicator variability within the index sampling area  (e.g., the central  basin of the lake)
         where permanently fixed monitoring sites cannot be established.
    •     Sampling along gradients from polluted to unpolluted  or impacted to natural  sites to 1) evaluate
         the responsiveness  of the indicator to stress,  2) aid in defining nominal and  subnominal
         classifications, and 3) evaluate the specificity  of the indicator to particular types of stress or change
         and the repeatability  of the indicator response in different regions or ecological  resource classes.

The optimal pilot research project design will depend on the specific  questions to  be addressed.  Two
examples from the EMAP-Estuaries resource group illustrate the types of studies that may be useful:
         Definition of the index period.  Levels of dissolved oxygen (DO) in estuaries are highly variable,
         yet DO also serves as an important condition and stressor indicator for assessing estuarine
         condition.  Therefore, field studies were conducted in 1990 to determine 1) the optimal boundaries
         for the summer index sampling period and 2) the utility of point-in-time measurements of DO.  At
         about 100 sites in the Virginian Province, three point-in-time  measurements  of DO were collected
         during three sampling intervals (early, mid-, and late summer).  Comparison of the DO cumulative
         distributions for the three periods provides information of the regional  stability of the DO indicator.
         In addition, DO was measured continuously at a subset of 30 sites, selected by experts as sites
         expected to experience problems with low DO.  These  continuous records will be used to both
         refine the index period and to evaluate the utility of point-in-time measurements as an indicator of
         the frequency, severity, and extent of low DO episodes.
         Indicator responsiveness  to stressors.  Using expert judgment, 24 sampling  sites were selected
         to reflect important gradients of both pollutant exposure (DO gradient) and habitat (salinity) within
         two geographic regions (latitudinal  gradient).  A variety of indicators (e.g., benthic biomass, species
         abundance) were sampled  at each site, three times during the summer index period.  Condition
         indicators that consistently  reflect the effects of pollutant gradients across a range of habitats and
         regions are obviously preferred and fulfill the  prime criterion for acceptance as EMAP core
         indicators (see Figure 5-2). In choosing  sampling sites  and regions for the pilot  research  project,
         resource groups should attempt to include the full range of conditions that are expected to be

                                  EMAP Indicator: Development  Strategy
                                                  54

-------
 I
I
o.
 S>
>
                                            1N3IQVUO
 0)
 o
'«*

 o
w
§
Q.
re
o
0)
O)
1
o
o
S"
Q.
H  .
Sf
CO O>
0) 0>
re o
1-
                                                                                                            O)
                                                                                                            I
                                                                                                            OT
I
1
Q
3
S
                                                                                                            111
V)
05
iZ
                                                        55

-------
               encountered during national implementation. Answers to some of the above questions (e.g., best index!
               period) may vary from region to region or among ecological resource classes or types.  If so, it may bel
               necessary to include multiple regions in the pilot research projects or to evaluate the indicator in regions thai
               are expected to represent the extreme conditions for the indicator. Multiregional pilot research projects!
               should be limited to investigations of issues that cannot be resolved within a single region and should be|
               designed to obtain only the minimum information needed  to complete the assessment.

               Table 5-2 illustrates decisions that resource groups should be able to make using the results of pilot research!
               projects.  To fulfill the second criterion in the center ("Accept if:") of Table 5-2, there must be quantitative]
               evidence that an indicator 1) responds to changing stressor levels, 2) responds to changes in the conditior
               of most resource classes, and 3) has signal-to-noise ratios both temporally and spatially stable enough I
               during the index period not to mask its responsiveness.  To fulfill criteria #3 and #6, data on the costs and j
               logistical  constraints associated with  sampling  must be compiled and evaluated.

               5.3.3  Regional Demonstration Projects

               The primary objective of regional demonstration projects is to provide preliminary estimates of a resource's
               condition over a standard Federal region  for one or more resource classes. Although these  projects are I
               used to confirm the results of site-specific pilot research projects on regional scales, they are designed to|
               evaluate whether data collected on these indicators can be interpreted regionally. These projects confirm the
               feasibility and utility of research indicators at regional scales over a broad range of conditions. These I
               projects focus only on indicators whose interpretability  (i.e., conceptual or mathematical relationship to
               identified assessment questions), sampling procedures, and methods have been well established. Regional
               demonstration projects are designed to further characterize  the indicator's among-site and between-year
               variability. Data obtained from regional demonstration projects are used to evaluate the indicator's ability 1

    •      be broadly applicable (e.g., are different  indicators needed for large lake, small lakes,
          rivers, and streams?),
    •      detect regional trends,
    •      integrate spatial and temporal influences and stressors,
    •      be used in diagnostic assessments of resource condition, and
    •      convey meaningful results to decision-makers  in the form of annual statistical summaries
          and periodic assessments.

A key function of these studies is the determination of the sampling intensity needed to estimate regional
status and trends  in resource condition and to detect associations among  regional  patterns of ecological

                                  EMAP Indicator Development Strategy
                                                   56

-------
condition and anthropogenic stresses.  From a logistical viewpoint, these projects establish EMAP's
infrastructure for conducting  regional monitoring activities through the field implementation activities that are
necessary to conduct regional  demonstrations.

