[From the U.S. Government Printing Office, www.gpo.gov]







               Biological Habitat Quality Indicators
               for Essential Fish Habitat
               Workshop Proceedings

               14-15 July 1997
               Charleston, South Carolina



               Edited by

               S. Ian Hartwell


























                       1%




                  ?'4?,5 Of
               U.S. Department of Commerce
               National Oceanic and Atmospheric Administration
               National Marine Fisheries Service

   SR          NOAA Technical Memorandum NMFS-F/SPO-32
   11          September 1998
   .A2
   N67
   no. 32






























































                         A copy of this report may be obtained from:

                         Habitat Conservation Office, F/HC
                         National Marine Fisheries Service, NOAA
                         1315 East-West Mghway
                         Silver Spring, MD 20910









                       Biological Habitat Quality Indicators
                       for Essential Fish Habitat
                       Workshop Proceedings


                       14-15 July 1997
                       Charleston, South Carolina


                       Edited by
                       S. Ian Hartwell
                       NOAA/NMFS
                       Habitat Conservation Office
                       1315 East-West Highway
                       Silver Spring, Maryland















                       NOAA Technical MemorandumNMFS-F/SPO-32
                       September 1998







                       U.S. Department of Commerce
                       A" -





                       William M. Daley, Secretary
                       National Oceanic and Atmospheric Administration
                       D. James Baker, Under Secretary for Oceans and Atmosphere
                       National Marine Fisheries Service
                       Rolland E. Schmitten, Assistant Administrator for Fisheries










                                                                Table of Contents


               Abstract     ..................................................................................       v


               Executive Summary           ....................................................................    vii

               Introduction     .............................................................................         1


               Index of Biotic Integrity Overview            ..................................................      5

               Workshop Objectives          ..................................................................     10

               Technical Presentations       ................................................................      12
                   Practical Application
                         Attaining Environmental Goals            .............................................    12
                         The Ohio EPA Biological Monitoring Program                  ..........................    14
                         Assessment of Watershed Development on Tidal Creeks                     ..............    49
                         Spatial Framework for EFH Data Collection and Analysis                     ............   54
                   Fish Indices
                         An Estuarine IBI for Chesapeake Bay              .....................................    59
                         Estuarine Biotic Integrity Index         .............................................    61
                         Ohio's Lake Erie and Lacustuary Monitoring Program                    ................    64
                   Benthic Indices
                         An Index of Benthic Condition            .............................................    77
                         A Benthic Index for Estuaries of the S.E. United States                ...............    80
                         Chesapeake Bay Benthic Community Restoration Goals                     ...............    87
                   Offshore Indices
                         Marine Biocriteria Survey Techniques             ....................................     90
                         The Benthic Response Index           ................................................     96

               Breakout Group Summaries             ..........................................................     98
                         Vegetated Habitats       ............................................................     98
                         Benthic Habitats      ...............................................................    100
                         Water Column Habitats          ......................................................    103
                         Synthesis    .......................................................................     105

               Workshop Consensus and Conclusions                 ............................................    114

               Follow-up Actions         .....................................................................    117

               Appendices
                         1        Workshop Agenda          ...................................................    120
                         2        List of Participants      ..................................................    122









                                                       Abstract


                   A national workshop to address development of biological indicators for habitat quality
            in Essential Fish Habitat (EFH) was held July 14-15, 1997 in Charleston, S.C. The workshop
            was attended by biologists from the National Marine Fisheries Service (NMFS) Office of
            Habitat Conservation, NMFS Science Centers, the National Ocean Service (NOS) Strategic
            Environmental Assessment Division and the Oceanic and Atmospheric Research (OAR) Sea
            Grant Program. The meeting included presentations by researchers from universities and
            Federal and state agencies who are performing bioindicator research and development in
            aquatic environments. These included projects in several benthic and pelagic estuarine habitats
            on the Atlantic and Gulf of Mexico coasts, coastal embayments, benthic habitats on the
            continental shelf in the Atlantic and Pacific Oceans, and rivers and open waters of the Great
            Lakes. In addition to ecological considerations, application of bioindicators to management
            needs, monitoring issues, and delineation of habitats into ecosystem units were addressed.
            Conceptual approaches for development of bioindicators of habitat quality for EFH,
            identification of current areas of research needs, and settings for potential pilot program
            initiation were developed. It was concluded that the Index of Biotic Integrity (IBI) approach
            will be useful by generating multimetric information to describe habitat quality in quantitative
            terms and for technical ecological assessment and research. Parameters for assessment metrics
            were developed for each of three general habitat types, vegetated, benthic, and pelagic. Areas
            requiring additional research include basic natural history information on species selected in
            the metric development process, quantification of their response to anthropogenic stress, and
            methods for delineating reference areas.
























                                                            v








                                                 Executive Summary

             Background
                    A major activity within the National Marine Fisheries Service (NMFS) is the
             implementation of the Essential Fish Habitat (EFH) requirements of the Magnuson-Stevens
             Fishery Conservation and Management Act of 1996. This legislation mandates that the regional
             Fishery Management Councils, in coordination with NOAA, amend each the 39 fishery
             management plans (FMPs) to include the best available information on habitat delineation for
             each of the approximately 600 managed species. The amended FMPs will include options and
             recommendations to minimize adverse effects on EFH and identify conservation and
             enhancement measures. These will include recommendations on activities or regulations that
             may impact water quality, so that NOAA can protect, conserve, restore and enhance essential
             habitats for each life stage of all managed species. The ultimate goal is to maintain the natural
             productivity of fish habitats at levels which will sustain populations at harvestable levels into
             the future.


                     A key requirement of the habitat assessment activities is an assessment of habitat
             quality. Habitat is defined as the combination of chemical, physical and biological components
             of the water and substrate in the local or regional ecosystem. The ultimate indicator of habitat
             quality is the response of the biological community to the interaction of stresses and resources
             available at a particular location and time frame. The biological community acts as the
             integrator of habitat quality. Coupled with habitat delineation, chemical analyses and physical
             characterization, biological indicators allow assessment of alteration of the environment
             including eutrophication, nonpoint source pollution, contamination, SAV loss, etc. Therefore,
             assessment of the condition of the biological community is an indicator of habitat quality, and
             can also be utilized to track preservation and/or restoration efforts. The value of biological
             criteria and biological assessment techniques has been demonstrated by their broad
             applicability not only to existing efforts to protect, restore, and manage aquatic resources, but
             in determining where management and restoration resources should be invested. Biological
             habitat quality indicators need to be developed for several types of marine environments to
             measure habitat quality in a variety of habitat types. The term biological integrity originates
             from the Federal Water Pollution Control Act amendments of 1972 and has remained a part of
             the subsequent reauthorizations. Efforts to construct a workable, practical definition of
             biological integrity have provided the supporting theory necessary for development of
             standardized measurement frameworks, techniques, and criteria for determining compliance
             with that goal. In 1981, Karr and Dudley defined biological integrity as "the ability of an
             aquatic ecosystem to support and maintain a balanced, integrated, adaptive community of
             organisms having a species composition, diversity, and functional organization comparable to
             that of the natural habitats of a region". This definition alludes directly to measurable
             characteristics of biological communities which are found in the least impacted habitats of a
             region. It was this definition and the underlying ecological theory which provided the
             fundamental basis for the development of numeric biological criteria in fresh water. The U.S.
             Environmental Protection Agency (EPA) essentially adopted this definition in the national

                                                              vii








                program guidance on biological criteria. Biological criteria and attendant monitoring and
                assessment designs provide a means to incorporate broader concepts of water resource integrity
                while preserving the appropriate roles of the traditional chemical, physical and toxicological
                approaches developed over the past three decades.

                       Biological evaluation of aquatic habitat integrity is made possible by monitoring
                aquatic communities directly. Community bioassessments differ from approaches which rely
                principally on target species or indicator organisms by utilizing the aggregated information
                across multi-species assemblages. The aggregation of key community attributes functions as an
                indication of the more complex ecosystem elements and processes which can not be measured
                directly or completely. At the same time information about individual species is not lost in the
                process and can be accessed at any time. Furthermore properly designed bioindicator systems
                extract ecologically relevant information and provide a synthesized, numerical result that can
                be understood by non-biologists.

                       Conceptually, an Index of Biotic Integrity (IBI) utilizes a set of categories which reflect
                important ecological parameters, for example, diversity, abundance and trophic function.
                Within each category a variety of metrics are generated. Diversity can be quantified by number
                of species, species richness or one of several diversity indices. Abundance can be assessed in
                terms of numbers of organisms or biomass. Functional metrics reflect interactions between
                community segments, for example predator prey ratios, density of predator species, etc. Other
                categories can be included, such as condition indices, prevalence of lesions, proportion of
                pollution tolerant species, etc. Each metric is a site-specific measurement which can be general
                or very specific. Individual metrics are given a rating score on a numerical scale which reflects
                its value relative to a reference value. The individual scores are then summed on a site-specific
                basis so sites can be compared to each other, or a reference site. An alternative approach to
                selecting and scoring the metrics has been to designate reference sites a-priori based on other
                factors such as demonstrated lack of chemical contaminants, eutrophication, physical
                disturbance etc. An array of candidate metrics are then calculated and a final subset are
                selected based on multivariate statistical evaluation of the data. This approach allows
                application of IBI assessment to habitats in which functional relationships of resident organisms
                are not fully understood, due to high complexity or lack of knowledge.

                       There are a variety of important technical questions that have to be addressed before
                this approach can be widely employed in marine environments. Many of these problems have
                been solved for fresh water environments. Some states utilize IBI monitoring for habitat
                quality assessment while others have integrated IBIs into water quality monitoring as a
                regulatory tool for enforcing water quality permits. Marine environments are more complex
                than streams, requiring the development of different approaches within the IBI framework. A
                variety of pilot projects in marine environments have been initiated which have experimented
                with different methods and procedures.




                                                               viii










            Technical Presentations
                   The NMFS Bioindicator Workshop was organized to itemize and begin to address these
            parameters, establish a consensus of scientific approach to development of biological habitat
            quality indicators in EFH, and identify research and monitoring needs for future projects. The
            focus was directed toward practical applications of bioindicators in marine systems and
            research needs to support this development. Subsequent discussion groups addressed which
            biological parameters would be practical and meaningful to measure in each of three broad
            categories of habitat (benthic, pelagic and vegetated).

                   Technical presentations included a conceptual overview of the IBI approach in a variety
            of marine systems, as well as an extensive evaluation of the practical application of
            bioindicators to statewide water quality monitoring and regulation in the Great Lakes region.
            In addition to water quality regulatory permit and enforcement activity, state agencies in Ohio
            have recognized the utility of bioindicators in the implementation of non-Clean Water Act
            management activities such as endangered species protection, targeted fisheries management,
            hydro-modification and wetland dredge and fill permit evaluations.

                   Shallow tidal creeks serve as conduits through which many pollutants enter estuaries.
            Creek sediments act as a repository for toxic chemicals and other contaminants. It is largely
            unknown how effective wetland management policies and programs are at protecting tidal
            creek habitats, or how to restore degraded creek habitats. A South Carolina Marine Resources
            Research Institute study has initiated development of a data base to develop the information
            needed to characterize the ecological values, identify major pollution threats associated with
            watershed development, assess the cumulative impacts and develop environmental quality
            criteria for sustaining nursery functions of tidal creeks and associated marsh habitats Results
            indicate that monitoring efforts for tidal creeks should focus on the upper reaches of primary
            tidal creeks and should include measures of the health of resident organisms, water and
            sediment quality, and selected population and community parameters of resident species.

                   Habitat delineation methods and data base development for IBI derivation are
            compatible with current data base and GIS activities currently under way between NMFS and
            the National Ocean Service (NOS) in response to implementation of the Magnuson-Stevens
            Act. The primary data layers currently in place are estuarine salinity zones and USGS
            Hydrologic Cataloging Units. Additional data for coastal and offshore spatial units and EPA
            river reach files are being included. A complete database exists only for the contiguous states
            at the present time.

                   Estuarine and marine habitats are more complicated than freshwater streams due to
            their larger scale, and diverse biological and physical components, including a wide
            phylogenetic diversity of biota, highly transient species, strong physical and chemical gradients
            in water and sediment quality, and a strong interaction between the pelagic and benthic
            communities. Nevertheless, the basic principles of IBI development can be applied to these
            systems. Estuarine fish bioindicators have been, or are being developed, in Connecticut,

                                                           ix








                Massachusetts, Chesapeake Bay, North Carolina, Florida and Texas. Investigations on the
                transferability of fish community bioindicator metrics for submerged aquatic vegetation (SAV)
                habitats developed for Cape Cod estuaries and tested in Chesapeake Bay, and from Chesapeake
                Bay pelagic habitats to coastal embayments have been instructive. The degree of modification
                to the metrics that was necessary to adapt the systems to different regions was relatively
                straight forward.

                        Fish and benthic invertebrate IBIs have been developed in freshwater environments to
                assess transitional zones going from lentic to lotic habitats (termed lacustuaries) and for near
                shore open-water habitats of the Great Lakes, analogous to estuaries and coastal zone habitats.
                The bioindicator systems have been demonstrated to be capable of quantitatively tracking
                habitat quality and are responsive to habitat quality changes resulting from watershed and
                riparian area management activities.

                        In some, but not all locations, benthic invertebrate bioassessment schemes have adapted
                somewhat different approaches than those utilized for fish community assessment. The benthic
                indicator development projects have employed complex mathematical schemes to develop
                metrics, due to the more complex and less well understood biological communities associated
                with benthic invertebrate communities. Current development projects in the New York/New
                Jersey harbors, the Virginian province Chesapeake Bay, SE Atlantic, and Gulf of Mexico rely
                heavily on the EPA Environmental Monitoring and Assessment Program (EMAP) data.
                Chemical contamination data has been used extensively to guide definition of reference sites.

                        Coastal benthic efforts on the Atlantic and Pacific continental shelves have taken
                divergent approaches from estuarine studies due to the more diffuse nature of impacts in off-
                shore habitats. However, gradients of habitat degradation can be identified and quantified. A
                great deal more development and research will be necessary to address the myriad of habitats
                present in off-shore areas.

                Metric Development
                        Three discussion groups were formed, for the purpose of coming to consensus on an
                array of biological metrics which would be practical and meaningful to measure in each of
                three broad habitat types (benthic, pelagic, vegetated). Important attributes of metrics included
                consideration of ecological relevance, practicality, and demonstrated relationship to
                anthropogenic degradation of aquatic habitats. The vegetated habitat category included
                submerged aquatic vegetation (vascular plants and algae), emergent wetlands, mangrove and
                kelp habitats. The benthic habitat category included soft (unconsolidated sediment), hard
                (surfaces to which benthic organisms can attach) and live bottom substrates (physical structure
                of the habitat was composed of, or built by, oysters, coral or benthic assemblages with
                significant three-dimensional relief). The water column habitat included the open water column
                habitats of freshwater streams, estuaries, near shore and coastal waters. A total of 36 potential
                metrics in four categories (diversity, abundance, function and condition) were enumerated.
                There was considerable overlap between metrics in the three habitat types in the diversity,

                                                                 x








              abundance and condition categories. It is instructive that there was very little overlap in
              functional metrics. Functional roles of a species in a habitat is much more site specific than
              other parameters. Overall, the metrics used in current programs do not cover as wide a range
              as the potential metrics considered in the break-out groups. The range and specificity of
              metrics utilized in fish IBI projects are greater than those used in benthic invertebrate projects.

              Conclusions
                     Based on knowledge gained from preliminary studies, the IBI approach will be useful
              for assessing habitat quality in two primary ways: it brings together multimetric information to
              describe the quality of the biological community in simple, yet quantitative terms, and can be
              used for technical ecological assessment or to formulate research hypotheses for testing. The
              approach was specifically designed to assess environmental harm resulting from anthropogenic
              stressors. In addition to the regulatory need for site specific biological measurements, it is
              useful to be able to represent the condition of complex ecosystems concisely, by means of
              composite indices or simple graphics, so that managers and non-specialists can readily evaluate
              and compare information, establish goals, and set priorities for remediation or protection.

                     It is not necessary to sample all subunits of an ecosystem. This would not be possible in
              any case, as all gear is selective to some degree. Assuming the ecosystem is integrated at some
              level, assessment of specific habitat types and/or locations should be adequate if methods are
              carefully selected.

                     NMFS should move forward to identify appropriate attributes that would constitute
              biological indicators of habitat quality for the following habitat types: SAV, riparian,
              estuarine benthic/water column, coastal benthic, and coral reef habitats. Ongoing activities
              around the nation that are involved in developing and applying biological indicators,
              biodiversity indices, and IBIs should be inventoried. A list of habitat priorities should be
              developed for investigation and feasibility studies.

                     NMFS must develop partnerships with other Federal, state, university and private
              groups that are involved or interested in developing and applying indices of biological
              integrity. Maximum use of ongoing programs should be made.

                     One difficulty with the application of IBI techniques to complex marine systems has
              been the relative lack of intimate knowledge of the ecological roles and interactions of specific
              species and/or functional guilds, compared to fresh water systems. Therefore, a basic element
              of any future IBI development work is simple taxonomic and natural history documentation of
              the species selected for use as markers of stress. Data gaps in life histories of critical species,
              including the degree of natural variation, must be identified and resolved.

                     A related problem is the definition of what constitutes a reference condition. A-priori
              selection of 'reference' sites based upon one set of parameters (e.g. contaminants) have not
              been tested for efficacy in habitats which may have been impacted by other stressors (e.g.

                                                               xi








               eutrophication). Ideally, a credible index should be responsive to any form of habitat
               degradation. A comparative assessment of the mathematical methods for derivation of
               reference sites and results has not been performed.

                       While the cumulative index may contain qualitative elements, the quantitative behavior
               of properly developed metrics in relation to each other, and our ability to assess them in
               relation to anthropogenic impacts is instructive. The detailed information from individual
               metrics is not lost in the process. The IBI approach provides a framework for assessing habitat
               quality in a consistent, technically defensible method. It has a demonstrated utility in fresh
               water environments as a technical assessment method and as a management tool.






































                                                             xii









                                                    1.0 Introduction


                    A major activity within the National Marine Fisheries Service (NMFS) is the
             implementation of the Essential Fish Habitat (EFH) requirements of the Magnuson-Stevens
             Fishery Conservation and Management Act. This legislation mandates that the NMFS amend
             each of its 39 Fishery Management Plans (FMPs) to include the best available information on
             habitat delineation, habitat needs, human impacts, and mitigative measures so that NMFS can
             protect, conserve, restore and enhance key habitats for each life stage of each managed species.