Table 5-3 illustrates key decisions that should  be able to be made using results of regional demonstration
projects. Criteria #1, -#3, and #4 in the acceptance column of Table  5-3 are absolutely critical tests for any
indicator to attain core status.  Figure 5-3 illustrates the temporal dynamics of a regional Index of Biotic
Integrity. These data suggest  that although a  summer index period would appear appropriate for this
indicator, a late spring index period  would not. Although criterion #2 (responsiveness to changing ecological
condition of the resource class) is critical for an indicator's long-term utility, in many cases full evaluation of
this criterion  may require regional monitoring of a suite of indicators for several years.  In such cases, an
indicator could be accepted and nominated for core status based on well documented  theoretical grounds.
Full evaluation of this criteria then would  be completed during its reevaluation as a core indicator.
Each indicator must be evaluated for all pertinent resources classes.  Although rejection  of an indicator for
one resource class should not affect decisions regarding the utility of that indicator for other resource
classes, indicators  that are effective for multiple resource classes are desirable and will always be  sought
(e.g., failure of nutrient concentration indicators to adequately characterize the condition of one agro-
ecosystem resource class should not result in  rejection  of this indicator in other agroecosystem  resource
classes).

5.4  Identification  of Indicators for Implementation

The  final phase in the indicator evaluation process is the selection of core indicators for routine regional and
national monitoring. In this stage of the indicator development  process, EMAP resource groups nominate
selected research indicators for long-term (i.e., 10-15 years  minimum)  core implementation.   These indicators
are then subjected to a formal programmatic and peer review process.  There must be substantial  and
documented  evidence that each proposed core indicator satisfies all of the critical indicator criteria
summarized  in Table 5-1 and that it does not provide merely redundant information to that provided by other
core indicators. Indicator status sheets (Table 5-4) should  be maintained and updated to summarize the
results  of the evaluation of each indicator and  to document all sources of information  used or generated in
the evaluation  of research indicators.  These indicator status sheets should  be included as an integral part of
the annual research plans and  the more comprehensive 5-year research and monitoring plans of each
resource group.

Although multiple  research indicators of the same or closely related  assessment endpoints may have been
evaluated by a resource  group, in general, only one of such multiple indicators will be implemented as a core

                                  EMAP Indicator Development Strategy
                                                   57

-------
Q)
M
m
i
(O
CO
en
1
i
»
 p
 
-------
indicator.  There are,  however,  two situations when multiple indicators of the same assessment endpoint
could be elevated to core status.  The first of these occurs when the indicators under consideration  are
                                                         F              '        *
multivariate or index indicators  that depend on the same set ;of field measurements. The second situation
occurs when  no one indicator is expected to respond to all of the resource's major stressors.  In this case,
however, it must be remembered  that each core indicator also must be responsive to changes in the
condition of the resource at large.  The fact that an indicator may not appear to be directly responsive to
particular stressors may be irrelevant  since it should be indirectly responsive to those stressors via its
responses to changes in the condition of other components of the resource.
                                 EMAP Indicator Development Strategy
                                                 59

-------
                                     6. Indicator Implementation

The third phase in EMAP's indicator development process,  is the national implementation  of the core
indicators that have undergone formal programmatic review and approval.  During this stage, EMAP resource
groups

    •     monitor core indicators nationally,
         prepare annual statistical summaries,  and
         prepare periodic resource assessment in conjunction with the EMAP-Assessment and
         Reporting coordination group.

In the implementation phase, there are explicit environment values and assessment questions associated
with each resource.  Conceptual models relate the condition indicators or indices to assessment endpoints
and to associated stressors.  Indicator variances have been quantified and partitioned into spatial and
temporal components.  Based on these variance estimates, both status estimation and trend detection data
quality objectives have been satisfied for each core indicator.  Regional demonstration  projects have been
conducted to document the feasibility of monitoring these indicators at regional and national scales.
Nominal-subnominal  ranges have been established  for at least selected environmental  values  and the
process is continuing for other segments  of society and societal values.

The annual statistical summary is the primary product of each resource group during this stage of the
program.  These reports  should summarize the current regional cumulative distribution  of  each core indicator
and identify the indicator's associated nominal-subnominal  thresholds.  Results of trends analyses for some
appropriate time window  should also be summarized in these reports.  If stressor indicators also have been
implemented, results of correlation analyses between condition and stressor indicator scores likewise should
be presented.  EMAP information will be shared with other monitoring  programs at the federal, regional,  and
state levels.
                                 EMAP Indicator Development Strategy
                                                  60

-------
                                       7. Indicator Reevaluation

This phase of the indicator development process is actually an ongoing activity that begins upon initial
implementation of the monitoring of core indicators at regional and national spatial  scales.  Core indicators
are periodically reevaluated to confirm their ability to  detect changes and  identify trends in the condition of
ecological resources  and to ensure that appropriate advances in technology and information  have been
considered for incorporation.  In this phase, it is important that EMAP balances continuity of methods to
maximize its capability to detect trends with procedures for refining or replacing  indicators that fail to perform
optimally. This phase is designed to critically review  the performance of core  indicators through time,
evaluate alternative indicators to address  emerging issues and; inadequate core indicator performance, add
new indicators as deemed desirable,  and  substitute superior indicators for less adequate core indicators.

Scientific advances and technological innovations  obviously will  occur during EMAP's implementation,  arid
these advances  may  improve the precision, accuracy, representation, cost effectiveness, and overall
applicability  of EMAP indicators.  These advances may necessitate modifying  specific indicators,  replacing
indicators with others that provide improved or equivalent information at reduced cost,  or adding indicators
that address  important emerging  issues. To accommodate these changes,  it will be necessary to specify
appropriate procedures to reexamine indicator performance  and usefulness.  This section presents a
preliminary outline of a systematic approach  to indicator reevaluation that will ensure the use of the best
possible set of indicators  for achieving EMAP's objectives.  This approach  should be revised  and expanded
as EMAP matures and as modification of the core indicators  is considered.

An EMAP core indicator should be revised only after  a thorough reexamination indicates that it is clearly
necessary (i.e., when revision results in a significant improvement in the quality  of the  assessment of status
and trends of ecological condition, without  diminishing the continuity of the assessment record).  Once the
current  EMAP program has been evaluated and modification of a core indicator is appropriate, a plan for
transition must be developed for  programmatic peer review.  Proposed changes will not be implemented
unless approved  by the peer review process.  Having established the need to modify the set of indicators,
the primary objective  is to accomplish a smooth transition.  Continuity of EMAP's data  bases  is extremely
important for ensuring that trends or changes in condition can be detected.  Situations  that may  require
reexamination of the core indicators include the following scenarios:

         A newly identified indicator appears to be superior to an indicator currently in use for measuring an
         assessment endpoint or stressor.  The decision to replace the current  indicator with the new one
         and to discontinue  monitoring the current indicator must be made after obtaining  adequate ;
                                 EMAP Indicator Development Strategy
                                                  61

-------
         information to ensure continuity of the monitoring record and comparability  of the new assessments
         with those that have relied on the old indicator.
    •     The environmental conditions have changed such that underlying mechanisms are altered and
         previous linkages between the indicators and environmental values may not be representative of
         the existing  situation.
    •     An improved method promises to provide similar quality data at lower cost  or higher quality data at
         a similar cost.  The impact of using the improved method to monitor condition and to detect trends
         must be evaluated before replacing the original method to ensure that data quality actually equals
         or exceeds that available  using the current method.

Evolution of assessment methods may also  result in changes for data analysis, presentation, or evaluation in
annual statistical  summaries (see Paulsen et al. 1991) and assessments.  For example, a revised conceptual
model of the resource might suggest a new  index or a revised nominal-subnominal threshold for assessing
ecological condition for selected resource classes.  Although these changes will not directly affect the set of
indicators, they may result in an opportunity  for modifying  or adding core indicators.  Therefore, the impact of
these changes on the assessment process should be evaluated as early as possible to increase the ability of
the indicators to provide  needed information.

7.1 Reevaluation Procedures
The primary approaches for evaluating core indicators include routine review, evaluation of assessment
products,  and searches for new ideas for indicators.  Such approaches call for continual tracking of the
published literature and ongoing research programs to identify promising new information.

Once a new or revised  indicator or method is identified, the process to evaluate  that idea should proceed  as
outlined in Sections 4 and 5 above.  Implementation of this review process ensures that indicators or
methods cannot be revised or replaced without  1)  carefully conducting the evaluations needed to ensure
that the new indicator or method provides a meaningful improvement in EMAP's monitoring  or assessment
capabilities and 2)  quantifying the relationship between the new and old indicators or methods (i.e.,
calibrating the new indicator).

Evaluation of proposed changes to core indicators  must  be conducted in such a way to allow not only  the
evaluation of the relative  merits of the new and old indicators or methods  but also the mathematical  or
statistical relationship between the two.  This evaluation  requires that pilot and regional demonstration
projects be designed and conducted to test for comparability and relative  responsiveness of the two
indicators or methods under a range of conditions.  Field demonstrations  should be conducted to test

                                  EMAP Indicator Development Strategy
                                                   62

-------
alternative  indicators or methods in one to several regions for a number of years to verify the consistency of
that relationship.  Pilot and regional demonstration projects should be conducted to calibrate the relationship
between the old and new indicators (or measurement techniques), and both the old and new (or modified)
indicators will  be monitored  long enough to ensure comparability of the data sets from both indicators before
phasing out the old indicator.  Once the spatial and short-term temporal relationships  between the
alternatives are well established,  simultaneous collection of data for both indicators may be desirable for an
extended period of time at a limited number of sites to ensure the similarity of the relationships over an
extended time period.