                    Implicit in the exercise of identification and delineation of EFH is that monitoring
             habitat quality is part of the process. While the productivity of fisheries is one of the ultimate
             management objectives of NMFS, the strength of the Magnuson-Stevens Act is that it is
             directed at protecting fisheries "habitat". Habitat assessment is easier than site-specific
             productivity assessment of multiple species at various times and specific life stages. Habitat
             quality assessment addresses the fundamental question of "how much of the habitat is still
             unimpaired?" and "what alterations are being imposed on that which remains?". A means to
             evaluate and monitor the ecological integrity of habitat is essential if that habitat is to be
             managed for fishery production or restored to productive habitat (Figure 1. 1). Therefore, the
             ultimate measure of EFH habitat quality is a measure of the condition of the biological
             community which inhabits it. This requires that the habitat be spatially and temporally
             delineated, and a method to continually assess biological condition be applied to it. Habitat
             quality monitoring programs should be incorporated into FMPs.

                    Sophisticated measures of habitat quality must be devised that reflect environmental
             coriditions and which are sufficiently robust to be used in a wide variety of physiographic
             regrions. A variety of habitat classification schemes have been devised for different regions and
             habitats (Brinson 1995, Davis and Harper 1996, Monaco et al. 1997, NOAA 1995, Osborne
             et al. 1991, Dethier 1990). These have incorporated a variety of indicators including basic
             water quality, physical and chemical parameters, and population metrics. These indicators are
             region specific in some cases. Indicators of biological integrity reflect parameters beyond those
             wLich only define the chemical and physical characteristics of the habitat, and should be used
             in concert with them to assess total habitat quality (Fausch et al. 1990). This will allow
             tracking the impacts of specific habitat stressors, such as contaminants, eutrophication, and
             wetland loss, and linkage of those stressors to ecological response. The ultimate indicator of
             habitat quality is the response of the biological community to the interaction of the stresses and
             recsources available at a particular location and time frame. The biological community acts as
             the integrator of habitat quality.

                    Many routine programs that provide data for the current generation of 'indicators' are
             related to monitoring for human health or regulatory control programs, as opposed to actual
             measures of environmental quality (EPA 1996). Frequently, they only reflect the size of
             regulatory programs (e.g.,number of permits), rather than actual loadings to, or impacts on the
             errvironment. Without substantial data manipulation and the imposition of significant

                                                              1








                assumptions, monitoring data may not be amenable to translation into actual metrics of impact
                (Warner et al. 1991).

                       To be useful for NMFS management application, a habitat quality assessment approach
                must be linked to, or at least correlated with, fishery production in that habitat. When NMFS
                is called upon to engage in ecosystem management decisions that affect fisheries habitat, and
                where tradeoffs for other resource demands, such as water use, forestry, development,
                resource extraction, etc. are being considered, the role of NMFS is to estimate what is at risk
                in terms of fishery production. In addition to fisheries, the role of the National Oceanic and
                Atmospheric Administration (NOAA), including NMFS, is to act as stewards of the marine
                habitat and to protect the multiple benefits derived from the ftinctioning of the intact ecosystem
                beyond direct production of food (e.g.,biodiversity, recreation, natural products, etc.). Habitat
                quality assessment paradigms must reflect the impact of various anthropogenic activities on
                fisheries productivity and the integrity of the marine ecosystem which supports that
                productivity.

                       This Proceedings document is organized into seven sections which provide a brief
                introduction to biological indicators, workshop objectives, technical presentations, workgroup
                products, conclusions and recommended follow-up activities. The technical presentations are
                presented as project summaries. The interested reader may contact the primary authors for
                more detailed documentation. The presentations were grouped into four categories, including
                practical applications and data base development, fish community studies, benthic community
                studies and continental shelf studies. The presentation on application of biological indicators in
                the state of Ohio is considerably longer than the other sections. While the other examples are
                of no less interest, the state of Ohio has successfully incorporated the basic scientific
                assessment of biological community condition into the very practical, real world regulatory
                framework for water quality monitoring and enforcement. It illustrates that biological
                indicators can be utilized for environmental management and regulatory needs, and how this
                has been accomplished in at least one state.



                                                           References


                Brinson, M. 1995. The HGM approach explained. National Wetlands Newsletter.
                       Nov-Dec:7-16.


                Davis, T.J. and J.R. Harper. 1996. Estuarine mapping and classification system for British
                       Columbia. Resource Inventory Committee, British Columbia Ministry of Environment,
                       Victoria, BC, Canada.

                Dethier, M.N. 1990. A marine and estuarine habitat classification system for Washington
                       State. Wash. Nat. Heritage Prog., Dept. Nat. Res., Olympia, Wash.


                                                                2









           Fausch, K.D., J Lyons, J.R. Karr and P.L. Angermeier. 1990. Fish communities as indicators
                  of environmental degradation. In Biological Indicators of Stress in Fish S.M. Adams
                  (ed.), Am. Fish. Soc. Symposium 8, Bethseda, Md.

           U.S. Environmental Protection Agency. 1996. Environmental indicators of water quality in the
                  United States. U.S. EPA, Office of Water, Wash., DC. EPA 841-R-96-002.

           Monaco, M.E., S.B. Weisberg and T.A. Lowery. (in press). Summer habitat affinities of
                  estuarine fish in USA mid-Atlantic coastal systems. Fish. Manag. and Ecol.

           NOAA. 1995. Environmental sensitivity index guidelines. NOAA Tech Memo., NOS, ORCA
                  92. Seattle: Hazardous Materials Response and Assessment Div., NOAA.

           Osborne, L.L. et al. 1991. Stream habitat assessment programs in states of the AFS North
                  Central Division. Fisheries, 16:28-35.


           Warner, K.A., S.I. Hartwell, J.A. Mihursky, C.F. Zimmerman and A. Chaney. 1991. The
                  lower Patapsco River/Baltimore Harbor contaminant data base assessment project.
                  Prepared for Balt. Regional Planning Counc. Chesapeake Research Consortium,
                  Solomons, Md. 128 pp.



























                                                       3







                                                                                                  Aw
                                                                   4".






                                                                    Manage Habitat
                             Define Habitat                                                             FP rotect Habitat

                          physical                              WHAT             HOW
                          chemical                            conservation    reserves
                          biological                          land use        permits
                          temporal                            releases        mitigation                F@estore Hab7itat
                          spatial



                                             Fssess     Habitat  I                                         Assess Habitat
                                                EVALUATION                                                  MONITORING










               Figure 1. 1  Diagrammatic representation of habitat management activities requiring habitat assessment.
                                                                                                                   , S.,























                                                                           4








                                    2.0 Index of Biotic Integrity Overview

                    COUPLED WITH HABITAT DELINEATION, CHEMICAL ANALYSES AND PHYSICAL
            CHARACTERIZATION, BIOLOGICAL INDICATORS ALLOW ASSESSMENT OF ALTERATION OF THE
            ENVIRONMENT INCLUDING EUTROPHICATION, NONPOINT SOURCE POLLUTION, CONTAMINATION,
            SAV LOSS, ETC. BIOLOGICAL HABITAT QUALITY INDICATORS NEED TO BE DEVELOPED FOR
            SEVERAL TYPES OF MARINE ENVIRONMENTS TO MEASURE HABITAT QUALITY IN A VARIETY OF
            HABITAT TYPES.


                    The index of biological integrity (IBI) approach has been demonstrated to be an
            effective tool to reflect the cumulative response of the aquatic community to the total
            environment, with all the attendant interactions and compensatory limits of populations and
            communities (Karr and Chu 1997). Biological integrity can be represented by indices which
            integrate the interaction of the total environment with specific populations and communities.
            They may include multiple parameters which assess productivity, trophic interactions and
            species richness in the community (Figure 2. 1). Bioindicators also have the potential to detect
            effects of trace level contamination and ephemeral events which may have long term effects on
            resident biota.


                    Assessing stream pollution was the driving force behind the original development of
            IBIs (Karr 1981). The IBI approach integrates a variety of other impact assessment methods
            which have been developed. These reflect a range of complexity, including indicator species
            or taxa, various species diversity indices, and multivariate methods (Deegan et al. 1997, Engle
            et al. 1994, Weisberg et al. 1997). Conceptually, an IBI utilizes a set of categories which
            reflect important ecological parameters, for example, diversity, abundance and trophic
            function. Within each category a variety of metrics are generated. Diversity can be quantified
            by number of species, species richness or one of several diversity indices. Abundance can be
            assessed in terms of numbers of organisms or biomass. Functional metrics reflect interactions
            between community segments, for example predator prey ratios, density of predator species,
            etc. Other categories can be included, such as condition indices, prevalence of lesions,
            proportion of pollution tolerant species, etc. Each metric is a site-specific measurement which
            can be general or very specific (e.g.,number of striped bass/kin').

                    Each metric is then given a rating score on an ordinal scale (1, 2, 3 or 1, 5, 10 etc.).
            This step is very important as it normalizes the various metrics on a common scale (Figure
            2.2). Thus, the measurements must be devised carefully, as they will be treated as being of
            equivalent ecological importance in the calculations, unless a weighting scheme is employed.
            In addition, they must reflect community response to stress. Assigning the score involves a
            good deal of ecological expertise (e.g., are 200 striped bass/km@ twice as good as 100/km@, or
            are they within the same range of habitat quality?). The individual scores are then summed on
            a site-specific basis so sites can be compared to each other based on percentile ranking of data
            relative to all stations, or relative to a reference site. Consistent sampling methods among
            sampling locations is crucial. An alternative approach to selecting and scoring the metrics has

                                                           5








                been to designate reference sites a-priori based on other factors such as demonstrated lack of
                chemical contaminants, eutrophication, physical disturbance etc. An array of candidate metrics
                is then calculated and a final subset is selected based on statistical evaluation of the data (Engle
                et al. 1994, Strobel et al. 1995) . This approach allows application of IBI assessment to
                habitats in which functional relationships of resident organisms are not fully understood, due to
                high complexity or lack of data.

                       There are a variety of important technical questions that have to be addressed before
                this approach can be employed in marine environments for gauging habitat quality in EFH.
                Many of these problems have been solved for fresh water environments (Karr and Chu 1997).
                Some states utilize IBI monitoring for habitat quality assessment (Ohio EPA 1988). Some have
                integrated IBIs into water quality monitoring as a regulatory tool for enforcing water quality
                permits. Marine environments are more complex than streams, requiring the development of
                different approaches within the IBI framework. A variety of pilot projects in marine
                environments have been initiated which have experimented with different methods and
                procedures (Deegan et al. 1997, Engle et al. 1994, Guillen 1997, Jordan et al. 1994, Lenat
                1993, Linder et al. 1997, Nelson 1990, Weisberg et al. 1997). The NMFS Bioindicator
                Workshop was organized to enumerate and begin to address these parameters, establish a
                consensus approach to development of biological habitat quality indicators in EFH, and
                identify research and monitoring needs for future projects.



                                                           References


                Deegan, L.A., J.T. Finn, S.G. Ayvazian and C.A. Ryder-Kieffer and J. Buonaccorsi. 1997.
                       Development and validation of an estuarine biotic integrity index. Estuaries 20:601-
                       617.


                Engle, V.D., J.K. Summers and G.R. Gaston, 1994. A benthic, index of environmental
                       condition of Gulf of Mexico Estuaries. Estuaries 17(2).

                Guillen, G.J. 1996. Development of a Rapid Bioassessment Method and Index of Biotic
                       Integrity for Tidal Streams and Bayous located along the Northwest Gulf of Mexico.
                       1996. TNRCC Special Report. Houston, Texas.

                Jordan, S. J., C. Stenger, M. McGinty, T. Arnold, S. Ives, D. Randall, B. Rodney, S. 1.
                       Hartwell. 1994. Estuarine habitat assessment and index of biotic integrity
                       demonstration and testing. Draft Final Report to U. S. Environmental Agency, Office
                       of Water. l7pp.

                Karr, J.R. 1981. Assessment of biotic integrity using fish communities. Fisheries. 6(6):21-27.

                Karr, R.R. and E.W. Chu. 1997. Biological monitoring and assessment: Using multimetric

                                                                6








                  indexes effectively. EPA 235-R97-001. Univ. Washington, Seattle, Wash.

           Lenat, D.R. 1993. A biotic index for the southeastern United States: Derivation and list of
                  tolerance values, with criteria for assigning water-quality ratings. J. N. Am. Benthol.
                  Soc. 12:279-290.


           Linder, C., M. McGinty, D. Goshorn and K. Price. 1997. Physical habitat and fish
                  assemblages: an investigation of the near-shore areas of the Chesapeake Bay and
                  Maryland's Coastal Bays. University of Delaware, Lewes, Delaware and Maryland
                  Department of Natural Resources, Annapolis, Maryland.

           Nelson, W.G. 1990. Prospects for development of an index of biotic integrity for evaluating
                  habitat degradation in coastal systems. Chem. and Ecol. 4:197-210.

           Ohio Environmental Protection Agency. 1988. Biological criteria for the protection of aquatic
                  life: Vol I. The role of biological data in water quality assessment. Ohio EPA, Div.
                  Water Qual. Monitoring and Assessment, Columbus, Ohio.

           Strobel et. al. 1995. Statistical Summary: EMAP-Estuaries Virginian Province, 1990-1993.
                  EPA/620/R-94/026.


           Weisberg, S.B., J.A. Ranasinghe, D.M. Dauer, L.C. Schaffner, R.J. Diaz and J.B. Firthsen.
                  1997.An estuarine benthic index of biotic integrity (B-IBI) for Chesapeake Bay.
                  Estuaries, 20: 149-158.


























                                                         7










                                                 Biological Index Metrics

                                        Diversity                     Abundance                         Trophic Level
                                  species 1  [-@,-0.0.0.00000         species 1 X

                                  species 2                           species 2 2X

                                  species 3                           species 3 4X

                                  species 4                                      7X


                                   Population Structure            Migrants vs Residents              Tolerant vs; Intolerant



                                      (3 0 Q
                                     (a (D (0 CD C) C@










                     Figure 2.1          Typical metrics utilized in IBI assessments.







                                                                                   8











                                    Index Calculation

                                              Criteria Values            Score


                                              low    med. high
                                metric 1        X      y     z              al
                Category I      metric 2        X      y     z              a2
                                metric 3        X      y     z              a3
                                metric 4        X      y     z              a4



                                metric 1        X      y     z              a,
                                metric 2        X      y     z              a2
                Category II     metric 3        X      y     z              a3
                                metric 4        X      y     z              a4
                                metric 5        X      y     z              a5
                                metric 6        X      y     z              a6


                Category III    metric 1        X      y     z              a,
                                metric 2        X      y     z              a2


                                       Index       =   f (an)



          Figure 2.2   IBI calculation scheme. X, Y, and Z are cutoff values unique to each metric.
          Each score (e.g., 1, 5, 10) is assigned based on the value of measured environmental measures,
          relative to the criteria. The final index is the sum of the scores, which may be weighted in
          some fashion.





                                                   9








                                                3.0 Workshop Objectives

                       The workshop was organized to develop a consensus within NMFS on the methods and
                utility of biological indicators for assessing habitat quality in EFH. The focus was directed
                toward practical application of bioindicators in marine systems and research needs. The
                workshop was arranged into one day of presentations by researchers in the field on conceptual
                approaches, experimental methods, pilot program results and, management applications. The
                following day, workshop participants broke into three discussion groups to deliberate which
                biological parameters would be practical and meaningful to measure in each of three broad
                habitat types (benthic, pelagic, vegetated) for the purpose of generating marine bioindicators.
                The three groups then compared and contrasted results. Finally, a general discussion of how to
                proceed with development of bioindicators for application to EFH was conducted. Several
                general questions were initially used as the basis for discussions of metric development and

                use.


                1 .    What categories, metrics and calculation methods will work in which environments?
                               Can the same metrics be used in an estuary in the Gulf of Mexico and an estuary
                               in North Carolina or Maine? The specific measurements within each category or
                               metric may have to be different in each case, because the biological communities
                               will differ in species composition in different regions. Should metrics be
                               expressed in terms of proportion or absolute numbers, e.g.,abundance.

                2.     Which environments are most feasible to assess this way?
                               Within the limits of manpower and resources, how much effort will be required
                               to obtain data with acceptable statistical power in differing habitat settings
                               (e.g., estuarine, open coastal, kelp bed).

                3.     How can individual metrics be assessed consistently over spatial and temporal
                       regimes?
                               Can the metrics be designed such that the need for 'professional judgement ' is
                               eliminated? Can the metrics be designed'so that a score of 28 in the Gulf of
                               Mexico indicates the same ecological quality as a similar score from an estuary
                               on the Atlantic coast or Alaska?


                4.     Can/should benthic-pelagic coupling be addressed?

                5.     What are the spatial scales over which bioindicators can be applied?
                               How will habitat be delineated? How large an area in the delineated habitat
                               should be assessed to confidently evaluate that habitat? Can results from a small
                               area be extrapolated to surrounding habitat, or must the entire region be
                               evaluated? What are the upper bounds of habitat area that can be assessed before
                               localized impairment becomes undetectable?


                                                              10









            6.     Can estuarine results be coupled with coastal habitat units? Can watershed results be
                   coupled with estuarine habitat units?

            7.     Can biological indicators be developed that are responsive to specific stressors?

            8.     Should redundancy be avoided or ignored in metric selection?


            9.     How should reference condition/sites be determined?
                          How clean is clean?


                                            WORKING DEFINITIONS


            Assemblage The association of interacting populations of organisms in a selected habitat.

            Attribute     A measurable factor in the biological assemblage.

            Biological Assessment       An evaluation of the condition of a habitat based on
                          measurements of attributes of the biological assemblage.

            Biological Integrity The ability of an ecosystem to support and maintain a stable community
                          of organisms having the structural and functional organization comparable to
                          that of an undisturbed habitat within a region.

            Category      A group of metrics which express a characteristic of habitat (e.g.,diversity,
                          abundance, function, etc.)

            Community     An assemblage of populations of organisms which either reside in, or utilize a
                          specific habitat, within a particular time frame.

            Ecological Integrity The condition of an ecosystem as measured by the chemical, physical,
                          and biological attributes.

            Habitat       The combined ecological features of an area, including chemical, physical and
                          biological components.

            Index         An integrated expression of habitat condition incorporating multiple metrics.

            Metric        A specific biological attribute, with a demonstrated empirical response to a
                          gradient of anthropogenic disturbance.









                                            4.0 Technical Presentations


               4.1 ATTAINING ENVIRONMENTAL GOALS: BIOLOGICAL MONITORING AND
                                      ASSESSMENT IN THEORY AND PRACTICE


                                                      James R. Karr


                                          University of Washington, Seattle, WA

                      Models guide much that we do in basic and applied ecology, including efforts to protect
               environmental quality. Models--whether conceptual, physical, or mathematical--can be wrong
               when they focus on the wrong endpoint or when they do not incorporate critical system
               components or processes. When models are not tested for their relevance in the real world,
               they can lead us astray as they squander both financial and environmental resources. It is
               especially regrettable when models lead us to ignore biological common sense or when
               scientists and managers focus on statistical significance rather than magnitude of effect and its
               biological consequence.

                      Because ambient biological monitoring focuses our attention on the most integrative
               endpoint (biological condition), we can use biological monitoring to test our models and assess
               the extent to which our policies protect ecological health. Biological monitoring has evolved
               rapidly during the twentieth century as knowledge has changed, and human-imposed stresses
               have become more complex and pervasive. Multimetric biological indices, like the index of
               biological integrity (IBI), integrate knowledge from earlier monitoring approaches while
               avoiding indicators that are flawed theoretically (Karr and Chu, 1997).

                      Developing effective multimetric biological indices involves five major activities:

               1 .    Classifying environments to define homogeneous sets within or across regions (e.g.,
                      large or small streams, warmwater or coldwater streams).

               2.     Selecting measurable attributes that provide reliable and relevant signals about the
                      biological effects of human activities.

               3.     Developing sampling protocols and designs that ensure that those biological attributes
                      are measured accurately and precisely.

               4.     Defining analytical procedures to extract and understand relevant patterns in the data
                      gathered.

               5.     Communicating the results to citizens and policy makers so that all concerned
                      stakeholders can contribute to environmental policymaking.


                                                            12













                                                     References


            Karr, J.R. and E.W. Chu. 1997. Biological Monitoring and Assessment: Using Multimetric
                   Indexes Effectively. Univ. Washington, Seattle, WA. EPA 325-R97-001.











































                                                         13









                4.2      THE OHIO ENVIRONMENTAL PROTECTION AGENCY'S BIOLOGICAL
                            MONITORING PROGRAM, IBI MEASURES AND THEIR POSSIBLE
                                     APPLICATION TO ESTUARINE ENVIRONMENTS


                                              Roger F. Thoma and Chris 0. Yoder

                                                   Ohio EPA, Columbus, OH


                Introduction
                       The value of biological criteria and biological assessment techniques has been
                demonstrated by their broad applicability not only to existing efforts to protect, restore, and
                manage aquatic resources, but in determining where management and restoration resources
                should be invested. The majority of the attention given to biological criteria thus far has dealt
                with how they fit into existing Water Quality Criteria (WQC) and National Pollutant Discharge
                Elimination System (NPDES) permit frameworks. While this is certainly an important set of
                issues, it would be a mistake to emphasize only this one program area, as it has the
                demonstrated ability to be useful in virtually any issue involving the management of water
                resources where a goal is to protect, enhance, or restore aquatic communities and ecosystems.
                We define the management of aquatic resources here as being broader than the traditional
                purview of water quality management. Efforts to attain the goals espoused by the Clean Water
                Act (CWA) and other initiatives (e.g., maintenance and recovery of aquatic biodiversity)
                should recognize the potentially broad role that biological criteria and assessment have in each
                area. We believe that biological criteria and the attendant concepts of regionalization and
                reference sites have a broad applicability beyond the CWA.

                       The Ohio EPA water programs have relied extensively on ambient bioassessments since
                the late 1970s. The program areas within which biological criteria have found the most
                widespread uses are the biennial water resource inventory (305b report), water quality
                standards (aquatic life use classifications), NPDES permits (includes enforcement and litigation
                support), the construction grants program (now the State Revolving Loan Fund), the Ohio
                Nonpoint Source Assessment (CWA section 319), evaluation of wet weather flow impacts
                (stormwater, CSOs), the state certification of CWA section 404 permits (401 program) and
                petitioned ditches, ranking of Comprehensive Environmental Response, Compensation, and
                Liability Act (CERCLA) sites, and comparative risk. In addition the biological data has
                proved useful to other state agencies including the Ohio Department of Natural Resources
                (rare, threatened, and endangered species, scenic rivers, nonpoint source management,
                fisheries management) and the Ohio Department of Transportation (environmental impact
                statements).

                       There is little question that aquatic resources have been and continue to be degraded by
                a myriad of land use and resource use activities. Benke (1990) summarized the status of the
                nation's high quality rivers and streams concluding that fewer than 2% remain in this category.
                Judy et al. (1984) indicate that the declining status of surface waters across the U.S. is largely

                                                                14









              the result of nonpoint source impacts. A continued reliance on technology based and even
              water quality-based solutions to these problems will simply be insufficient. Water resources in
              Ohio and elsewhere have historically been and will continue to be impacted by human activities
              beyond those targeted by the NPDES permit process. These remaining problems are
              comparatively more complex and subtle, but are no less important or real. In fact, it is these
              more subtle and diffuse impacts which imperil aquatic resources to the point where additional
              species are declining in distribution and abundance, this in addition to those already declared as
              rare, threatened, or endangered (Ohio EPA 1992).

                     A monitoring approach, integrating biosurvey data that reflects the integrity of the
              water resource directly, with water chemistry, physical habitat, bioassay, and other monitoring
              and source information, must be central to accurately defining these varied and complex
              problems. Such information must also be used in tracking the progress of efforts to protect
              and rehabilitate water resources. The arbiter of the success of water resource management
              programs must shift from a heavy reliance on administrative activity accounting (numbers of
              permits issued, dollars spent, or management practices installed) and a pre-occupation with
              chemical water quality alone to more integrated and holistic measurements with overall water
              resource integrity as a goal. Biocriteria seems an essential component in making this shift.

                     Emphasizing aquatic life use attainnient is important because: 1) aquatic life criteria
              oftentimes result in the most stringent requirements compared to those for the other use
              categories, (i.e., protection for the aquatic life use criteria should assure the protection other
              uses); 2) aquatic life uses apply to virtually all waterbody types and the diverse criteria (i.e.,
              includes conventional pollutants, nutrients, toxins, habitat, physical, and biological factors,
              etc.) apply to all water resource management issues; and, 3) aquatic life uses and the
              accompanying chemical, physical, and biological criteria provide a comprehensive and
              accurate ecosystem perspective towards water resource management which promotes the
              protection of ecological integrity.

                     Finally, biocriteria can aid greatly the visualization of aquatic resource values and
              attributes. This is a critical need if we are to change the prevailing view of watersheds and
              streams as merely catchments and conveyances for municipal and industrial wastes, excess
              surface and subsurface drainage, or as obstacles to further land developments. In an effort to
              stem the virtually unabated loss of riparian habitat and watershed integrity, Ohio EPA has
              adopted a stream protection policy which sets forth guidelines under which various activities
              will need to be conducted in order to conserve biological integrity. Without biocriteria and the
              case examples developed over the past 14 years this would not have been possible and any
              opportunity to affect these degrading influences would have been lost.

                     While we have demonstrated how biological criteria can be developed and used within a
              state water resource management framework, some important challenges remain. The
              cumulative costs associated with environmental mandates, many of which consist of
              prescription-based regulations, have recently come into question. Both the regulated

                                                               15








               community and the public desire evidence of "real world" results in return for the expenditures
               made necessary by federal and state mandated requirements. Biological criteria seem
               particularly well suited to meet some of these needs in that the underlying science and theory is
               robust (Karr 1991) and biocriteria certainly qualifies as "real world".

                      While no single environmental indicator can "do it all", particularly in the more
               complex situations (i.e., multiple discharges, habitat alterations, presence of toxic compounds,
               etc.), it is obvious that biological criteria have a major role to play. A lack of information
               from, or an over-reliance on any single class of indicators can result in environmental
               regulation that is less accurate and either under- or overprotective of the water resource.
               Accounting for cost is not only a matter of dollars spent, but is also a question of
               environmental accuracy and technical validity. In short, a credible and genuinely cost-effective
               approach to water quality management should include an appropriate mix of chemical,
               physical, and biological measures, each in their respective roles as stressor, exposure, and
               response indicators. Comprehensive monitoring designs using such cost-effective indicators
               must become a part of the "cost of doing business" and perhaps at the expense of programs
               where new evidence suggests that the resources devoted are disproportionate to the magnitude
               of the present problems (e.g., point sources vs. nonpoint sources).

                      Based on our experience over the past 17 years it is evident that including a biological
               criteria approach in a state's monitoring and assessment effort can foster a more complete
               integration of important ecological concepts, better focus water resource policy and
               management, and enhance strategic planning. Some specific examples include:

               1 .    Watershed Approaches to Monitoring, Assessment, and Management: The monitoring
                      and assessment design inherent to biological criteria is fundamentally watershed
                      oriented and will yield information on a watershed basis.

               2.     Integrated Point, Nonpoint, and Habitat Assessment and Management: Biological
                      criteria integrate the effects of all stressors over time and space, and the attendant use
                      of chemical, toxicological, and physical tools enables the association of probable causes
                      of observed impairments. This should provide a firm setting for the collaborative use
                      of the same information for the management and regulation of both point and nonpoint
                      sources (including habitat), two disciplines which have thus far been operated as
                      independent programs.

               3.     Cumulative Effects: Biological communities inhabit the receiving waters all of the time
                      and reflect the integrative, cumulative effect of various stressors. Such information
                      provides a basis for management programs to evaluate different problems in relative
                      terms.


               4.     Biodiversity Issues: The basic biological data provides information about species,
                      populations, and communities of concern and also provides the opportunity to focus

                                                             16









                    beyond ecosystem elements, but include an assessment of processes as well.

             5.     Interdisciplinary Focus: Because of the inherently integrative character of the
                    biosurvey monitoring and assessment design, a biological criteria approach provides the
                    opportunity to bring ecological, toxicological, engineering, and other sciences together
                    in planning and conducting assessments, interpreting the results, and using the
                    information in strategic planning and management actions.

                    ExaLuples of Non-Clean Water Act Uses
                    Biocriteria, because they measure the overall condition of aquatic communities and
             hence reflect the condition of the entire aquatic resource, are potentially useful outside the
             traditional purview of CWA programs. One of these areas is with nongame species,
             particularly the rare, endangered, threatened, and special status species listed by government
             agencies. Presently, in Ohio, 25 species are listed as endangered, 8 species as threatened, 13
             species as special interest, 5 as extirpated, and 2 as extinct; this represents more than 30% of
             the Ohio fauna. Of the 41 species listed by Ohio EPA as extremely intolerant, intolerant, and
             sensitive (Ohio EPA 1987), 25 are listed as endangered, threatened, or special status. Sixteen
             additional species are in the process of significant declines, some of which are declining more
             rapidly than others (Rankin et al. 1992). This increases to more than 40% the fraction of the
             Ohio fish fauna which is potentially imperiled. If introduced species and those species that are
             on the fringe of their natural range are excluded, these percentages become even higher.
             These trends are potentially symptomatic of other environmental problems that could
             eventually emerge to affect attributes of surface waters which are of more direct human
             interest. Fish species that depend on relatively clean, silt free substrates, the continuous
             presence of good quality water, good instream. cover, and headwater stream habitats seem to be
             most seriously affected. This information was provided by the biosurveys conducted by Ohio
             EPA over the past 14 years, thus the multiple use of the same data is exemplified. It also
             demonstrates the opportunity to utilize the dimensions of the data in ways which would
             otherwise become collapsed in the IBI evaluations. Nongame aquatic communities are not only
             indicators of acceptable environmental conditions for themselves, but also indicate that the
             water resource is of an acceptable quality for wildlife and human uses since they have the
             ability to integrate and reflect the sum total of disturbances in watersheds. While individual,
             site-specific watershed and water body disturbances themselves may seem trivial to some, the
             aggregate result of these individual impacts emerges in the form of a degraded and declining
             fauna on a regional or watershed scale. We will have a very difficult time demonstrating this
             problem if we do not employ monitoring and assessment efforts which generate this type of
             information in a scientifically credible manner which the public will accept.

                    Another potential use for biocriteria is in the management and assessment of lotic
             fisheries. The smallmouth bass (Micropterus dolomieui) is one of the most important game
             species in Midwestern rivers and streams. Furthermore, it is a species which requires little or
             no external support in the form of supplemental stocking. However, like any other valued fish
             species it does have specific habitat and water quality requirements. We examined the

                                                             17








                relationship between the occurrence and abundance of smallmouth bass with the Index of Biotic
                Integrity (IBI) throughout the state. The overall pattern is that this species reaches is highest
                abundance and occurs most frequently at sites with IBI scores at least in the fair range and
                preferably in the good and exceptional range. As expected, the species declines sharply as the
                IBI indicates increasingly degraded conditions (i.e., poor or very poor). This analysis
                demonstrates the relevance of the IBI to and correlation with tangible resource benefits of
                direct importance to resource users specifically and the public in general.

                       Activities requiring a permit under Section 404 of the CWA must be certified as
                meeting provisions of the WQS by the Ohio EPA. These are referred to as 401 certifications
                which largely pertain to wetlands and stream habitat impacting activities. These are the third
                leading cause of nonpoint source impact which has undoubtedly resulted in some of the most
                irretrievable impairments to aquatic life uses in Ohio. Biological evaluations of 401
                certification issues has greatly increased since the adoption of numerical biocriteria and the
                attendant field evaluation techniques. This is presently the best legal means by which Ohio
                EPA can protect habitat quality. Biological criteria are especially useful in this process since
                habitat is a predominant factor in determining the ability of an ecosystem to support a
                structurally and functionally healthy assemblage of aquatic life. Furthermore, by using the
                result of the work that supported the development of the Qualitative Habitat Evaluation Index
                (QHEI; Rankin 1989), the ecological consequences of projects involving the degradation of
                lotic habitat can be predicted. This allows Ohio EPA to prevent unnecessary degradation of
                aquatic habitat and communities.

                Major Factors That Determine Water Resource Integrity
                       Multiple factors in addition to chemical water quality are responsible for the continuing
                decline of surface water resources in Ohio (Ohio EPA 1997) and the U.S. (Judy et al. 1984;
                Benke 1990). These include the modification and destruction of riparian habitat, sedimentation
                of bottom substrates, and alteration of natural flow regimes on a watershed scale. Because
                biological integrity is affected by multiple factors, controlling chemicals alone does not assure
                the restoration of biological integrity (Karr et al. 1986). Biological criteria and attendant
                monitoring and assessment designs provide a means to incorporate broader concepts of water
                resource integrity while preserving the appropriate roles of the traditional chemical, physical
                and toxicological approaches developed over the past three decades.

                       The health and well-being of aquatic biota is an important barometer of whether we are
                achieving the Clean Water Act goal of maintaining and restoring the biological integrity of the
                nations's surface waters. This concept underlies the basic intent of state water quality
                standards. Yet this tangible end-product of Clean Water Act regulatory and water quality
                planning and management efforts is frequently not linked to source control and other
                performance measures. Simply stated, biological integrity is the combined result of chemical,
                physical, and biological processes in the aquatic environment. The interaction of these
                processes is readily apparent in the functioning of ecosystems. Thus management efforts
                which rely solely on comparatively simple, surrogate approaches to assessment and

                                                                18








            management carry a significant risk of failure in attempting to achieve the restoration of
            ecological integrity (Karr 1991). Therefore, ecological concepts, biological criteria, and
            attendant monitoring and assessment tools must be further incorporated into the management of
            surface water resources.


            Understanding Biological Integrity: A Prerequisite to Biological Criteria
                    The term biological integrity originates from the Federal Water Pollution Control Act
            (FWPCA) amendments of 1972 and has remained a part of the subsequent reauthorizations.
            Early attempts to define biological integrity in practical terms were inconclusive. Although
            one of these efforts failed to produce a consensus definition or framework, several contributors
            urged that an ecological approach be employed (Ballentine and Guarria 1975). Biological
            integrity has since been considered relative to: 1) conditions that existed prior to European
            settlement; 2) the protection and propagation of balanced, indigenous populations; and, 3)
            ecosystems that are unperturbed by human activities. These criteria (especially 1 and 3) easily
            could be construed as referring to a pristine condition that exists in few, if any, ecosystems in
            the conterminous United States. Subsequent to this initial effort, a U.S. EPA sponsored work
            group concluded that biological integrity, when defined as some pristine condition, was
            difficult and impractical to define and measure (Gakstatter et al. 1981). The pristine vision of
            biological integrity was considered as a conceptual goal towards which pollution abatement
            efforts should strive, but the group also realized that past, present, and future impacts to
            surface waters may prevent the full realization in many parts of the U.S. More recently,
            efforts to construct a workable, practical definition of biological integrity have provided the
            supporting theory that necessarily precedes the development of standardized measurement
            frameworks, techniques, and criteria for determining compliance with that goal. Our concept
            of biological integrity follows the definition of Karr and Dudley (1981)   ...... the ability of an
            aquatic ecosystem to support and maintain a balanced, integrated, adaptive community of
            organisms having a species composition, diversity, and functional organization comparable to
            that of the natural habitats of a region". This definition alludes directly to measurable
            characteristics of biological communities which are found in the least impacted habitats of a
            region. It was this definition and the underlying ecological theory which provided the
            fundamental basis for the development of numerical biological criteria using a regional
            reference site approach in Ohio. U.S. EPA adopted a facsimile of this definition in their
            national program guidance on biological criteria (U.S. EPA 1991).


                    Biological Criteria
                    Biological evaluation of aquatic life integrity is made possible by monitoring aquatic
            communities directly. Community bioassessments differ from approaches which rely
            principally on target species or indicator organisms by relying on the aggregated information
            across multi-species assemblages. The aggregation of key community attributes functions as
            an indication of the more complex ecosystem elements and processes which are not measured
            directly and completely. At the same time information about individual species is not lost in
            the process and can be accessed at any time.


                                                             19








                        Six criteria that biological monitoring programs should be judged against have been
                 defined (Herricks and Schaeffer 1985). These requirements are:

                 1 .    The measure(s) used must be biological.

                 2.     The measure(s) must be interpretable at several trophic levels or provide a connection
                        to other organisms not directly involved in the monitoring.

                 3.     The measure(s) must be sensitive to the environmental conditions being monitored.

                 4.     The response range (i.e., sensitivity) of the measure(s) must be suitable for the intended
                        application.

                 5.     The measure(s) must be reproducible and precise within defined and acceptable limits
                        for data collected over space and time.