Revision of core indicators requires the conduct of all assessments described  in Section 5.2.4 and that all
criteria for adoption of the changes are satisfied.  The advantages of new indicators or methods must be
significant and must represent improvements over existing indicators. These advantages must be well
documented, and the documentation of the research efforts (laboratory,  pilot research, and  regional
demonstration projects) must include the following information:

         Quantification  of the calibration between the old and new indicators or methods under the full range
         of conditions observed during EMAP's monitoring to date.
   •     Evaluation of how the proposed change in indicators or methods would affect the annual statistical
         summary and data interpretation  and integration  (may require recalculation  of indices).
                                  EMAP Indicator Development Strategy
                                                  63

-------
                               8.  Integration Among Resource Groups

As discussed in Section 2.3, eight broad ecological resource categories  have been defined within EMAP:
agroecosystems, arid ecosystems,  forests, estuaries,  Great Lakes, surface  waters, wetlands, and
landscapes.  Indicator development and implementation within each of these resource categories is the
responsibility of an individual EMAP resource group.  All ecological communities, ecosystems, and biomes
are open thermodynamic and material flow systems that can be sustained only by maintaining critical
intersystem connections and interactions. If EMAP is to achieve its programmatic goals, its indicator
development process, implementation, and assessments must reflect and embody this principle.
Consequently, indicator development and implementation of individual EMAP resource groups must be
integrated and coordinated.

This integration occurs at two levels: 1) during indicator selection to ensure that all important intergroup
linkages are considered and 2) during data interpretation. The second level of integration is beyond  the
scope of this document. Tasks relating to an integrated  interpretation  of EMAP's monitoring  results are the
responsibility of the EMAP-Assessment and Reporting coordination group.  This group recently published an
EMAP Assessment Framework document that discusses this data  interpretation process (EPA 1994).
However, the utility of each indicator for interpreting resource status and trends  is an  important consideration
in the indicator selection process. Thus, close cooperation between resource groups and EMAP-Assessment
and Reporting is essential.  Integrating monitoring  results among ecological resource  categories will enable
EMAP to address a wide range of issues including:

    •     source apportionment and diagnostic analyses across resource boundaries  (e.g., nonpoint  sources
         to surface waters),
    •     the status at a regional scale, encompassing all ecosystem types,
         the extent and magnitude  of environmental problems that impact multiple ecological resource
         categories,
    •     the effectiveness of regulatory  actions, and
         emerging  environmental problems or new assessment questions that EMAP can address.

The primary approach to achieving  an integrated set of indicators across all resource  groups is through
communication and information exchange.
8.1 Framework for Indicator Integration
                                 EMAP Indicator Development Strategy
                                                 64

-------
Integration of indicator development  and application of indicators across all EMAP resource groups involves
a number of factors including 1) maintaining inter-group communication and interaction to foster development
of indicators that will integrate ecosystem level information  among  different resource groups, 2) assimilating
new knowledge, 3) consistent interpretation  of critical  indicator development concepts, 4) consistent
collection  and  use of stressor indicator data, 5) collaboration in identifying special condition indicators that
integrate across EMAP resource groups (e.g., wide ranging or migratory organisms that use multiple habitats,
landscape  pattern  metrics, soil carbon, etc.), and 6) co-locating sampling units for special  studies.

EMAP's ability to diagnose plausible causes of changing trends will be facilitated  if the pathways of
interaction  between ecological resources are explicitly identified. EMAP needs to examine the stresses that
each ecological  resource receives  as a result  of processes or conditions in other resource categories.  For
example, the nutrient  balance of a lake may be highly dependent on nutrient fluxes in the surrounding forest.
Nutrient flux from the  forest may be  measured as a condition indicator in the forest, but as a stressor
indicator for the lake.   Such identification will help clarify the auxiliary  stressor  data requirements of each   ;
EMAP resource group and the level  of mutual assistance that  is needed to acquire such information.
Consistency in using auxiliary information will  improve the abilities  of all EMAP resource groups  to detect
spatial and temporal associations among condition and  stressor indicators and the natural and anthropogenic
stressors affecting them, particularly  intersystem problems  and issues.

Substantial consistency already exists among  EMAP resource groups in their definitions of environmental
values, formulation of assessment questions, and identification  of major stressors. This parallelism provides
EMAP with opportunities for identifying plausible causal relationships on large  regional scales.  For example,
if EMAP-Surface Waters detects that nutrients are significantly increasing  in streams across a broad region,
but data from  EMAP-Agroecosystems indicate no increase in nutrient export from agricultural lands, then
other nonpoint sources (e.g., atmospheric loadings) or point sources (e.g., discharges) may be responsible
for the observed trends in aquatic  nutrients. Conversely, noncomplementary data may indicate that the
conceptual models being used are inappropriate or incomplete and should be  reexamined.

EMAP resource groups proceed at different rates to implement the indicator development strategy.  The
differences in  phasing among EMAP resource groups contributes to the overall indicator development
because it allows the pioneering groups to pass on lessons learned to other groups.  This process began at
ah indicator strategy workshop in Las Vegas,  Nevada, in June 1990, and it should be maintained and
fostered in EMAP.  Circulation  of annual research  plans, communication of lessons learned, and informal
intergroup  discussions provides important information for efficient indicator development.
8.2  Types of Indicators that Integrate Across Ecological Resources

                                  EMAP Indicator Development Strategy
                                                  65

-------
EMAP has two general types of indicators: condition and stressor.  Although all indicators are one of these
two indicator types, many indicators also contribute to an integrative understanding of status and trends and
contribute to diagnostic evaluations.