                 6.     The variability of the measure(s) must be low.

                        Karr et al. (1986) evaluated the applicability of the IBI (based on stream fishes) to these
                 criteria and found that it satisfied the six requirements. These evaluation mechanisms which are
                 based on the recent improvements in'ecological theory (re: Karr and Dudley 1981) provide a
                 more comprehensive analysis of community information than do single dimension measures
                 such as diversity indices, species richness, indicator species, numbers, biomass, etc.
                 Furthermore the IBI type measures extract ecologically relevant information and provide a
                 synthesized, numerical result that can be understood by non-biologists.


                        Reference Condition and Reference Site
                        Although there is agreement that biological criteria should be based on data collected
                 from reference sites, there exists technically different approaches. Two of these, the U.S.
                 EPA Rapid Bioassessment Protocols (RBP; Plafkin et al., 1989) and the regional reference site
                 approach (Hughes et al. 1986) are the most commonly used. The RBP specifies the selection of
                 a 5ingle upstream or nearby reference site from which the results at unknown test sites are
                 evaluated. One problem with selecting a-gngle reference site is that the reference site could
                 differ in more than the imposition of an impact. In regulatory applications one potential
                 liability is that debates will center on whether the single reference site is sufficiently similar to
                 the impacted site rather than focusing on whether the test site departs from the reference
                 condition. If the single reference site is not representative of the impacted test sites then the
                 resulting biological criteria will be either under or over-protective.

                        We have encountered situations in Ohio where insufficient knowledge about regional
                 expectations resulted in misinterpretations about the severity of impacts in streams. A regional
                 reference framework offers a substantial advantage for the interpretation of community
                 responses beyond the derivation of biocriteria. By offering a more robust framework based on

                                                                20








            multiple and regionally attuned reference sites, the chance for deriving an inappropriate
            biocriterion is greatly reduced. Benefits will also be realized by having an approach within
            which the same framework and information can apply to the different programs in which the
            protection of aquatic life is a goal.

                    The selection of reference sites from which attainable biological performance can be
            defined is a key component in deriving biological criteria. Hughes et al. (1986) described at
            least seven different approaches that have been used to estimate attainable biological conditions
            in surface waters. Regional reference sites can have a dual role as the arbiter of regionally
            attainable biological performance (which is the basis for numeric biological criteria) and as an
            upstream reference (more commonly referred to as a control) for determining the significance
            of any longitudinal changes. It is important to realize this duality and the differences between
            each role.


                    Control sites are applied in the longitudinal upstream/downstream design characteristic
            of most water quality studies in lotic systems. While it is possible for reference sites to double
            as upstream control sites, the reverse is not always true. The following is a synopsis of the
            important and distinctive characteristics of each:


                    Reference Sites:
                            "least impacted" sites are located throughout a homogeneous region (i.e.,
                            ecoregions);
                            biological performance across multiple sites defines expectations and variability;
                            benchmark levels (e.g., 25th percentile) of performance are used to establish
                            numerical biocriteria within an established system of tiered aquatic life uses
                            codified in the WQS;
                            data from the reference sites are used to calibrate the IBI and ICI on a statewide
                            basis - biocriteria are established on both an ecoregional and statewide basis;
                            re-monitoring of reference sites occurs on a periodic basis (i.e., once every 10
                            yrs.) which provides the opportunity to make periodic adjustments to the
                            indices, biocriteria, or both.


                    Control Sites:
                            used primarily in an upstream/downstream format to evaluate longitudinal
                            changes;
                            does not serve as an arbiter of use potential or use attainment - however, the
                            level of attainable performance for site-specific and antidegradation applications
                            is defined;
                            are important in point source monitoring and evaluation.

                    Ideally, reference sites for estimating attainable biological performance should be as
            undisturbed as possible and be representative of the watersheds for which they serve as
            models. Such sites can serve as references for a large number of habitat types if the range of

                                                            21








                physical characteristics within a particular geographical region are included (Hughes et al.
                1986). It is for this reason, among others, that the selection of only the most pristine sites as
                references is inadvisable. To do so would artificially restrict reference conditions to only
                rarely occurring benchmarks for evaluating progress or deterioration (Hughes et al., 1986).
                While it is recognized that individual water bodies differ to varying degrees, the basis for
                having regional reference sites is the similarity of watersheds within defined geographical
                regions. Generally less variability is expected among surface waters within the same region
                than between different regions. This is because surface waters, particularly streams, derive
                their basic characteristics from their parent watersheds. Thus streams draining comparable
                watersheds within the same region are more likely to have similar biological, chemical, and
                physical attributes than from those located in different regions.


                Framework for Deriving Numerical Biological Criteria
                       The derivation of biological criteria for Ohio surface waters is based on the biological
                community performance which can be attained at regional reference sites. The numerical
                biological criteria that result from the application of this framework represent the ecological
                structure and function that can reasonably be attained given present-day background
                conditions. Although these criteria are not an attempt to define pristine, pre-Columbian
                conditions, the framework design includes a provision for future changes to the criteria which
                would take place if changes in background conditions occur.

                       The framework within which biological criteria are established and used to evaluate
                Ohio rivers and streams includes the following major steps:

                1 .    selection of indicator organism groups;

                2.     establish standardized field sampling, laboratory, and analytical methods;

                3.     selection and sampling of least impacted reference sites;

                4.     calibration of multi-metric indices (e.g., 1131, ICI);

                5.     set numeric biocriteria based on attributes of tiered aquatic life use designations;

                6.     reference site re-sampling (10% of sites sampled each year);

                7.     make periodic adjustments to the indices, biocriteria, or both as determined by
                       reference site re-sampling results.

                       The key steps in this process are illustrated in Figure 4.2.1 I-VI, and presume that
                narrative statements of biological community condition (i.e., designated aquatic life uses)
                already exist in the WQS and that a regionalization scheme (e.g., ecoregions) is also included.

                                                                22









             1.     Indicator Assemblages
                    Our experience has shown that at least two assemblages should be monitored. Fish and
             macroinvertebrates were chosen as the routine organism groups for Ohio to monitor because
             both groups met the above criteria, have been widely used in environmental assessment, and
             there is an abundance of information about their life history, distribution, and environmental
             tolerances. The need to use two assemblages is apparent in the ecological differences between
             them, differences that tend to be complimentary in an environmental evaluation. The recovery
             rates differ between these two groups which can provide insights about whether or not a
             pollution problem has been completely abated. The value of having two assemblages
             independently showing the same result cannot be overstated and lends considerable strength to
             an assessment. However, differential responses can lead to the definition of problems that
             might otherwise have gone undetected in the absence of information from one or the other
             organism group. The differing sensitivities of the two groups is not the same to all substances
             or in every situation. Thus the resultant information can influence decisions to control certain
             substances or processes that might have been overlooked or underrated in an evaluation based
             on only one group. The use of these two groups is somewhat analogous to the use of a fish
             species and an invertebrate species as standard bioassay test organisms.

             2.     Field and Laboratory Methods and Logistics
                    The choice of field sampling methods is a cornerstone aspect of using and implementing
             bioassessments and biocriteria. Although a variety of methods and techniques are available,
             the choice of which ones to use should be dictated by the conditions that exist in a particular
             state or region. There are a number of equally valid techniques, some of which will
             undoubtedly work better in some habitats and/or regions of the U.S. than in others.

                    In selecting the appropriate field and laboratory methods there were several
             considerations. These include:


             1 .     the need to produce assessments which are capable of discriminating the various
                    impacts that occur in Ohio surface waters;

             2.     scientific validity; and,


             3.     cost-effectiveness.


                    These are inherently competing objectives. Elaborate and highly detailed assessments
             are not very cost-effective, yet the need for scientific validity prescribes an inherent level of
             rigor and complexity in the assessment and hence a higher cost. In contrast, assessments
             which lack sufficient detail and rigor may cost less, but lose in cost-effectiveness by lacking
             the scientific validity necessary to discriminate impacts which actually exist. Given the
             economic, social, and environmental consequences of the decisions being made with the data
             and results, it seems wiser to opt for a more complex   and rigorous assessment.



                                                              23








                 3.     Criteria for Selecting Reference Sites
                        The selection of reference sites is another cornerstone issue in biocriteria derivation.
                 Should reference sites be selected primarily on a cultural basis without prior detailed
                 knowledge of the reference site sampling results? Or, should the sampling results be used to
                 assist in the selection of reference sites? We believe the latter approach may induce some
                 unintentional bias into the biological criteria calibration and derivation process because of the
                 inherent tendency to select the best sites instead of a more representative, balanced cross-
                 section of sites that reflect both typical and exceptional communities. In extensively disturbed
                 regions and uniquely undisturbed regions, the method of reference site selection will likely be
                 less of an issue because of the relatively homogenous conditions. However, in regions that
                 have a gradient of disturbances, the method of selection becomes more critical.

                        A notched box-and-whisker plot method was used to portray the reference site results
                 for each biological metric by ecoregion. These plots contain sample size, medians, ranges with
                 outliers, and 25th and 75th percentiles. Box plots have an important advantage over the use of
                 means and standard deviations (or standard errors) because a particular distribution of the data
                 is not assumed. Furthermore, outliers (i.e., data points that are two interquartile ranges
                 beyond the 25th or 75th percentiles) do not exert an undue influence as they do on means and
                 standard errors. In establishing biological criteria for a particular area or ecoregion we
                 attempted to represent the typical biological community performance, not the outliers. The
                 latter can be dealt with on a case-by-case or site-specific basis if necessary.


                        The Role of Reference Results in Biocriteria Derivation. The data obtained from
                 sampling regional reference sites provides the basis for deriving numerical biological criteria.
                 Reference sites serve a fundamental purpose by providing the database for calibrating the
                 multi-metric indices and deriving the ecoregional numeric biocriteria. The reference database
                 was used to establish the actual IBI, Invertebrate Community Index (ICI), and the Modified
                 Index of well being (MIwb) biological criteria for each of three applicable aquatic life uses,
                 three site types (IBI and MIwb), and across the five ecoregions; of Ohio. This was done out of
                 necessity on a statewide basis, but it could be organized differently. This is where broader
                 calibration regions that extend beyond state boundaries could be useful.

                        It is imperative that reference sites meet the aforementioned criteria and thus be
                 representative of the attainable biological community performance respective of habitat type
                 within each ecoregion. The initial selection of reference sites occurred during the Stream
                 Regionalization Project (SRP) of 1983-84. The results of this effort are reported in Larsen et
                 al. (1986) and Whittier et al. (1987). While the 1983-84 SRP focused on watersheds with
                 drainage areas of 10-300 square miles the reference site network was consequently
                 supplemented with data from additional locations with drainage areas of 1-200 square miles
                 sampled during 1981-89 (Ohio EPA 1987, 1989). These included reference sites on larger
                 streams, mainstem rivers, and headwaters streams throughout the state. The transitional
                 sections of Lake Erie tributaries, the Ohio River, and inland lakes and reservoirs were not
                 included in this analysis. However, work is underway to address these areas within the next

                                                                 24









            three to five years.

                    How MaLiy Reference Sites Are Enough         We have frequently been asked this
            question as most are interested in deriving technically valid biological criteria at the lowest
            cost. Logically, enough reference sites must be selected to account for the range of natural
            variability among the least impacted reference sites within a region. Increased variability
            among reference sites, if it originates from natural sources and not sampling error, indicates
            the need to employ a stratification scheme among the reference sites for the purpose of
            biocriteria derivation. Stratification of natural variability is an essential component of
            biological criteria development if the resultant criteria are to become managerially useful. Our
            approach accomplished this through the use of tiered use designations, site types, and
            ecoregional stratification. Additional stratification variables could include mean annual
            temperature ( e.g., warmwater versus coolwater streams; Lyons 1992) and gradient (e.g., low
            gradient versus high gradient streams; Leonard and Orth 1986).

                    High variability among reference sites without obvious natural causes could be a result
            of sampling problems which an increased number of reference sites would not correct.
            However, assuming proper stratification and a valid sampling approach we can then determine
            the minimum number of reference samples needed to arrive at a biocriterion (e.g., 25th
            percentile for a use designation) which adequately represents the potential biological
            performance of a region. The range of natural variability will not be encompassed with an
            insufficient reference database on which stratified expectations are to be based. This could
            result in biocriteria that are either under or over-protective of the biological performance
            defined by the designated aquatic life uses.

                    To illustrate the effect of reference site sample size on the Ohio EPA IBI biocriteria, we
            randomly selected sites from our reference database for each ecoregion and site type
            combination and, without replacement, recalculated the 25th percentile warmwater habitat
            (WWH) biocriterion after samples were added in increments of five. The procedure was
            performed for 50 trials over 15 different sets of reference sites (5 ecoregions X 3 site types per
            each ecoregion). The results were plotted on a three dimensional bar chart with the frequency
            at which a 25th percentile biocriteria value was randomly selected versus sample size. The
            analog of an asymptotic relationship of a 25th percentile 1131 value with increasing sample size
            defined the minimum number of reference sites which are needed to achieve a biocriterion
            value which encompasses the inherent background variability.

                    Our criterion to determine when the analog to an asymptotic relationship was reached is
            where the variation in the 25th percentile value narrowed to one predominant value in terms of
            the number of observations per aggregation category. Of the 15 sets of reference samples
            tested (5 ecoregions X 3 site types per each ecoregion) this point ranged from a low of 10-15
            samples for headwater sites in the Interior Plateau ecoregion to 75-80 samples for boat sites in
            the Eastern Corn Belt Plain (ECBP) ecoregion. The Huron/Erie Lake Plain (HELP) ecoregion
            appeared to require the fewest reference samples to reach the point of diminishing return and

                                                            25








                 the ECBP ecoregion appeared to require the most reference samples. The other ecoregions
                 tended to be intermediate between the HELP. and ECBP.


                        Ecoregions with widespread and uniform land disturbance, such as the HELP
                 ecoregion, require fewer samples to characterize the prese reference condition while those
                 with a greater degree of natural heterogeneity (i.e., ECBP) require the most samples. Most of
                 the reference sites were sampled twice which makes the safe minimum number of sites for the
                 Ohio ecoregions from as few as 5-8 sites per ecoregion per site type/stream size strata for the
                 more homogeneous ecoregions to as many as 38-40 sites per ecoregion per site type/stream
                 size strata for the more heterogeneous ecoregions. This may illustrate the need for further
                 landscape stratification via sub-ecoregions. We believe that if uncertainty exists about the
                 variability within an ecoregion more sites should be used than too few. In our experience this
                 would be approximately 35-40 sites per ecoregion per site type. This may vary across the
                 nation as these figures are most representative of the Midwestern U.S.

                        A failure to stratify variability where the clear need for a stratification scheme exists
                 risks inaccurate biocriteria. that may be under-protective of sites with greater biotic potential
                 and over-protective of sites with lower biotic potential that otherwise would have been
                 adequately protected by lower criteria. In contrast, attempts to stratify regions where little
                 difference exists may lead to unnecessary regulatory complexity and an unsound and arbitrary
                 scientific basis for biocriteria development.

                        The minimum number of reference sites also depends on the statistics upon which the
                 criteria will be based. Extreme percentiles (e.g., 5th, 95th), because they represent the tails of
                 distribution ftinctions, are characterized by wider confidence bounds around the threshold
                 statistic and will require a larger number of sites before a stable asymptote is reached, whereas
                 the median of the same distribution will reach an asymptote with fewer samples (Berthouex and
                 Hau 1991).

                 4.     Calibration of Multi-metric Indices for Drainage Area
                        In order to establish biological criteria that are reflective of the legislative goal of
                 attaining and restoring biological integrity in surface waters, a calibration of multi-metric
                 indices is needed. The practical definition of biological integrity as the biological performance
                 exhibited by the natural or least impacted habitats of a region provides the underlying basis for
                 designing a reference site sampling network. This is not an attempt to characterize pristine or
                 totally undisturbed, pre-Columbian environmental conditions as such exists in only a very few
                 places, if any, in the conterminous U.S. (Hughes et al. 1982). The landscape and aquatic
                 ecosystems of Ohio have been significantly altered during the past 150-200 years. This
                 includes massive deforestation and conversion to agricultural and urban land use, extensive use
                 of rivers and streams for wastewater discharges, extensive drainage and elimination of more
                 than 90% of the wetlands, and extensive modification of stream and river habitats through
                 channelization, impoundment, and encroachment on the riparian zone. Together these
                 activities have radically altered the lotic ecosystems of Ohio, much of which is essentially

                                                                26









            irreversible. Thus expectations of how a biological community should perform are determined
            by the demonstrated attainability of natural communities at least impacted reference sites within
            a particular biogeographical region.

                    The reference site results were pooled on a statewide basis prior to constructing the
            drainage area scatter plots. Calibrating on a statewide (or other large area basis) as opposed to
            an ecoregion by ecoregion basis gives the resultant index important resolution between
            ecoregions. For example, it is useful to know that an index value of 30 means something
            different in the HELP ecoregion as compared to the WAP ecoregion while retaining
            comparability on a statewide basis. Having to deal with multiple, ecoregion-specific indices
            and resultant biocriteria values on a statewide and inter-regional basis would make
            communication and comparison much more difficult. Ideally, index calibration should occur
            on a broad spatial basis other than that defined by political boundaries. This is an area for
            further research and an opportunity for interstate cooperation.


            5.      Set Numeric Criteria
                    Once the task of calibrating the biological indices is completed the task of deriving the
            numerical biological criteria can proceed. However, on what basis were the decisions to select
            a baseline numerical criterion value for each index made? As was previously mentioned, Ohio
            EPA has employed a system of tiered aquatic life uses in the state WQS since 1978. These use
            designations are essentially narrative goal statements about the type of aquatic community
            attributes which are envisioned to represent each use. For the purposes of establishing
            numerical biocriteria, the two most important uses are WWH and Exceptional Warmwater
            Habitat (EWH). These use designations contain a narrative goal statement and specifies the
            numeric index thresholds which serve as the numeric biocriteria for each use. Numerical
            biological criteria for the WWH use designation, which is the most commonly applied aquatic
            life use in Ohio, were established as the 25th percentile value of the reference site scores by
            index, site type (fish), and ecoregion. The resultant numeric biocriteria for the WWH use vary
            by ecoregion in accordance with the narrative definition and the reference site results for each
            site type. It was felt that most of the least impacted reference results should be encompassed
            by the baseline WWH use designation for Ohio's inland rivers and streams. The selection of
            the 25th percentile value is analogous to the use of safety factors, which is commonplace in
            chemical water quality criteria applications, and has previous precedents such as the 75th
            percentile pH, temperature, and hardness used to derive unionized ammonia-nitrogen and
            heavy metals design criteria for wasteload allocations, using > 20 % mortality for determining
            significance in bioassay results, or even the 10 -6 risk factor for human exposure to carcinogens.
            In this sense the 25th percentile acts as a safety factor in the derivation process. Choosing the
            25th percentile as the minimum WWH criterion is conservative and reduces the influence of
            any unintentional bias induced by including potentially marginal sites.