8.2.1  Indicators Linking Resource Groups

Some indicators are an output from one resource category and an input to another (e.g., nitrogen in fertilizer
used in agroecosystems  and as subsequent runoff to surface waters and wetlands) and, therefore, interface
one EMAP resource group with another.  Such indicators measured using the EMAP monitoring design
generate data that can be used by more than one EMAP resource group.  For example, an index of soil
erosion measured by EMAP-Forests  or EMAP-Agroecosystems would provide EMAP-Wetlands, EMAP-
Surface Waters, and EMAP-Estuaries with an  indicator of the potential  inputs of sediment,  nutrients, or
pesticides.  Because soil and sediments represent sinks for chemical contaminants in all ecosystems, soil
and sediment contaminant  data can be used as important links between ecological resources.  For example,
soil and sediment contaminant measurements  would be of importance  not only  in evaluating forest status but
also in assessing potential  effects on aquatic receiving  systems.

8.2.2  Indicators Shared by Resource Groups

Common  or shared indicators are measured in multiple resource categories using similar techniques.
Examples include wildlife biomarkers,  landscape attributes, and commonly used metrics of population or
community status, such as relative species abundance.  By using consistent sampling and analysis tech-
niques in  all resource categories, interpretation of multiresource patterns in ecosystem status and trends is
facilitated.  Landscape-level indicators  (e.g., mosaic diversity, patch fractal dimensions) may be applicable for
many or all resource categories as measures of habitat quality or as surrogates for other indicators that are
more  difficult to measure (e.g., wildlife  density).  Biomarkers  (e.g., DNA alterations, cholinesterase levels) are
common indicators  that can be used as a metric of exposure  to metals or organic constituents, whether the
organism  is a plant, fish, or mammal.

A second  type of shared indicator are  migratory indicators.  Migratory  indicators provide quantitative
information on organisms that move across resource boundaries,  that is, from one resource category to
another and back again (e.g., honey bees, migratory birds, white-tailed deer). These indicators would be
expected  to reflect  changes in exposure or habitat in one or more ecological resources,  and in some cases
they might indicate  cumulative impacts in  several resource classes or categories within or outside a region.
Indicators that integrate the effects of ecological resource conditions in multiple  regions, resource classes, or
resource subclasses (e.g., some birds, amphibians, top carnivores)  might be particularly useful in detecting

                                 EMAP Indicator Development  Strategy
                                                  66

-------
the cumulative effects of changes in more than one resource category.  Observing  such indicators may lead
to the detection of stress pathways  that had not previously been recognized  (e.g., DDT and reduced
reproduction in raptors due to eggshell thinness).

8.3 Use of Conceptual  Models to Facilitate Integration

As discussed in Sections 3 and 4, conceptual models are an important tool for formalizing  possible relations
among indicators, assessment endpoints, and stressors and for identifying data or knowledge gaps that could
be filled through the selection  and development of additional indicators.  In a similar manner, conceptual
models also play a key role in identifying indicator-endpoint-stressor relationships and interactions across
resource groups.

Each EMAP resource group should  develop one or more conceptual models that emphasizes the major
inputs, outputs, and structural  and functional attributes of interest for the resource class. These resource-
specific conceptual  models then provide the basis for integrating needs and  results, (see Section 3 for the
framework).

As a first step toward  integration among resource groups, the individual models developed by each group will
be compared.  This process will 1) help to formalize expected relationships among  indicators in different
resource groups, 2) establish consistency in the ways that conceptual  resource models are developed and
used to aid in  indicator selection, 3) encourage  the identification  and use of linking, common, and migratory
indicators, 4) identify commonalities in methods and indicator use among resource  groups, and 5) ensure
that important  processes and linkages are considered within the EMAP monitoring  network.  Developing,
updating, and  revising the structural aspects and the inputs and outputs of the individual conceptual models
will form the framework for coordinating EMAP indicator development,  as outlined in Section  3.

8.4 Coordination  of Indicator Development Among Resource Groups

Five major tasks are planned to facilitate coordination and integration of the  indicator development process
among EMAP's resource groups:
    1.    Compile and cross reference lists of assessment endpoints,  environmental values,  stressors, and
         indicators proposed by each resource group to identify areas of similarity or commonality.
         Assessment endpoints and environmental problems in different resource categories are, in general,
         highly  interdependent and linked to common stressors.  Foliar damage,  fish loss, and estuarine
         eutrophication,  for example, can all be related to atmospheric deposition.  Compilation of the

                                 EMAP Indicator Development Strategy
                                                  67