                    Ohio EPA employs three indices as part of the numerical biological criteria: the IBI,
            the ICI, and the MIwb. The MIwb does not require a spatial calibration prior to use.
            However, both the IBI and ICI require calibration in order to establish individual metric

                                                           27








                scoring criteria tailored to the reference conditions. The sample value of each of the 12 1131
                metrics is compared to the range of values from the least impacted reference sites located
                within the same ecoregion. Each IBI metric receives a score of 5, 3, or 1, based on whether
                the sample value approximates (5), deviates somewhat from (3), or strongly deviates (1) from
                the range of reference site values. The maximum IBI score possible is 60 (i.e., all 12 metric
                scores = 5) and the minimum is 12 (i.e.,, all metric scores = 1).

                        To determine the 5, 3, and 1 values for each IBI metric the reference site data base was
                first plotted against a log transformation of drainage area, the latter serving as an indicator of
                stream size. Other measures that have been used as an indicator of stream size include stream
                order (Fausch et al. 1984) and stream width (Lyons 1992). The decision to use drainage area
                was based on the availability and ease of calculation and relevance to stream size. Stream
                order was viewed as being too coarse (Hughes and Omernik 1981) and stream width is simply
                not representative of stream size given the widespread historical modification of streams
                throughout Ohio. In other regions of the U.S. these and other parameters may be appropriate
                for use in the calibration process. Additional dimensions could include temperature, gradient,
                elevation, and lake acres or shoreline distance. The one concept which continues to surface
                throughout this process is that these are decisions which can only be made reliably by regional
                experts.

                        The plots for each metric were visually examined to determine if any relationship with
                drainage area existed. If a relationship was observed a 95 % line of best fit was determined and
                the area beneath trisected into three equal portions following the method recommended by
                Fausch et al. (1984). Wading and headwaters data was combined for in-common metrics to
                determine the slope of the 95 % line even though scoring for these metrics was performed
                separately; all boat site IBI metrics were calibrated separately. The IBI metric scores (i.e., 5,
                3, or 1) for a sample are determined by comparing the site value to the trisected scatter plots
                constructed from the reference site data base for each applicable metric. Certain meirics that
                showed no positive relationship with drainage area required the use of an alternate trisection
                method. Horizontal 5 % and 95 % lines were determined and the area between trisected. A
                bisection method was used only for the number of individuals metric. For two others (top
                carnivores, anomalies) the reference site data base was examined and scoring criteria
                established following Karr et al. (1986) and Ohio EPA (1987). The resultant 5, 3, and I
                values for these metrics are the same across drainage areas. A similar method of trisection was
                used by Hughes and Gammon (1987) for a modified IBI used in the lower 280 kin of the
                Willamette River, Oregon.

                        The principal measure of macroinvertebrate community performance used by the Ohio
                EPA is the Invertebrate Community Index (ICI) which was originally developed by Ohio EPA
                (Ohio EPA 1987, DeShon, 1995). The ICI is an adaptation of the IBI concept to
                macroinvertebrate cominunities. The ICI consists of 10 structural and functional community
                metrics, each with four scoring categories of 6, 4, 2, and 0 points in order to result in scores
                which were comparable to the fish IBI scores. The point system is structured to operate the

                                                                 28








             same as the IBI. The summation of the individual metric scores (determined by the relevant
             attributes of an invertebrate sample with consideration given to drainage area) results in the ICI
             value. To determine the 6, 4, 2, and 0 values for each ICI metric, the reference site database
             was plotted against drainage area. Each metric was visually examined to determine if any
             relationship existed with drainage area. When it was decided if a direct, inverse, or no
             relationship existed, the appropriate 95 % line was estimated and the area beneath quadrisected.
             Certain percent abundance and taxa richness categories were not quadrisected since the data
             points showed a tendency to clump at or near zero. In these situations, a quadripartite method
             was used where one of the four scoring categories included zero values only, and, in two
             cases, the remaining scoring categories were delineated by an equal division of the reference
             data points.

                   A modified approach was necessary for determining the HELP ecoregion biocriteria.
             The HELP ecoregion is affected by significant and widespread historical land use and stream
             channel modifications dating to the 19th century. Setting the WWH criteria for the IBI and
             MIwb in this ecoregion involved detailed consideration of the extensive and essentially
             irretrievable physical stream habitat and watershed modifications. Based on the Qualitative
             Habitat Evaluation Index scores (Rankin 1989), the field observations of Ohio EPA biologists,
             and the descriptions of land use patterns (Whittier et al. 1987), none of the wading and
             headwaters reference sites in the HELP ecoregion reflected least impacted conditions relative
             to that observed in the other Ohio ecoregions. This distinction is made necessary by the
             widespread degree to which macrohabitats have been altered among the headwater and
             wadeable streams in the HELP ecoregion. Intensive row crop agriculture and attendant
             subsurface drainage practices (i.e., channel maintenance and tiling) have left few if any
             streams that match the intended definition of least impacted. As a result IBI and Mlwb values
             from the wading and headwaters reference sites of this ecoregion reflect these environmentally
             degrading influences. Deriving the WWH wading and headwaters sites biocriteria involved an
             examination of IBI and MIwb results from all sites sampled during 1981-89 (Ohio EPA 1987,
             1989). IBI and Mlwb values that marked the upper 10% (90th percentile) of all sites sampled
             were selected as an alternative to the 25th percentile of the HELP reference sites which yielded
             lower values. The information contained in selected historical descriptions of streams in this
             ecoregion (Meek 1889, Trautman 1981, Kirsch 1895, Trautman 1939, 1981, Smith 1968,
             Trautman and Gartman 1974) was influential in making judgements about attainable WWH
             expectations in this ecoregion. Even with this adjustment the resulting IBI and MIwb criteria
             are the lowest in the state. Although the ICI values are likewise low for the HELP ecoregion
             the primary sampling technique is not nearly as affected by the macrohabitat modifications.
             Thus the 25th percentile from the reference sites was chosen as the WWH criterion for the ICI.
             Establishing biocriteria for the HELP ecoregion is an example of the dilenima posed by
             extensively disturbed areas - maintaining a balance between setting a goal for watershed
             restoration efforts and the pragmatic implications of maintaining present-day socioeconomic
             activities.


             6-7.   Maintenance of the Reference Site Network and Periodic Adjustments

                                                          29








                       The adoption of numerical biological criteria includes the task of maintaining the
               reference data base which includes a planned re-sampling of all sites within a prescribed time
               frame. A concern which is frequently expressed is that by basing aquatic community
               performance expectations on contemporary conditions defined by present day reference sites,
               aquatic life goals are somehow being frozen in time. This is why the concept of continual
               maintenance monitoring must be included as a part of the overall regional reference site
               approach. In Ohio, we have chosen to sample approximately 10% of the reference sites each
               year within the organization of the Five-Year Basin Approach. This will provide an
               opportunity to examine regional background aquatic community performance at periodic
               intervals (e.g., once/ten years) and make appropriate adjustments to the calibration of the
               multi-metric indices, the numerical biological criteria, or both.

               Future Considerations and Potential Improvements
                       Calibration
                       The determination of the 95 % line is one of the most important parts of the calibration
               process. While the line-of-best-fit method is presently accepted (Fausch et al. 1984), it is not a
               strict statistical derivation. As an experimental approach to possibly improve the objectivity of
               the 95 % line determination we applied the technique described by Blackburn et al. (1992) in
               which a series of regression lines are determined across the upper surface of the wedge of
               points that result from the scatter plots of drainage area-dependent IBI metrics. Thus far we
               have determined this for the fish species richness metric. The results indicate a line that is not
               substantially different from the line-of-best-fit method. While this seems to initially confirm the
               line-of-best-fit method it appears to offer important advantages, the most obvious of which is a
               statistically objective method for determining the 95 % line. One important drawback,
               however, is the inability of the statistic to determine when and where the slope of the line
               should change. This was done by visual interpretation for several of the IBI and most of the
               10 metrics.


                       Calibration issues which need further examination include determining the degree of
               convergence between the 5, 3, and 1 lines at the lower drainage areas, the non-linear
               distribution of the scatter plots for the "percent of" metrics, and how to determine scoring for
               metrics which have no apparent relationship with stream size. Other considerations include the
               consistent designation of trophic guilds, tolerance rankings, refined metrics, refined metric
               scoring, and regional calibration. For example, differences exist in the designation of feeding
               and tolerance guilds between states which share similar faunas. In addition, criticism has been
               leveled at intolerant species designations as reflecting rare, threatened, and endangered status
               more so than true environmental tolerance. While we have dealt with most of these issues in
               Ohio, these and other issues will arise elsewhere, thus regional consistency in achieving a
               resolution of these issues will be needed.


                       The Ohio case example represents an effort to derive numerical biocriteria on a state-
               specific basis. This particular effort was constrained to data available or obtainable within the
               state boundaries. Political boundaries, however, seldom coincide with geographical or faunal

                                                             30








             region boundaries. Thus, consideration should be given to an alternative method to establishing
             what we term here as calibration regions. A calibration region is an area with a logical
             commonality with regards to faunal associations, species richness, waterbody type, and major
             drainage networks. Ideally, defining these areas would be done on the basis of regional
             attributes such as faunal similarity, aggregations of ecoregions and sub-ecoregions, and other
             relevant factors. For example a calibration region for the Midwestern U.S. might include the
             northern portion, or subsets therein, of the Ohio River drainage basin (all sub-drainages north
             of the Ohio River mainstem) which would include portions of five states. In order to begin
             coping with the regionally unique aspects of faunal composition, stream and river
             characteristics, and watershed characteristics, this type of framework seems essential if we are
             to maximize the utility and validity of biocriteria as water resource management decision
             criteria. Such a regional framework, while fostering interstate cooperation, would also provide
             a scientific forum for indicator selection and development, methods standardization, reference
             site selection, and calibration of multi-metric evaluation mechanisms. Stratification beyond
             this geographic level could be accomplished through the use of ecoregions, sub-ecoregions,
             and tiered aquatic life use designations (i.e., designated uses). This framework would also be
             adaptable to emerging national monitoring frameworks such as the U.S. EPA Environmental
             Monitoring and Assessment Program (EMAP), the U.S. Geological Survey National Water
             Quality Assessment (NAWQA), and the National Biological Survey (NBS). In fact, this seems
             to be a logical prerequisite to the analysis of the data from these efforts. Finally, regional
             calibration areas would provide a means of jurisdiction over the logistical and technical issues
             which inevitably arise within national monitoring and assessment programs.

                     Biological Index Variability
                     A frequent criticism of ambient biological data is that it is too variable to function as a
             reliable component of surface water resource assessment. Natural biological systems are
             indeed variable and seemingly noisy, but no more so than the chemical and physical
             components that also exist within aquatic ecosystems. Certain dimensions of ambient
             biological data are quite variable, particularly population or sub-population level parameters.
             Single dimension community measures can also be quite variable. The new generation
             community evaluation mechanisms such as the IBI and ICI are sufficiently redundant so as to
             compress and dampen some of the aforementioned variability. Rankin and Yoder (1990)
             examined replicate variability of the IBI from nearly 1000 sites throughout Ohio and found it to
             be quite low at least impacted sites. Coefficient of variation (CV) values were less than 10% at
             IBI ranges indicative of exceptional biological performance and less than 15 % for the good
             performance range. This is lower than the variability reported for chemical laboratory
             analyses and inter-laboratory bioassays (Mount 1987). Variability as portrayed by CV values
             increased at IBI ranges indicative of increasingly impaired biological performance. Low
             variability was also found for the ICI with a CV of 10. 8 % for 19 replicate samples at a
             relatively unimpacted test site (DeShon, 1995). The variability of the MIwb was determined to
             be on the order of +0.5 MIwb units (Ohio EPA 1987). Other investigators have reported
             similarly low variability with other biological indices (Davis and Lubin 1989, Stevens and
             Szczytko 1990). Fore et al. (1993) used different statistical techniques and determined a

                                                              31








                variability of +3 1131 units using the Ohio database. Cairns (1986) suggested that differences in
                variability rather than differences in averages or means might be the best measure of stress in
                natural systems. Variability must begin to be recognized as a part of the signal rather than
                noise alone (Karr 1991). Not only is the variability of the measures used as biological criteria
                low, the degree of variability encountered can also be a useful assessment and interpretation
                tool.


                        Ohio EPA has addressed the variability inherent to biological measures in three general
                ways:


                1.      Variability is compresse through the use of multi-metric evaluation mechanisms such
                        as the IBI and ICL


                2.      Variability is stratified by the tiered use classification system, ecoregions, biological
                        index calibration, and site type.

                3.      Variability is controlled through standardized sampling procedures which address
                        seasonality, effort, replication, gear selectivity, and spatial concerns.


                Initial Decisions and Other Considerations
                        There are a number of fundamental decisions which need to be made early in the
                development of biocriteria. This is a critical juncture in the process since these initial
                decisions will determine the overall effectiveness of the effort well into the future. Decisions
                about which sampling methods and gear to use, seasonal considerations, which organism
                groups to monitor, which parameters to measure, which level of taxonomy to use, etc. will
                need to be made. The axiom follows           when in doubt choose to take more measurements
                than seem necessary at the time since information not collected is impossible to retrieve at a
                later date". This does not apply equally to all parameters. For example, seasonality is a well
                understood concept, therefore it is not necessary to sample in multiple seasons for the sake of
                data redundancy. However, parameters which require little extra effort to acquire should be
                included until enough evidence is amassed to evaluate its relative worth. One example in Ohio
                is with external anomalies on fish. A decision was made to record this information even
                though it was not immediately apparent what use this information would have. This one
                parameter has proven over time to be one of our most valuable assessment tools. For
                macroinvertebrates the issue of identifying midges to the genus/species level (as opposed to the
                family level) proved likewise to be a far sighted decision given the value of this group in
                diagnosing impairments. Samples could have been archived for later processing, but the
                logistical burdens that this would entail later on are even more undesirable.

                        Another important consideration is assuring that qualified and regionally experienced
                staff are available to implement the monitoring and assessment activities. Ecological
                assessment is no less in need of skill and experience than are other professions. However,
                biological field assessment is somewhat unique in that an equivalent level of expertise is

                                                                 32








             needed in the field since many of the critical pieces of information are recorded and, more
             importantly, interpreted there. There is simply no substitute for direct experience in the field -
             this is not a job to be left to technicians. In addition, it is only prudent that the same
             professional staff who collect the field data also interpret and apply the information derived
             from the data in a "cradle to grave" fashion. Thus the same staff who perform the field work
             also plan that work, process the data into information, interpret the results, and apply the
             results via assessment and reporting. Such staff, particularly the more experienced ones, also
             contribute to policy development.

             Logistics and Costs
                    The approach used by Ohio EPA to collect macroinvertebrate and fish community data
             is intended to secure an adequate sample, but not necessarily an exhaustive inventory. Fish
             relative abundance data is collected using standardized, pulsed D.C. electrofishing techniques.
             In an analysis of resources expended during FFY (Federal Fiscal Year; October I - September
             30) 1987 and  1988 the following was revealed:

                            8.44 WYE (work year equivalents) were used to collect 1277 samples at 617
                            sites.
                    0       An average of 0.014 WYE or 29.1 hours/site were expended to plan, collect,
                            analyze, interpret data, and produce reports at an average cost of $740/site.
                            This translates into 1-3 hours/sample with a field crew sampling 3-6 sites/day
                            by working 10-14 hours/day.
                            Post-field season laboratory effort ranges from 1-3 weeks.
                            A field crew consists of one full-time biologist and two interns.

                    The approach used by Ohio EPA to collect macroinvertebrate community data-is
             intended to secure an adequate sample, but not an exhaustive inventory of all taxa possible.
             Relative abundance data is collected using a standardized approach (artificial substrates;
             DeShon (1995). In an analysis of resources expended during FFY 1987 and 1988 the
             following was revealed:

                    0       5.02 WYE (work year equivalents) were used to collect 323 samples at 323
                            sampling sites, setting or retrieving 4-6 sites per day.
                    0       An average of 0.015 WYE or 33.2 hours/site were expended to plan, collect,
                            analyze, interpret data, and to produce reports at an average cost of $824/site.
                    0       Laboratory effort is 12-20 hours/sample for artificial substrates and 2-6 hours
                            for qualitative samples only.
                    0       A field crew consists of one full-time biologist and one intern.

                    Concern is frequently expressed not only about the practical utility of biological field
             data, but the resources needed to implement such programs (Loftis et al. 1983, U.S. EPA
             1985). Whole effluent toxicity evaluation has been advocated partly because it is viewed as
             more cost-effective than biological field evaluations (U.S. EPA 1985). Our experience with

                                                            33








                using a standardized and systematic application of biological field monitoring techniques
                integrated with the traditional chemical/physical and bioassay assessment techniques allows a
                detailed comparison of the costs involved with each component. Out of nearly 100 WYE
                (Work Year Equivalents) that were devoted to surface water monitoring and laboratory
                activities within the Division of Water Quality Planning and Assessment in FFY (Federal
                Fiscal Year) 1987 and 1988, 19.34 WYE or just over 19 % of the total was devoted to ambient
                biological monitoring. When considered on the basis of agency-wide water programs this
                percentage is approximately 6%.

                        Table 4.2.1 gives the unit cost of the four monitoring components that are being
                compared and evaluated. Costs are broken down by sample collection, laboratory analysis,
                individual test, and evaluation as appropriate for each component. Included in the cost figures
                are all equipment, supplies, logistical, administrative, data analysis, and interpretation
                activities. Chemical/physical water quality costs were derived from grab samples taken from 3
                to 8 times at each site during surnmer-fall low flow periods (mid-June through mid-October).
                Bioassay costs were on the tests routinely performed by Ohio EPA: 48 hour screening tests,
                48 and 96 hour definitive tests, and seven-day acute/chronic tests. The seven-day tests were
                further subdivided between those analyzing daily composites (i.e., seven-day renewal test) and
                those designed to test one 24-hour "megagrab" sample.