-------
         proposed endpoints,  stressors, and indicators serves as the first step towards identifying areas of
         overlap or inconsistencies in approach among  resource groups. A preliminary listing of the
         environmental values identified by each  resource group is provided in Table 3-1.
   2.    Conduct workshops that include  non-EMAP scientists to identify linking  and stressor indicators  that
         EMAP deems necessary or has overlooked.  Interaction  matrices are commonly used to develop
         and chart linkages in computer or simulation models. This same method  can be used to identify
         possible linkages among resource groups for indicator development.  The conceptual models
         identify major inputs  and outputs for each resource group and provide a starting point for
         discussions.
   3.    Develop conceptual models that identify cross-resource linkages and relationships.
   4.    As appropriate, propose alternate assessment  endpoints and indicators that would provide
         information similar to that provided by the resource group but improve comparability with endpoints
         and indicators being  monitored by other resource groups.  For  example, sustaining biological
         integrity is an environmental value common to all resource groups. Thus, to the degree possible,
         biodiversity should be assessed  using similar assessment endpoints and  indicators in each group.
         Greater compatibility of indicators, endpoints, and stressor information used in the different
         resource  groups results in easier and more direct program integration and cross-resource analyses.
         Direct comparison of responses and effects among resource categories allows weight of evidence
         and process  of elimination  approaches for diagnosing possible causal factors and mechanisms,
         thereby providing greater confidence in EMAP's results.
   5.    For indicators selected by more than one group, examine and compare the proposed field sampling
         and measurement methods and  suggest modifications needed  to improve comparability  among
         groups.  Comparable methods and units are also important if comparisons are to be made across
         resource  groups. In some  cases, further research may be needed  to develop methods applicable
         to several resource categories.  For example, nutrient and pesticide exports from terrestrial
         systems are typically measured using different techniques and  expressed in different units
         compared to estimates of inputs of these constituents to aquatic systems.  Selection of the optimal
         approach for satisfying the  needs of both the terrestrial  and aquatic resource groups may require
         additional simulation analyses or field testing.

Efforts related to each of the above tasks will evolve continually, and the lists, matrices, and models will  be
updated as needed.

8.5 Problems  Associated with Differences in Spatial and Temporal  Scales
                                 EMAP Indicator Development Strategy
                                                  68

-------
For some stressor and exposure indicators, temporal displacement might be desirable; the observed
response may be displaced in time from the perturbation  that caused the response.  For example, soil
erosion indices measured during the spring period in agroecosystems and forests, combined with nutrient
export coefficients, might correspond better with estimates of summer chlorophyll  concentrations in lakes
than would export estimates for the  summer period.  Selection of the optimal  index period for indicator
measurements must consider, therefore, hypothesized stress-response relationships  and the potential
displacement  that might be required to associate a dependent response variable with an independent
stressor or exposure indicator.  Expert opinion,  obtained during workshops and peer review (see Sections
4.3.2, 7.1, and 8.1), will aid in evaluating the importance and effects of temporal displacement.

Spatial  displacement is more  difficult to evaluate and address.  Paired comparisons are generally used for
association, and  regression analysis is used to  relate dependent and independent  variables.  Indicators
linking multiple resource categories,  however, may riot be collocated.  Data analysis techniques are being
developed, therefore, to deal with non-collocated data, relying largely on aggregation of regional or
subregional  EMAP results before conducting diagnostic analyses. For example, aggregation of data by
subregions was used during the National Acid Precipitation Assessment Program  to identify a linear
relationship  between sulfate deposition and surface water sulfate concentrations.

Analysis of associations at regional or subregional scales is consistent with EMAP's design objectives for
determining  regional patterns  and trends in the  status of ecological resources.   Most environmental  analyses
to date, however, have focused on causal  relationships at local or site-specific scales.  Extension of these
techniques to  larger spatial  scales will require new perspectives and approaches for data aggregation and
analysis and, where possible,  the development  of new indicators better suited for  application  and inter-
pretation on regional scales.  The potential utility of regional-level analyses also hinges, in part, on the
degree  of intra- versus inter-regional indicator variability, and the spatial scale over which indicator values
and variability  are relatively  homogeneous.  Indicator integration both within and across ecological  resource
categories represents  a major challenge; however, the benefits  resulting from a successfully integrated
program are substantial. The process of integration will require considerable amounts  of cooperation,
communication, and coordination.
                                  EMAP Indicator Development Strategy
                                                  69

-------
                                       9. Concluding Remarks
9.1 Planned  Reviews

Each EMAP resource group will formally reevaluate its indicator suite every five years. Proposed revisions to
the EMAP core indicators and their associated justifications should be included in the 5-year plan.  These
proposed revisions will be peer reviewed along with the rest of the program.

The research plan, indicator status sheets, and indicator data  base will be updated at each evaluation stage
of proposed new indicators and  methods.  It will also be necessary to ensure that the documentation for
each of the new indicators identifies the reasons for the investigation (e.g., identified  gaps,  inadequate
precision of current indicators).  All information  developed through comparison of alternative methods should
also be summarized in these documents.

9.2 Evolving Process

Lists of environmental values, assessment questions, and major stressors are not static; they each must  be
reevaiuated periodically as new issues  emerge, environmental values shift, and experience is gained with the
use and interpretation of EMAP monitoring data.  In addition, unforeseen stressors may begin to operate on
ecological resources, or ecosystem relationships  may change.  Either of these circumstances could require
alterations to the suite of indicators in order for adequate monitoring  of changes in the status and trends  in
resource condition to continue.

Like the lists of values,  assessment questions,  and stressors,  conceptual models linking these components
should not be viewed as static.  Lists of environmental values and assessment questions should be reviewed
both internally and externally via workshops, annual budget plans, and 3- to 5-year research plans.
Likewise,  conceptual models that are used to identify and interpret research indicators for these assessment
needs should be reevaiuated  periodically for their utility, validity, and completeness.

The periodic review of core indicators must incorporate new and  evolving environmental issues and values,
questions, emerging  stressors and new knowledge about linkages among indicators  to ensure EMAP
continually provides information useful  in making decisions about environmental protection and management.
This includes identifying and developing indicators that relate  directly to those variables or factors used in the
decision-making  and resource management process.
                                  EMAP Indicator Development Strategy
                                                   70

-------
Sharing of information,  including information on EMAP indicators, with other programs will continue to expand
as EMAP evolves.  Improved information management, technology transfer and reporting linkages among
EMAP and other agency/organization  monitoring efforts will facilitate this sharing of information.
                                 EMAP Indicator Development Strategy
                                                 71

-------
                                              10. References


Barnthouse, L.W. 1992.  The role of models in ecological risk assessment: A 1990's perspective. Environmental
       Toxicology and Chemistry 11:1751-1760.

Evenari, M., I. Noy-Meir, and D.W. Goodall. 1986. Ecosystems of the World: Hot Deserts and Arid Shrublands.
       12B.  Amsterdam,  Netherlands: Elsevier.

Fausch, K.D., C.L. Hawkes,  and M.G.  Parsons. 1988. Models That Predict Standing Stock of Stream Fish From
       Habitat Variables. PNW-GTR-213:  U.S. Department of Agriculture,  U.S. Forest Service, Pacific  Northwest
       Research Station.

Franklin, J.F., C.S. Bledsoe, and J.T. Callahan. 1990. Contributions  of the long-term ecological  research
       program.  Bioscience 40:509-523.

Grover, H.D., and H.B. Musick. 1990. Shrubland encroachment  in southern  New Mexico, U.S.A.: an analysis of
       desertification  processes  in the American Southwest. Climate Change 16:165-190.

Hughes, R.M.  1989. Ecoregional  biological criteria. In: Water Quality Standards for the 21st Century,  147-151.
       EPA68C80006, EPA/600/D-89/167.  Washington, DC:  U.S. Environmental  Protection Agency, Office of
       Water.

Holland, A.F. 1990. Near Coastal Program Plan for 1990: Estuaries. EPA/600/4-90/033. Narragensett, Rl: U.S.
       Environmental Protection Agency, Office of Reasearch and Development, Environmental Research
       Laboratory.

Hunsaker, C.T., and D.E. Carpenter, eds. 1990. Ecological Indicators for the Environmental Monitoring  and
       Assessment Program. EPA/600/3-90/060. Research Triangle Park, NC:  U.S. Environmental Protection
       Agency, Office of Reasearch and Development,  Atmospheric Research and Exposure Assessment
       Laboratory.

Karr, J.R., D.D. Fausch, P.L. Angermeir, P.R. Yant, and I.J. Schlosser.  1986. Assessing Biological Integrity in
       Running Waters: A Method and Its Rationale. Special Publication No. 5. Champaign, IL:  Illinois  Natural
       History Survey.

Karr, J.R.  1991. Biological integrity: A long-neglected aspect of water resource management.  Ecological
       Applications 1:66-84.

Karr, J.R., and D.R. Dudley. 1981. Ecological perspective  on water quality goals. Environmental Management
       5:55-68.

Kelly, J.R., and M.A. Harwell. 1990. Indicators of ecosystem recovery. Environmental Management 14:527-545.

Kepner, C.H., and B.B. Tregoe. 1981. The New Rational Manager.  Princeton,  NJ:  Princeton Research Press.

Knapp, C.M., D.R. Marmorek, J.P. Baker,  K.W. Thornton,  J.M. Klopatek, and D.F. Charles. 1991. The Indicator
       Development  Strategy for the Environmental Monitoring and Assessment Program. EPA/600/3-91/023.
       Corvallis, OR: U.S. Environmental  Protection Agency, Environmental Research Laboratory.

McKenzie, D.H., D.E. Hyatt, and V.J. McDonald,  eds. 1992. Ecological Indicators.   Vol. 1 & 2.  New York, NY:
       Elsevier Applied Science.
                                     EMAP Indicator Development Strategy
                                                     72

-------
Messer, J.J. 1990.  EMAP indicator concepts.. In: Ecological Indicators for the Environmental Monitoring and
       Assessment Program, C.T. Hunsaker and D.E. Carpenter,  eds., 2-1 through 2-26. EPA/600/3-90/060.
       Research Triangle  Park, NC:  U.S. Environmental Protection Agency, Atmospheric Research and
       Exposure Assessment Laboratory.

Messer, J.J., R.A. Linthurst, and W.S. Overton.  1991. An EPA Program for monitoring ecological  status and
       trends. Environmental Monitoring and Assessment 17:67-78.

Minshall, G.W., K.W. Cummins, R.C. Petersen,  C.E. Gushing, D.A. Burns, J.R. Sedell, and R.L. Vannote.  1985.
       Developments in stream ecosystem  theory.  Canadian Journal of Fisheries and Aquatic Sciences
       42:1045-1055.