                        An initial comparison of the cost of each component is evident from an examination of
                Table 4.2. 1. Fish and macroinvertebrate evaluation costs/site are comparable. Obtaining
                chemical/physical water quality data was approximately twice that of either biological method
                alone, but only 5 % more than both organism groups together. Comparison of bioassay costs
                was most appropriately done on a point source entity evaluation basis. For example, three to
                six biological sites are usually required to evaluate the impact from a single point source (a
                cost of $4692 to $9384 for both fish and macroinvertebrates), whereas a definitive and/or
                seven-day bioassay test costs $1848 and $3052, respectively. However, these bioassay costs
                represent those for a single test, not a complete, three test evaluation. U.S. EPA (1985)
                protocols specify three tests per evaluation per entity making the bioassay evaluation cost
                $5544 for a definitive evaluation and $9156 for a seven-day evaluation. Thus, sampling both
                fish and macroinvertebrates is comparable to a definitive or seven-day bioassay evaluation, and
                more so if a seven-day renewal test is employed.

                        Using an example situation, the cost to evaluate three entities discharging to a small
                river for acute and chronic toxicity using the seven-day static test would be $27,468 ($54,954
                for a seven-day renewal test). Fish and macroinvertebrates sampled at 18 locations in the
                mainstem would cost $28,152. Furthermore, some of the 18 biological sampling locations
                would also be devoted to monitoring influences other than toxicity from point sources. For
                example the influence of factors that exert their effects by means other than toxicity (e.g.,
                habitat, sediment, nutrient enrichment, flow alterations, low dissolved oxygen, etc.) will be
                apparent in the biological data and such results should play a key role in decision-making about
                water quality based effluent limits and other management needs. In this example, it was

                                                                34








             estimated that 12 of the biological sites would be necessary to determine the cumulative impact
             of the three point sources which results in a comparative biological component cost of
             $18,768. It is recognized that the chemical-specific and toxicity evaluations perform a
             uniquely essential function in attempting to separate relative contributions from interacting
             sources. This is an oft cited shortcoming of biosurvey results although response signatures are
             discernible in the data (Yoder 1991, Yoder and Rankin 1995a). The quality of the eventual
             decisions about water quality standards and discharge limitations (chemical or otherwise)
             would suffer significantly without the information provided by an integrated evaluation
             including chemical, biological, and toxicity measures (Yoder 1991).

                    States that do not operate extensive ambient bioassessment networks will need to be
             prepared for some rather sizeable start-up costs. While the cost analysis incorporated start-up
             costs for equipment and supplies, these were amortized over 5 or 10 years depending on the
             expected life of an item. Start-up equipment and supplies, for most states, could total from
             $200,000 to $500,000 depending on the number of field crews involved.

             Data Management and Information Processing
                    Once field data is collected, processed, and finalized the next step is to reduce the data
             to scientifically and managerially useful information. The principal Ohio EPA data
             management system for fish, macroinvertebrate, and habitat data (Ohio ECOS) includes
             storage, processing, and analysis routines. Once data is tabulated in the field (fish and habitat)
             and laboratory (macroinvertebrates) and documented via chain of custody procedures, the data
             is entered directly into the electronic database. Basic information includes the field crew,
             waterbody name, date, and time. Site location is indicated by river mile (distance upstream
             from mouth) and latitude/longitude, both of which are determined from USGS 7.5 minute
             topographic maps. A basin-river code system is used to electronically identify individual
             streams, rivers, and lakes. Sampling information includes method or gear type and other
             information relevant to the use of each. Ohio ECOS generates data summaries and reports for
             a variety of community measures, community composition, or individual species/taxon
             analyses.


             Conclusions
                    Biological criteria are an emerging and increasingly important issue for EPA, the states,
             and the regulated community. The use of biocriteria through bioassessments is growing
             nationwide as more states and local organizations shift their monitoring and assessment efforts
             in this direction. However, much remains to be done, particularly in the area of national and
             regional leadership. Technical guidance and expertise is needed to ensure a nationally
             consistent and credible approach and to resolve outstanding technical concerns listed by Yoder
             and Rankin (1995a). Resolving outstanding policy issues such as EPA's policy of independent
             applicability needs to be accomplished in such a manner as to encourage, not discourage, states
             to participate. In an era of declining government resources ways to accomplish the "increases"
             needed in biological monitoring to support the biocriteria approach must be developed. Based
             on our experience in Ohio the staffing of state programs should be a minimum of one work

                                                             35








                year equivalent for every 1200 miles of perennial streams and rivers. This estimate may vary
                by region and should additionally incorporate lake acres in states with a predominance of this
                water body type (Yoder and Rankin 1995b). The potential for bioassessments and biocriteria
                to modify the present capital and resource intensive system of tracking environmental
                compliance on a pollutant specific basis needs to be considered by EPA. This should prove to
                be a more cost and information effective approach to managing the nation's water quality
                programs.


                        Ohio EPA has been monitoring the condition of Ohio's surface waters intensively since
                the late 1970s. Biological assessment has always been emphasized and this was further
                formalized with the adoption of numerical biological criteria in 1990. Beginning in 1990 Ohio
                instituted a "5-Year Basin" approach to monitoring and NPDES permit reissuance. This
                schedule has been devised so that monitoring data is collected in advance of permit reissuance,
                implementation of best management practices for nonpoint sources, or other management
                actions which benefit from monitoring information. The 15 plus years of using an integrated
                biosurvey approach to monitor major sources of pollution has put Ohio EPA in a position to
                determine the effectiveness of water quality-based pollution controls. This effort has resulted
                in a shift away from a sole reliance on regulatory and administrative activities as the principal
                measures of success to the inclusion of environmental results oriented measures.


                        The eventual attainment of the CWA goal of biological integrity means more than
                achieving a higher level of species diversity, numbers, and/or biomass. In fact there are
                situations when increases in any one of these attributes may signify degradation. Managers
                also must strive for more than the protection target species, an effort which sometimes receives
                a disproportionate share of scarce resources. Merely conserving imperiled species, while
                nonetheless essential, is alone insufficient for maintaining and restoring biological integrity.
                Conservation policy needs to promote management practices which maintain and restore
                biological integrity, prevent endangerment, and enhance the recovery of species and
                ecosystems (Angermier and Williams 1993). The goals of water resource management must
                begin to focus additionally on the maintenance of self-sustaining and functionally healthy
                aquatic communities. Achieving this state of aquatic ecosystem integrity will "bring along"
                these other goals as well since functionally healthy communities include the elements of
                biodiversity and rare species that the more narrowly focused management efforts are striving to
                attain. Biological criteria can and should play an important role in meeting these challenges.

                        Ohio EPA has placed a high emphasis on monitoring as being much more than a data
                gathering activity by integrating the results into the entire water quality management process.
                Because of the investment made in monitoring over the past 15 years, we are now reaping
                benefits by being able to quantify improvements resulting from our regulatory efforts of the
                past 20 years, producing accurate estimates of goal attainment and non-attainment (Rankin et
                al. 1992), and in being able to approach new and emerging issues from a sound environmental
                basis. Invaluable insight into the potential uses of biological criteria, their advantages, and
                their limitations has been gained. A wide array of different types and degrees of

                                                                 36









            environmental perturbation (both chemical and nonchemical) have been observed and
            evaluated. This experience provided the basis for many of the concepts and findings that are
            presented herein.

                   A growing body of information shows that other factors in addition to chemical water
            quality are responsible for the continuing decline of surface water resources in many cases
            (Judy et al. 1984, Rankin and Yoder, 1990). Because biological integrity is affected by
            multiple factors in addition to chemical water quality, controlling chemicals alone does not in
            itself assure the restoration of biological integrity (Karr et al. 1986). If we are to make
            progress in the restoration and protection of aquatic ecosystems our concerns must incorporate
            a broader focus on the water resource as a whole. The concepts inherent to biological integrity
            implicitly include such holism. Whole effluent toxicity testing offers an improvement over a
            strictly chemical approach, but alone lacks the ability to broadly assess ecosystem effects,
            particularly those caused by physical, episodic, and nontoxic chemical impacts. Biological
            criteria and the attendant biosurvey approach to monitoring and assessment provides a means
            to incorporate the broader concept of water resource integrity while preserving the traditional
            chemical/physical and toxicological approaches of the past three decades.




                                                       References


            Angermier, P.L. and J.E. Williams. 1993. Conservation of imperiled species and
                   reauthorization of the endangered species act of 1973. Fisheries 18(7):34-38.

            Ballentine, R. K. and Guarrie L. J., Eds., 1975. The integrity of water: a symposium, U.S.
                   Environmental Protection Agency, Washington, D.C., 230 pp.

            Blackburn, T. M., Lawton, J. H., and Perry, J . N., 1992. A method for estimating the slope
                   of upper bounds of plots of body size and abundance in natural animal assemblages,
                   Oikos, 65: 107-112.


            Benke, A.C. 1990. A perspective on America's vanishing streams. J. N. Am. Benth. Soc.,
                   9 (1): 77-88.

            Berthouex P.M & I. Hau. 1991. Difficulties related to using extreme percentiles for water
                   quality regulations. Research journal of the water pollution control federation
                   63:873-879.


            Cairns, J., Jr. 1986. Freshwater. In Proceedings of the workshop on cumulative
                   environmental effects: a binational perspective. CEARC, Ottawa, Ontario and NRC,
                   Washington, D.C. 175 pp.


                                                           37









                Davis, W.S. and A. Lubin. 1989. Statistical validation of Ohio EPA's invertebrate
                       community index, pp. 23-32. in Davis, W.S. and T.P. Simon (eds.). Proc. 1989
                       Midwest Poll. Biol. Mtg., Chicago, 111. EPA 905/9-89/007.

                DeShon, J.D. 1995. Development and application of the invertebrate community index (ICI),
                       pp. 217-243. in W.S. Davis and T. Simon (eds.). Biological Assessment and Criteria:
                       Tools for Risk-based Planning and Decision Making. Lewis Publishers, Boca Raton,
                       FL.


                Fausch, K. D., Karr J. R., and P. R. Yant. 1984. Regional application of an index of biotic
                       integrity based on stream fish communities. Transactions of the American Fishery
                       Society 113: 39-55.

                Fore, L.S., J.R. Karr, and L.L. Conquest. 1993. Statistical properties of an index of biotic
                       integrity used to evaluate water resources. Can. J. Fish. Aquatic Sci.

                Gakstatter, J. and others. 1981. A recommended approach for determining biological
                       integrity in flowing waters. U.S. EPA, Corvallis, Oregon, 26 pp.

                Herricks. E.E. and D.J. Schaeffer. 1985. Can we optimize biomonitoring? Environmental
                       Management 9: 487-492.

                Hughes, R. M., D. P. Larsen, and J.M. Omernik. 1986. Regional reference sites: a method
                       for assessing stream pollution, Environmental Management, 10: 629.

                Hughes, R. M. and J. R. Gammon, 1987. Longitudinal changes in fish assemblages and water
                       quality in the Willamette River, Oregon, Transactions of the American Fishery Society,
                       116, 196.


                Hughes, R. M., Gakstatter, J. H., Shirazi, M. A., and Omernik, J. M., 1982. An approach
                       for determining biological integrity in flowing waters, in Inplace resource inventories:
                       principles and practices, a National workshop, Braun, T. B., Ed., Society of American
                       Foresters, Bethesda, MD.

                Hughes, R. M., and Omernik, J. M. 1981. Use and misuse of the terms watershed and
                       stream order, in Warmwater Streams Symposium, Krumholz, L., Ed., American
                       Fishery Society, Bethesda, MD.

                Judy, R. D., Jr., P. N. Seely, T. M. Murray, S.C. Svirsky, M. R. Whitworth, and L.S.
                       Ischinger. 1984. 1982 national fisheries survey, Vol. 1. Technical Report Initial
                       Findings. U.S. Fish and Wildlife Service, FWS/OBS-84/06.

                Karr, J. R. 1991. Biological integrity: A long-neglected aspect of water resource

                                                             38








                   management. Ecological Applications l(l): 66-84.

            Karr, J. R., K. D. Fausch, P. L. Angermier, P. R. Yant, and I. J. Schlosser. 1986.
                   Assessing biological integrity in running waters: a method and its rationale. Illinois
                   Natural History Survey Special Publication 5: 28 pp.

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

            Kirsch, P. H. 1895. A report upon investigations in the Maumee River during the summer of
                   1893, Bulletin of the U. S. Fishery Commission, 16, 315.

            Larsen, D. P., Omernik, J. M., Hughes, R. M., Rohm, C. M., Whittier, T. R., Kinney, A.
                   J., Gallant, A. L., and D. R. Dudley. 1986. The correspondence between spatial
                   patterns in fish assemblages in Ohio streams and aquatic ecoregions. Environmental
                   Management 10: 815-828.

            Leonard, P. M., and D.J. Orth 1986. Application and testing of an index of biotic integrity
                   in small, coolwater streams. Trans. Am. Fish. Soc. 115: pp. 401.

            Loftis, J. C., Ward, R. C., and Smillie, G. M. 1983. Statistical models for water quality
                   regulation, Journal of the Water Pollution Control Federation, 55, 1098.

            Lyons,J. 1992. The length of stream to sample with a towed electrofishing unit when fish
                   species richness is estimated. N. Am. J. Fish. Manag. 12:198-203.

            Lyons, J. 1992. Using the index of biotic integrity (IBI) to measure environmental quality in
                   warmwater streams of Wisconsin. Gen. Tech. Rep. NC-149. St. Paul, MN: USDA,
                   Forest Serv., N. Central Forest Exp. Sta. 51 pp.

            Meek, S. E. 1889. Notes on a collection of fishes from the Maumee valley, Ohio,
                   Proceedings of the U. S. National Museum, 2, 435.

            Mount, D.I. 1987 (unpublished). Comparison of test precision of effluent toxicity tests with
                   chemical analyses. U.S. EPA, Env. Res. Lab., Duluth, Minnesota.

            Ohio Environmental Protection Agency. 1987. Biological criteria for the protection of
                   aquatic life: volume 11. Users manual for biological field assessment of Ohio surface
                   waters. Division of Water Quality Monitoring and Assessment, Surface Water Section,
                   Columbus, Ohio.

            Ohio Environmental Protection Agency. 1989. Biological criteria for the protection of
                   aquatic life. Volume III: Standardized biological field sampling and laboratory methods

                                                         39








                       for assessing fish and macroinvertebrate communities. Ohio EPA division of water
                       quality planning and assessment, Ecological assessment section, Columbus, Ohio.

               Ohio Environmental Protection Agency. 1992a. Biological and habitat investigation of greater
                       Cincinnati area streams: the impacts of interceptor sewer line construction and
                       maintenance, Hamilton and Clermont Counties, OH, OEPA Tech. Rep. EAS/1992-5-1.

               Ohio Environmental Protection Agency. 1992b. Ohio water resource inventory. Volume I,
                       status and trends. E.T. Rankin, C.O. Yoder and D.A. Mishne (editors). Ohio EPA,
                       Division of water quality planning and assessment, Ecological assessment section,
                       Columbus, Ohio.


               Ohio Environmental Protection Agency. 1997. Ohio water resource inventory, volume I,
                       summary, status, and trends. Rankin, E.T., Yoder, C.O., and Mishne, D.A. (eds.),
                       Ohio EPA Tech. Bull. MAS/1996-10-3-I, Division of Surface Water, Columbus, Ohio.
                       190 pp.

               Plafkin, J. L. and others. 1989. Rapid Bioassessment Protocols for use in rivers and streams:
                       benthic macroinvertebrates and fish. EPA/444/4-89-001. U.S. EPA. Washington, D.C.

               Rankin, E. T. 1989. The qualitative habitat evaluation index (QHEI), rationale, methods,
                       and application, Ohio EPA, Division of Water Quality Planning and Assessment,
                       Ecological Assessment Section, Columbus, Ohio.

               Rankin, E. T. and C. 0. Yoder. 1990. A comparison of aquatic life use impairment detection
                       and its causes between an integrated, biosurvey-based environmental assessment and its
                       water column chemistry subcomponent. Appendix I, Ohio Water Resource Inventory
                       (Volume 1), Ohio EPA, Div. Water Qual. Plan. Assess., Columbus, Ohio. 29 pp.

               Rankin, E.T., C.O. Yoder and D.A. Mishne. 1992. Ohio water resource inventory, Volume
                       1: Summary, Status and Trends, 1992. Division of Water Quality Planning and
                       Assessment, Ecological Assessment Section. Columbus, Ohio.

               Smith, P.W. 1968. An assessment of changes in the fish fauna of two Illinois rivers and its
                       bearing on there future. Transactions Illinois State Academy of Science, 61(l): 31-45.

               Stevens, J.C. and S.W. Szczytko. 1990. The use and variability of the biotic index to
                       monitor changes in an effluent stream following wastewater treatment plant upgrades,
                       pp. 33-46. in Davis, W.S. (ed.). Proc. 1990 Midwest Poll. Biol. Mtg., Chicago, Ill.
                       EPA-905-9-90/005.


               Trautman, M. B., 1939. The effects of man-made modifications on the fish fauna in Lost and
                       Gordon Creeks, Ohio, between 1887-1938, Ohio Journal of Science, 39, 275.

                                                             40









            Trautman, M.B. 1981. The Fishes of Ohio. Ohio State University Press. Columbus, Ohio.
                   782 pp.

            Trautman, M.B. and R.K. Gartman. 1974. Re-evaluation of the effects of man-made
                   modifications of Gordon Creek between 1887 and 1973 and especially as regards its
                   fish fauna. Ohio J. Sci. 74(3): 162-173.

            US Environmental Protection Agency. 1985. Technical support document for water
                   quality-based toxics control. EPA-440-4-85-003. U.S. Environmental Protection
                   Agency, Office of water, Washington, D.C.

            U.S. Environmental Protection Agency. 1991. Environmental monitoring and assessment
                   program. EMAP - surface waters monitoring and research strategy - fiscal year 1991,
                   EPA/600/3-91/022. Office of Research and Development, Environmental Research
                   Laboratory, Corvallis, OR. 184 pp.

            Whittier, T. R., Larsen, D. P., Hughes, R. M., Rohm, C. M., Gallant, A. L. and J. R.
                   Omernik. 1987. The Ohio stream regionalization project: A compendium of results.
                   Report No. EPA-600/3-87/025. U. S. Environmental Protection Agency, Corvallis,
                   Oregon.

            Yoder, C.O. 1991. The integrated biosurvey as a tool for evaluation of aquatic life use
                   attainment and impairment in Ohio surface waters. Pages 110-122 in Biological
                   criteria: research and regulation, Proceedings of a symposium, 12-13 December 1990,
                   Arlington, Virginia. EPA-440-5-91-005. U.S. Environmental Protection Agency,
                   Office of water, Washington, D.C.