Moore, J.C., and P.C. de Ruiter. 1991. Temporal and spatial heterogeneity of trophic interactions within below-
       ground food webs.  Agriculture, Ecosystems, and Environment 34:371-397.

NAS (National  Academy of Sciences). 1975. Planning for Environmental Indices.  Washington, DC.: Planning
       Committee  on Environmental Indices to the Environmental  Studies Board.

Norton,  S.B., D.J. Rodier, J.H. Gentile, W.H. van der Schalie, W.P. Wood, and M.W. Slimak. 1992. A framework
       for ecological risk assessment at the EPA.  Environmental  Toxicology and Chemistry 11:1663-1672.

Olsen, A.R., ed. 1992.  The Indicator Development Strategy for the Environmental Monitoring and Assessment
       Program. EPA/600/3-91/023.   Corvallis,  OR: U.S. Environmental Protection Agency, Office of Research
       and Development,  Environmental Research Laboratory.

OMB (Office of Management and Budget). 1974. Standard Federal Regions.  OMB  Circular A-105.  Washington,
       DC:  OMB.  [April 4]

Ott, W.R. 1978. Environmental Indices:  Theory and Practice. Ann Arbor, Ml: Ann Arbor Science Publications.

Overton, W.S., D. White, and D.L. Stevens,  Jr.  1990. Design Report for EMAP (Environmental  Monitoring  and
       Assessment Program). EPA/600/3-91/053. Corvallis, OR:  U.S. Environmental Protection  Agency,
       Environmental Research Laboratory.

Paulsen, S.G., D.P. Larsen, P.R. Kaufmann, T.R. Whittier, J.R. Baker, D.B. Peck, J. McGue, D. Stevens, J.L
       Stoddard, R.M. Hughes, D. McMullen, J. Lazorchak, and W. Kinney  (with contributions  by: W.S.
       Overton, J. Pollard, D. Heggem, G. Collins, A. Selle,  M. Morrison,  C.B. Johnson, S. Thiele, R. Hjort, S.
       Tallent-Halsell,  K. Peres, S. Christie, and J. Mello). 1991. Surface Waters Monitoring and Research
       Strategy—Fiscal Year 1991. EPA/600/3-91/022. Corvallis, OR: U.S. Environmental  Protection Agency,
       Environmental Research Laboratory.

RAF (Risk Assessment Forum). 1992. Framework for Ecological Risk Assessment.  EPA/630/R-92/001.
       Washington,  DC: U.S. Environmental Protection  Agency.

Rapport, D.J.,  H.A. Regier, and T.C. Hutchiinson. 1985. Ecosystem behavior under stress. American Naturalist
        125:617-640.

Rapport, D.J. 1989. What constitutes ecosystem health? Perspectives in Biology and Medicine 33(1): 120-132.

Reilly, W.J. 1989. Measuring for environmental  results.   EPA Journal May/June:2-4.

Ryder, R.A., and S.R. Kerr. 1978.  The adult walleye in the percid community—A niche definition based on feeding
        behavior and food  specificity. In: Selected Coolwater Fishes of North America, R.L. Kendall, ed.,
        American Fisheries Society Special  Publication  11:39-51.

                                    EMAP Indicator Development Strategy
                                                     73

-------
SAB (Science Advisory Board). 1990. Evaluation of the Ecological Indicator Report for EMAP: Report of the
       Ecological Monitoring Subcommittee of the Ecological Processes and Effects Committee.  EPA-SAB-
       EPEC-91-001.  Washington,  DC: U.S. Environmental Protection Agency.

Suter, G.W. 1990. Endpoints for regional ecological risk assessments. Environmental Management 14:9-23.

Urquhart, N.S., W.S.  Overton, and D.S. Birkes. 1993. Comparing  sampling designs for monitoring ecological
       status and trends:  Impact of temporal patterns.  In: Statistics and the Environment, V. Barnett and K.F.
       Turkman, eds., 71-85.  New  York, NY: John Wiley and Sons.

U.S. Environmental Protection Agency,  1992.  Framework for Ecological Risk Assessment.  EPA/630/R-93/001.
       Risk Assessment Forum.  U.S. Environmental  Protection Agency.  Washington,  DC 20460.

Westman, W.E. 1985. Ecology, Impact Assessment, and Environmental Planning. New York, NY: Academic
       Press.

White,  D., A.J. Kimerling, and W.S. Overton. 1992. Cartographic and geometric  components of a global sampling
       design for environmental monitoring. Cartography and Geographic Information  Systems 19:5-22.

Wiens, J. 1989. Spatial scaling in  ecology.  Functional  Ecology 3:385-387.

Wischmeier, W.H., and D.D. Smith.  1978. Predicating  Rainfall Erosion Losses -A Guide To Conservation
       Planning. Agriculture Handbook No. 537, Washington, DC: U.S. Department of Agriculture.
                                    EMAP Indicator Development Strategy
                                                     74
                                                                &U.S. GOVERNMENT PRINTING OFFICE: 1994 - 550-OOI/OOWZ

-------