            Yoder, C. 0. and E. T. Rankin. 1995a. Biological criteria program development and
                   implementation in Ohio, pp. 109-144. in W. Davis and T. Simon (eds.). Biological
                   Assessment and Criteria: Tools for Water Resource Planning and Decision Making.
                   Lewis Publishers, Boca Raton, FL.


            Yoder, C.O. and E.T. Rankin. 1995b. Biological response signatures and the area of
                   degradation value: new tools for interpreting multimetric data, pp. 263-286. in W.
                   Davis and T. Simon (eds.). Biological Assessment and Criteria: Tools for Water
                   Resource Planning and Decision Making. Lewis Publishers, Boca Raton, FL.









                                                        41







                    Table 4.2.1     Cost comparison of macroinvertebrate community and fish community
                    evaluations with chemical/physical grab sampling and acute and acute/chronic bioassay tests.


                                     Sample                 Analytical                 Cost per                   Cost per
                                  Collection            Cost (Laboratory)            Test /Sample               Evaluation



                                                         Macroinvertebrate Communi1y

                                              Artificial Substrates (includes qualitative sample)
                                      N/A                       $397                      $824                      $824
                                                             Qualitative Sample Only
                                      N/A                       $150                      $275                      $275


                                                                 Fish Community

                                                                 Cost per sample
                                      N/A                       N/A                       $340                      $340
                                                                    Cost per site
                                      N/A                       N/A                       $340                      $740


                                            Chemical/Physical Water Qualfty (4.6 saMples/site

                                     $1124'                     $5292                     $359                     $1653

                                                                      Bioassay

                                                                      Screening'
                                      $261                      N/A                       $1191d                   $3573
                                                                      Definitive  4
                                      $261                      N/A                       $1848                    $5544
                                                                      Seven-day5
                                      $261                      N/A                       $3052                    $9156
                                                                      Seven-day6
                                      $1973                     N/A                       $6106                   $18318

                    'includes cost of sample collection and data analysis only; based on an average frequency of 4.6 samples/site in
                    1987 and 1988;
                    2analytical costs based on each sample being analyzed for 5 heavy metals ($7.00 ea.), 4 nutrients ($10.00 ea.),
                    COD or BOD ($20.00 ea.), and 2 additional parameters ($20.00 for both); $115 per sample;
                    348 hour exposure to determine acute toxicity;
                    448 and 96 hour exposure to determine LC50 and EC50;
                    5seven-day exposure to determine acute and chronic effects using a single 24-hour sample; cost based on analysis
                    of one pipe only; costs for chemical analyses in sole support of the test are not included.
                    6seven-day exposure using a composite sample collected daily (renewal); other factors apply.

                                                                          42













                                                              00
                                                      n
                                                                           0
                                                                           0    -^0


                                                           0


                                                                        0


                                                                                   0





                                                                  0




                                                                                 0 Headvater Sites
                                                                                    Wading Sites
                                                                                    Boat Sites


                            1. Select & sample reference sites



             Figure 4.2.1 Key steps in the process of establishing biological criteria to evaluate
                             Ohio rivers and streams.


















                                                                43







                Figure 4.2.1 (cont.)




                           40
                                   Wading & Headwater Sites
                U1
                LU         30                                                                         5   7.

                LU
                CL         20
                                                                                                      3

                <          10
                0

                             0
                                                       10                      100                       1000


                                                 DRAINAGE AREA (SQ MI)

                             11. Calibration of IBI metrics












                                                               44













           Figure 4.2.1 (cont.)


           /it. Calibrated IBI modified for Ohio waters



                    Metric                5        3           1

                Number of Species      Varies x Drainage Area
                No. of Darter Spp.     Varies x Drainage Area
                No. of Sunfish Spp.     >3        2-3        <2
                No. of Sucker Spp.     Varies x Drainage Area
                Intolerant Species
                  >100 sq. mL           >5        3-5        <3
                  <100 sq. mL          Varies x Drainage Area
                % Tolerant Species     Varies x Drainage Area
                %Omnivores             <19        19-34     >34
                %insectivores
                  <30 sq. mL           Varies x Drainage Area
                  >30 sq. mL           >55        26-55     <26
                %Top Carnivores         >5         1-5       <1
                %Simple Lithophils     Varies x Drainage Area
                %DELTAnomalies         >1-3       0.5-1.3    <0.5
                Relative Abundance     >750       200-750 <200


















                                        45









            Figure 4.2.1 (cont.)




                                REGIONAL REFERENCE SITES:
                                      IBI (Wading Site Type)
                  60





                  so





                  40





                  30



                  20          0

                           HELP         /P      EOLP      WAP       ECBP
                  10      (n = 20)   (n = 50)   (n = 64) (n = 105) (n = 155)


                    0
                                          ECOREGION

            IV. Establish ecoregional pattems/expectations
                                                                   qi6







                                                    46








                  Figure 4.2.1 (cont.)





                                           Fish - Boat Sites                                        Fish - Wading Sites

                                                                                                              IBIlIwb



                                     .-3418.6
                                                                    4018.7                  . . . . . . . . . .                3917.9



                                                             ........ ..
                                                                           %                                                ......   %5
                                       4218.5               M 4018.6                               4018.3




                                                                                                  4018.1                             501
                                                                          481                                                               E9W
                                                                                ERW
                                                                          9.6


                                     Fish      Headwater Sites                                       Macroinvertebrates

                                                      JBI                                                          ICI


                                         28
                                                                     40                                                          34



                                                                                                                                   w..

                                                             %  .....    %                                                % ...
                                                             R
                                          40                                                         36                        36

                                                                       M.


                                                                       50] EWH                                                       46@ ERW
                                                                     F                                                             F
                      El   Huron Erie Lake Plain - HELP                Eastern-Ontario Lake Plain - EOLP            Eastern Corn Bell Plains - ECBP
                      EJ   interior Plateau - IP                       Western Allegheny Plateau - RAP


                        V. Derive numeric biocriterfalcodify in WQS
                                                                                                                                 34

































                                                                                            47










      Figure 4.2.1 (cont)






              SCIOTO RIVER: 1979-1991

                 JACKSON       Bowers
         60      PIKE WWTP     Landrill
              WHI=  SOUTHERLY WWTP 4
            STREEi                    EWH Criterfon
                                       (IBI = 48)
         50                @A9
              "on r Q
         40      boo@
                                      WWH Critedon
         30              Walnut Cr.    (IBI = 42)


         20                           --jj@- 1979
             Impounded                    1991
         12
          140   130 120  110   100  90
                   RIVER MILE



      V1. Numeric biocriteria used in assessments











                         48










            4.3 ASSESSMENT OF THE IMPACT OF WATERSHED DEVELOPMENT ON THE
                             NURSERY FUNCTIONS OF TIDAL CREEK HABITATS.


              A. F. Holland, G. H. M. Riekerk, S. B. Lerberg, L. E. Zimmerman, and D. M. Sanger,

                 S. C. Dept. of Natural Resources, Marine Resources Research Inst., Charleston, SC

                   Meandering shallow tidal creeks and the associated intertidal salt marshes are dominant
            features of Southeastern estuaries and provide critical nursery habitat for many species of fish,
            crabs, and shrimp. These shallow tidal creeks are also conduits through which many pollutants
            enter estuaries, with creek sediments serving as a repository for toxic chemicals and other
            contaminants discharged into estuarine environments.

                   Resource management and regulatory agencies responsible for protecting estuarine
            environments do not know if the policies and programs they have implemented are adequately
            protecting tidal creek habitats. These agencies also lack the knowledge required to restore
            degraded creek habitats.

                   In 1994, the South Carolina Marine Resources Research Institute initiated a study,
            called the Tidal Creek Project (TCP), to develop the information needed to: (1) characterize
            and define the ecological values of tidal creeks and associated marsh habitats;(2) identify the
            major pollution threats to tidal creeks associated with watershed development; (3) assess the
            cumulative impacts of watershed development on tidal creek habitats including the living
            resources that use them as nurseries; and (4) develop environmental quality criteria for
            sustaining tidal creek nursery functions. This study was funded jointly by the Charleston
            Harbor Project (1994-1996) and the Marine Recreational Fisheries Advisory Board (1995-
            1996).

            Approach
                   The general study approach used was to sample and contrast the physical, chemical,
            and ecological characteristics of tidal creeks draining relatively pristine, undeveloped
            watersheds (called reference creeks) and creeks draining highly developed watersheds (called
            developed creeks). Associations between physical, chemical, and ecological characteristics of
            creeks and the various types of human development and land cover that occurred were also
            evaluated. This sampling approach is generally referred to as the comparative watershed
            assessment approach.

                   Creeks in the developed watershed class were selected to represent the major types of
            development that occur in the South Carolina coastal zone including: (1) industrial
            development, (2) urban development, (3) suburban development, and (4) agriculture. Creeks
            in the reference class were either predominately forested and/or salt marsh. Watersheds of
            similar sizes and physical characteristics were evaluated from both the reference and developed
            classes. The tidal creeks sampled included representatives of the major salinity zones (brackish

                                                          49








                water to near full strength sea water) and sediment types (sand, mixed, and mud sediments)
                that occur in South Carolina.


                        The accuracy, precision, representativeness, completeness, and comparability of the
                information produced by the TCP were evaluated through a formal Quality Assurance (QA)
                Program. This QA program was designed to ensure the information produced by the TCP was
                adequate for addressing study objectives and developing environmental policy. A
                computerized relational data base system was also established to facilitate efficient storage,
                retrieval, and analysis of the data produced. This data base provides a means through which
                the data can be accessed by other researchers or regulatory and resource management agencies.
                A copy of the TCP data base will be provided to state and federal agencies upon request.

                Findings
                        Salinity was identified as the major factor controlling the distribution and abundance of
                living resources in shallow tidal creeks. Salinity fluctuated over greater ranges and was
                generally more variable in creeks with developed watersheds than in reference creeks. The
                increased variability and extreme fluctuations in the salinity of developed creeks appeared to be
                related to the "flashier" runoff associated with the increased amount of impervious surface in
                developed watersheds (e.g., roofs, roads, parking lots). Creeks that were dominated by salt
                marshes and limited freshwater inputs had relatively stable salinity distributions.

                        Dissolved oxygen (DO) concentration is a fundamental requirement for maintaining
                balanced, indigenous populations of fish, shellfish, and other aquatic biota in shallow tidal
                creeks. Pollution related decreases in DO is generally considered to be the greatest threat to
                the environmental quality of estuaries. DO in tidal creeks fluctuated with phase of the moon,
                time of day, and stage of the tide. DO in both reference and developed creeks frequently did
                not meet state water quality standards (4 mg/1), with the lowest and most stressftil DO to living
                aquatic resources occurring during early morning and night-time low tides. DO in developed
                creeks was less predictable and had larger amounts of unexplained variance than DO in
                reference creeks. About 68 % of the variance in the DO of reference creeks was associated
                with natural cycles. Only about 20% of the variance in DO of developed creeks could be
                attributed to natural cycles. Living resources inhabiting developed creeks were exposed to
                stressful low DO more frequently than living resources inhabiting reference creeks. Tidal
                creek ecosystems in both reference and developed watersheds appeared to consume more DO
                than they produced. Point-in-time measurements of tidal creek DO does not adequately
                represent the exposure of living resources to stressful low DO events.

                        Sediment characteristics were also identified as an important environmental factor
                influencing the distribution of the living resources in shallow tidal creeks. Sediments in
                developed creeks were generally composed of more sand and had larger site-to-site variation in
                physical characteristics than reference creeks. The greater sand content and more variable
                sediment characteristics in creeks located on developed watersheds were probably associated
                with alterations in erosion and deposition processes associated with watershed development.

                                                                50









                    Tidal creek sediments are repositories for pollutants. Trace metal and organic
            contaminant concentrations in sediments of the upper reaches of developed creeks, particularly
            those with industrialized watersheds or long histories of high density urban and suburban
            development, were enriched with chemicals to levels known to adversely affect living
            resources. Enrichment levels ranged from 2-10,000 times the values observed in reference
            creeks or deeper areas of South Carolina estuaries. Contaminants of particular concern were
            copper, lead, chromium, polycyclic aromatic hydrocarbons (PAHs), polychlorinated biphenyls
            (PCBs), and older pesticides, including DDT and chlordane. Low density suburban
            development did not result in sediment contaminant levels that were of concern. The single
            agriculture watershed sampled did not provide an adequate representation of sediment
            contaminants in agricultural watersheds because pollution inputs, mainly pesticides, are
            episodic and do not persist in sediments.

                    The distribution of contaminants in tidal creeks varied with the type of development and
            kind of contaminant. For example, PAHs, which are mainly derived from street runoff and
            specific point sources, generally had the highest concentrations in sedimentary environments of
            upper reaches of creeks. Pesticides concentration in at least one suburban watershed was
            frequently highest in the salt marsh adjacent to houses.

                    Sediment bioassays indicated that the cumulative amounts of chemicals in sediments of
            the upper reaches of developed creeks, particularly industrialized creeks, were adversely
            affecting key biological processes. Sediment bioassays from reference creeks did not suggest
            exposure to these sediments resulted in acute or chronic impacts on living resources.

                    The kind of benthic prey available to fish, crabs, and shrimp using tidal creeks as
            nurseries varied with salinity and sediment characteristics. Human activities associated with
            watershed development did not adversely affect the biodiversity of benthic organisms in
            creeks. Long-term salinity distributions and estuary-wide water quality were more important
            in controlling biodiversity of benthos in tidal creeks than were the local processes occurring
            within creeks.


                    The abundance of benthic organisms in tidal creek habitats was mainly controlled by
            salinity, sediment characteristics, and location within tidal creeks. These three factors
            accounted for between 7 and 84% of the variance in the abundance of benthic populations.
            After accounting for the effects of salinity, sediment characteristics, and location within a
            creek on benthic distributions, both increases and decreases in the abundance of benthic
            populations were found in developed watersheds. The greatest differences occurred in the
            upper regions of developed creeks where benthic population abundances were generally
            reduced, particularly at sites with a long history of industrial or urban development.

                    Results of a benthic recruitment experiment demonstrated that benthic resources
            maintained high population levels in creeks by continually recruiting to bottom sediments over
            the summer. This continual recruitment over the summer provided a renewable source of food

                                                            51








                for fish, shrimp, and crabs using tidal creeks as nurseries. Salinity, sediment characteristics,
                location within creeks (upper or lower reaches), and predation by fish and shrimp all had large
                influences on benthic recruitment success and colonization processes. After accounting for the
                variation in recruitment due to these natural factors, human alterations of tidal creek
                watersheds were found to adversely affect the recruitment processes for the numerically
                dominant benthic organism reproducing during the summer. Recruitment of these organisms
                was greatly reduced in developed creeks.

                       Mummichogs and grass shrimp, the preferred prey of many species of recreationally
                important fish including juvenile red drum, spotted seatrout and flounder, were the dominant
                fish and crustaceans collected in seine samples from tidal creeks during the summer. Penaeid
                shrimp and spot were the dominant recreationally important living resources that were found in
                tidal creeks. Much of the variation in the abundance of fish and crustaceans that occurred from
                creek-to-creek was associated with variation in sediment characteristics and salinity
                distributions. After accounting for creek-to-creek variation due to salinity and sediment
                distributions, no differences in abundance of the numerically dominant species of fish and
                crustaceans and the kinds/diversity of the fish and crustaceans were found between developed
                and reference creeks. The abundance of selected key species were, however, reduced in
                specific creeks with long histories of industrial and urban development.

                       Although no differences in abundance of numerically abundant fish were observed
                between creeks located in developed watersheds and reference creeks, the numerically
                dominant resident fish (i.e., mummichogs) collected from creeks with developed watersheds
                generally were characterized by poorer physiological condition (i.e., skinnier) and had blood
                that was not as vigorous as fish from reference creeks. The differences in the blood vigor
                between developed and reference creeks was most pronounced in male fish and suggests that
                immune system of resident fish is compromised in developed watersheds.

                       Fish and crustaceans in size ranges sought by fishermen were rarely collected from tidal
                creeks. These biota are apparently not be able to tolerate the low DO and other environmental
                conditions that occur in tidal creeks during summer.


                Conclusions and Recommendations
                       The cumulative impact of development has adversely affected the health of individual
                resident fish and altered distributions of the type of prey available to fish, shrimp, and crabs
                that use shallow tidal creeks as nurseries. These alterations, however, do not appear to be
                substantial enough to adversely affect the populations of recreationally and commercially
                important living resources that use creeks as nurseries. The number of creeks that are affected
                in South Carolina is small and the regions of creeks that are the most severely affected is
                confined to the headwaters which is not the preferred nursery habitat for living resources.
                Living resources from adjacent habitats continually repopulate impacted regions of creeks.

                       We believe the alterations to tidal creeks identified above are "early warnings" of more

                                                               52








            widespread degradation that will occur if the pollution inputs are not reduced. It is interesting
            that these are the same symptoms that were identified for Chesapeake Bay and other
            Northeastern estuaries in the early to mid 1970s before it became obvious that the living
            resource populations of the Bay were declining.

                   The data base that has been created for primary tidal creeks provide critical baseline
            information for a broad range of tidal creeks located in developed and undeveloped
            watersheds. This data base is a research platform for designing and conducting a broad array
            of future environmental research. Scientists from other institutions and geographical areas are
            being encouraged to use these data as part of their assessment and research programs.

                   Because tidal creek ecosystems are consumers of DO, they require adequate amounts of
            DO to sustain their functions. Water quality management agencies should ensure that DO
            allocation schemes provide sufficient DO to tidal creeks.

                   Factors that contribute to low DO in tidal creeks have not been identified or evaluated.
            Currently, we do not know if the observed alterations to DO dynamics in developed tidal
            creeks is associated with increased loadings of oxygen consuming pollutants, increased
            loadings of nutrients (nitrogen and phosphorous) that stimulate excessive growth of primary
            producers, modifications to the hydrodynamics of tidal creeks from development of the
            watershed, and/or some other unidentified cause. Until the low DO in tidal creeks can be
            linked to contributing factors, it is unlikely that policies which prevent DO problems can be
            developed. A DO budget for tidal creeks and the associated salt marshes to define their
            relative importance as consumers and identify the major factors controlling low DO conditions
            needs to be developed. Development of a DO budget is a critical step in the development of
            DO standards that will ensure that nursery ftinctions provided by tidal creeks are sustained as
            South Carolina's coastal watersheds are developed.

                   Additional research on the chronic, sublethal effects of chemical contamination to the
            health of individual organisms in tidal creeks needs to be conducted. Priority research topics
            include evaluation of the effects of contamination on immune systems, genetic adaptations of
            resident living resources to chronic exposure of high levels of chemical contaminants,
            bioaccumulation/trophic transfer of contaminants as a means of export, and in situ effects of
            contaminant exposure on survivorship, growth, and production of valued living resources
            (e.g., juvenile red drum).

                   Based on the data collected to date, status and trends monitoring efforts for tidal creeks
            should focus on the upper reaches of primary tidal creeks and should include measures of the
            health of resident organisms, water and sediment quality, and selected population and
            community parameters of resident living resources. The objective of tidal creek monitoring
            programs should be to assess the proportion of creeks that have degraded characteristics.




                                                          53









               4.4          A PROPOSED SPATIAL FRAMEWORK FOR ESSENTIAL FISH
                                    HABITAT DATA COLLECTION AND ANALYSIS


                                              Mark Monaco and Paul Orlando


                         NOAA, Strategic Environmental Assessment Division, Silver Spring, MD

                      A spatial framework was proposed for organizing and analyzing data to describe and
               identify Essential Fish habitat (EFH) for the Nation's riverine, estuarine, and coastal waters.
               A prerequisite for implementing habitat management approaches is a comprehensive set of
               spatial units for mapping the areal extent of fish species, habitats, and watershed stressors in
               rivers, estuaries, and offshore areas.


                      To enable collection and organization of EFH data within Fishery Management Plans
               (FMPs) at national, regional, and local spatial scales, the proposed framework must extend
               from rivers to the continental shelf (Figure 4.4. 1). Thus, the proposed spatial framework
               includes four geographic areas: rivers, estuaries, estuarine watersheds, and offshore zones.
               For estuaries and watersheds, spatial boundaries have already been defined by the National
               Ocean Service (NOS) through its Coastal Assessment and Data Synthesis Framework
               (CA&DS). The CA&DS includes more than 130 estuaries and watersheds along the Atlantic,
               Gulf of Mexico, and Pacific coasts (Figure 4.4.2). The EPA River Reach file of stream
               segments as well as offshore segments (e.g., biogeographic zones, NMFS depth strata, 10
               minute grids) can be added to the CA&DS to complete the required spatial coverages.

                      The existing framework reflects the evolution and maturity of a national program for
               estuaries initiated in 1985. The framework has been widely distributed and is currently used by
               numerous agencies including EPA, USGS, MMS, and several state agencies. Information
               available through Federal and state agencies can be readily "tagged" to CA&DS spatial units
               and integrated with NOAA data sets already in the CA&DS. Numerous data sets describing
               estuarine resources, habitat, and watershed uses have already been developed using the
               proposed spatial framework. Nearly all of these are national data sets. Because the thematic
               data are aggregated by common spatial units, they can be used to make comparisons, rankings,
               statistical correlations, and other analyses related to resource use and environmental quality.

                      The CA&DS is available as a digital (ArcInfo EOO) product. Since 1985, NOS has
               used the CA&DS to assemble national data sets on estuarine resources, habitat, water quality,
               and watershed activities. Data regarding species distributions and their associated habitats
               could be organized by any of the existing spatial units or any that will be added to the
               CA&DS. An example product that integrates the CA&DS with biological information is NOS's
               Estuarine Living Marine Resources program (ELMR) program (e.g., Jury 1994, Monaco and
               Christensen 1997). More detailed descriptions of these spatial units and available data sets are
               given below.


                                                            54








            Existing Spatial Units for Estuaries and Watersheds
                   Two fundamental "building blocks" in the CA&DS are estuarine salinity zones and
            USGS Hydrologic Cataloging Units (HUCs). At present, these are the smallest geographic
            units in the CA&DS and are readily aggregated into larger units that define estuaries and
            watersheds, respectively.

                   Estuarine Salini1y Zones
                   Each estuary is subdivided into three zones between the head of tide and its ocean
            boundary based on average annual and depth-averaged salinity conditions. These zones
            correspond to the following salinity regimes: Tidal Fresh (0.0 to 0.5 ppt), Mixing (0.5 to 25
            ppt), and Seawater (> 25 ppt). Two major NOAA data sets use the salinity framework to
            aggregate information. These include the ELMR data set for species distribution and
            abundance and the National Estuarine Eutrophication Survey of dissolved oxygen, nutrient
            concentrations, algal blooms, and ecological shifts.

                   While the existing 3-zone salinity structure provides a consistent and logical approach
            for synthesizing biotic information in estuaries, more refined spatial and temporal salinity units
            may be useful for some EFH applications. To that end, refined salinity distributions have
            already been completed for approximately 50 estuaries in the South Atlantic and Gulf of
            Mexico regions. For these systems, seasonal salinity contours have been constructed at 5 ppt
            increments from the head of tide to the ocean boundary for both surface and bottom layers of
            the water column. Seasonality was defined by the 3-month high salinity period, the 3-month
            low salinity period, and the two transitional periods. These refinements are required for
            approximately 80 North Atlantic, Mid-Atlantic, and West Coast estuaries.


                   Watersheds
                   Physical boundaries for estuarine watersheds were based on the USGS HUC system.
            Typically, catalog units occupy about 700 square miles and represent all or part of a surface
            drainage or a distinct hydrologic feature. For each estuary, the watershed includes all catalog
            units that drain to the estuary. In large watersheds, a distinction is made between the portion
            of the drainage area that is immediately adjacent to tidally-influenced waters (termed the
            "Estuarine Drainage Area", EDA) and the more distant regions adjacent to tidal-fresh streams
            (termed the "Fluvial Drainage Area", FDA). All 130+ estuaries in NOAA's National
            Estuarine Inventory have EDAs. Nearly one-half of the 130+ estuaries have FDAs. Several
            major NOAA data sets use the EDA/FDA framework to organize information. Among these
            are pollutant sources and loadings, population trends, land use, wetland distributions, and
            physical/hydrologic data.

            Spatial Units Now Being Added to the CA&DS
                   Coastal and Offshore Spatial Units
                   This component of the proposed EFH spatial framework is relatively undeveloped, but
            can readily accommodate any proposed organizational units. We propose that the coastal and
            offshore EFH data be organized by depth strata and or grid cells as most of state and Federal

                                                         55








             monitoring programs organize or sample fisheries data by depth zones or grids. For example,
             the NMFS northeast coast bottom trawl surveys use approximately 57 depth strata for fishery
             independent monitoring, while the joint NOAA and Canadian Dept. of Fisheries and Ocean,
             East Coast of North America Strategic Assessment Project organizes environmental data by 10
             X 10 minute grid cells. It is likely that a combination of approaches will be required to define
             spatial structures in coastal and marine areas due to the diversity of habitats, oceanographic
             currents, sampling programs, and data availability across the Nation. In addition, EFH in
             marine areas could be aggregated/organized by oceanographic features, such as large marine
             ecosystems (LMEs) (Sherman et al. 1990).


                  EPA River Reach File
                  To help accommodate riparian issues, anadromous fish habitats, and other freshwater-
             related concerns, EPAs River Reach file is being added to the CA&DS. This system, which
             divides rivers into reach segments, includes nearly all but the smallest streams within a
             watershed. The system is hierarchical and encodes river reaches as primary, secondary,
             tertiary, or quaternary depending on how far removed the stream is from the major tributary.

                  Comments and suggestions made at the IBI workshop will be incorporated into a
             complete draft to be reviewed by agencies and institutions involved in the EFH initiative (e.g.,
             NMFS, American Fisheries Society, Atlantic States Marine Fisheries Commission). We
             suggest the way forward on this work is to incorporate comments from the community on the
             feasibility and usefulness of developing a consistent spatial framework to collect and organize
             data and information to support the EFH initiative. Ultimately, consensus should be obtained
             on the spatial structures necessary to meet the EFH objectives of: 1) describing, identifying,
             and mapping EFH; 2) inventorying habitat impacts; and 3) developing corrective actions to
             conserve and enhance habitats



                                              References


             Jury, S.H., J.D. Field, D.M. Nelson, and M.E. Monaco. 1994. Distribution and abundance
                  of fishes and invertebrates in North Atlantic estuaries. ELMR report no. 13. Silver
                  Spring, MD: National Oceanic and Atmospheric Administration, Strategic
                  Environmental Assessments Division. 221 pp.

             Monaco, M.E. and J.D. Christensen. 1997. The NOAA/NOS Biogeography Program:
                  Coupling species distributions and habitat. In: Boelert, G.W. and J.D. Schumacher
                  (eds.), Changing oceans and changing fisheries: Environmental data for fisheries
                  research and management. NOAA Technical Memorandum, NOAA-TM-NMFS-
                  SWFSC. pp. 133-138.

             Sherman, K., Alexander, and B.D. Gold. 1990 Large marine ecosystems. AAAS Press. Pub.
                  No. 90-30s. Washington, DC. 242 p.

                                                  56



















                           EMOver
                           Reach                                                     EFHSpatial
                           spatial                                                   l-fierxThr,

                                                                                     Mver nwilles
                                                                                     to aver
                                            1@@s Coastal and                         Oman ShdKvs
                                            FAuarine Drainage
                                            Area Spatial
                                            Rammork
                                            LSWs 7.5Q@



                                                              NQWs Estuadne
                                                              Safinity zone
                                                              Spatial
                                                              Flummork



                                                                              Coastal and
                                                                              Nkline spatial
                                                                              R-ammorlis
                                                                              (e.g,, Water Depth
                                                                              Zones, Grids)










              Figure 4.4.1          Proposed spatial framework to support EFH data and analysis.












                                                                    57
























                                                                            r7 77









                Figure 4.4.2    Estuarine drainage watersheds included in the Coastal Assessment and Data Synthesis Fram(

                                                                               58










            4.5 AN ESTUARINE INDEX OF BIOTIC INTEGRITY FOR CHESAPEAKE BAY
                                           TIDAL FISH COMMUNITIES


                                       Margaret McGinty' and Cecelia Linder'

                   'Maryland Department of Natural Resources, 580 Taylor Avenue, Annapolis, MD
                                        'University of Delaware, Lewes, DE

                   A fish Index of Biotic Integrity (IBI) was developed for tidal fish communities of
            several small tributaries to the Chesapeake Bay (Jordan et al., 1990, Vaas and Jordan, 1991,
            Carmichael et al., 1992a, b). It is based on the original IBI (Karr, 1981) and has a nine metric
            index that contains measures of species richness (number of species, number of species
            comprising 90% of the catch and the number of species in the bottom trawl), trophic structure
            (proportion of carnivores, planktivores, and benthivores) and abundance (number of estuarine
            fish, number of anadromous fish, and total fish with Atlantic menhaden removed). The IBI was
            tested using stepwise discriminant analysis to determine the weight of each metric relevant to
            the IBI score. This exercise showed that six of the nine metrics accounted for -95% of
            variability. Of these six, the anadromous fish metric showed to be the most influential metric
            on the IBI score (partial r=.59). However, the anadromous fish metric also strongly
            correlated with spring flow (r2=.95, p=.0001). Because IBI's are intended to identify
            biological impairment due to anthropogenic influences, it was undesirable that the most
            influential metric on the IBI was so strongly influenced by natural variation. With this
            realization, a reevaluation of the IBI was initiated to attempt to define metrics that were
            minimally influenced by natural variation. Following is a description of the procedures applied
            to redefining the estuarine fish IBI.

                   Data used for the reevaluation were from 12 tributaries sampled between 1989 and
            1995. The data were divided so that two data sets were available for the effort, a development
            set, and a test set. Stations included in the IBI development set were those for which consistent
            monitoring has been done. Reference sites were established a priori based on reference
            criteria. We attempted to model the criteria established in the Virginian Province EMAP effort
            (Weisberg et al. 1992). Criteria for bottom dissolved oxygen concentrations, sediment toxicity,
            algal blooms, and land use features were proposed to identify reference and degraded sites
            within the test data sets. The criteria selected did not clearly discriminate between reference
            and degraded conditions. Cluster analysis was applied to the data to group data into two
            clusters, reference sites and degraded sites. The results of the clustering grouped the small,
            predominately urban tributaries, including a site near Sparrows Point into one group, and the
            other sites which included larger scale agriculture dominated into a second group.

                   The data from these sites were used to calculate and test approximately seventy possible
            metrics. Approximately twenty possible metrics were selected from box and whisker plots. A
            metric was selected if the mean of the reference group was different from the means of the
            degraded, and if the upper quartile of the degraded group did not extend past the mean of the

                                                         59








                reference group. We had limited success in designating meaningful metrics from this
                procedure. Possible reasons for this are that watershed scale is influencing the results, and that
                the reference criteria used to establish the reference and degraded sites do not significantly
                influence mobile fish communities.


                       Presently, data are being evaluated as originally done, where reference and degraded
                conditions are assigned based on dominant land use within the watershed. We are accounting
                for the influences of seasonal flow patterns and watershed scale. Thus far, eight metrics have
                been shown to be statistically meaningful in discerning the differences in land use. They are
                presently being examined for ecological significance. We are also exploring methods to
                account for flow influence to retain some or all of the original metrics.



                                                          References


                Carmichael, J., B. Richardson, M. Roberts, and S. Jordan. 1992a. Fish Sampling in Eight
                       Chesapeake Bay Tributaries. Maryland Department of Natural Resources, Tidewater
                       Administration, Chesapeake Bay Research and Monitoring Division CBRM-HI-92-1.
                       Annapolis, MD.

                Carmichael, J., B. Richardson, S. Jordan. 1992b. Development and Testing of Measures of
                       Ecological Integrity and Habitat Quality for Chesapeake Bay Tidal Tributaries. Final
                       Report to Maryland Coastal Zone Management. Maryland Department of Natural
                       Resources, Tidewater Administration, Chesapeake Bay Research and Monitoring
                       Division. Annapolis, MD.

                Jordan, S.J., P.A. Vaas, and J. Uphoff. 1990. Fish Assemblages as Indicators of
                       Environmental Quality in Chesapeake Bay. IN Biological Criteria: Research and
                       Regulation, 1990.

                Karr, J.R. 1981. Assessment of biotic integrity using fish communities. Fisheries. 6(6):21-27.

                Vaas, P.A. and S.J. Jordan. 1991. Long Term Trends in Abundance Indices for 19 Species of
                       Chesapeake Bay Fishes: Reflections in Trends in the Bay Ecosystem. In: J.A. Mihursky
                       and A. Chaney (eds.). New Perspectives in the Chesapeake System: A Research and
                       Management Partnership. Proceedings of a Conference. Chesapeake Research
                       Consortium Publication No. 137. Solomons, Maryland, p. 539-546.

                Weisberg, S. B., A. F. Holland, K. J. Scott, H. T. Wilson, D. G. Heimbuch, S. C.
                       Schimmel, J. B. Frithsen, J. F. Paul, J. K. Summers, R. M. Valente, J. Gerritsen, R.
                       W. Latimer. 1992. Virginian Province Demonstration Report, EMAP - Estuaries -
                       1990. EPA/620/r-93/006. U.S. Environmental Protection Agency, Washington, D.C.
                       20460.


                                                              60










            4.6                      ESTUARINE BIOTIC INTEGRITY INDEX

                                      Melissa J. Weaver' and Linda A. Deegan@

                     'Ecology & Evolutionary Biology Dept., Univ. of Tennessee, Knoxville, TN
                      'The Ecosystems Center, Marine Biological Laboratory, Woods Hole, MA

                   An Estuarine Biotic Integrity Index (EBI) has been developed and validated for
            Southern New England (Deegan et al. 1993, Deegan et al. in press). The EBI is an assessment
            of the condition of estuarine ecosystems based on the abundance, diversity, and composition of
            the fish community. Fish integrate and reflect the condition of and linkages between
            ecosystems and serve as indicators of the biotic integrity of a whole region. The EBI focuses
            on submerged rooted aquatic vascular plant (SRV) habitats because they are critical habitats for
            fish and are sensitive indicators of anthropogenic stress. Both the EBI and its' metrics are
            well-correlated with habitat and water quality, but the EBI does better than its' individual
            metrics in predicting ecosystem health, as indicated by fish production (Deegan et al. in press).
            The EBI as developed in Southern New England consists of eight metrics which include both
            functional grouping and specific species as indicators of estuarine health: total number of
            species; number of estuarine spawners, estuarine residents, and nursery species; number of
            species which make up 90% of individuals; % benthic dependent (feeding, spawning,
            dwelling, etc.) based on the number or biomass of individuals; and, % with disease.
            Individual metrics and the overall index show a strong correlation with habitat degradation.
            Habitats that were classified as impacted on the basis of year-round measurements of chemical
            and physical characteristics (algal blooms, macroalgae, low dissolved oxygen, high nutrients,
            dredged channels) had highly modified fish communities. These changes in the biotic
            community were reflected in low EBI values. Differences between moderate and low quality
            habitats are most pronounced near the end of the summer reflecting the cumulative effects of
            habitat degradation. The EBI and its metrics are well-correlated with habitat quality in
            moderate quality embayments such as Waquoit Bay. Thus, the EBI can be used to evaluate the
            current status of Southern New England estuaries.

                   For the EBI to be useful, it must not only reflect the current status of fish communities,
            but it must track changes in habitat quality over time, be applicable over a wide range of
            estuaries and habitat quality within the same geographical region, and be transferable to other
            regions. To test whether the EBI reflects long-term changes in habitat we compared habitat
            quality and fish communities at sites in Waquoit and Buttermilk Bays from the late 1980's to
            the mid-1990's. The EBI provides corroborative evidence that habitat quality within Waquoit
            Bay has continued to degrade and that efforts to control nitrogen inputs into Buttermilk Bay
            have prevented further degradation of habitats and maintained stable fish populations (Chun et
            al. 1996). Evaluation of the applicability of the EBI across a wide range of habitat quality
            within a geographic range requires that the mechanisms that relate the fish community structure
            with stress, degradation, and loss of functions are similar throughout the quality range. In
            1996 we sampled estuaries (23 sites from 13 embayments) in Buzzards Bay, Southern New

                                                          61








                England for which there is extensive background information about nitrogen loading and other
                stressors. These sites include nearly pristine sites and severely degraded sites and extend the
                range of habitat degradation compared to the original study. These data will allow us to test
                the response of the fish community to more extreme conditions and in estuaries that differ not
                only in levels of anthropogenic stress but also in flushing rate, exposure to wave action,
                morphology, sediment, macroalgae and eelgrass abundance, amount of marsh edge, and fish
                species. Moreover, we will be able to test the general applicability of the EBI throughout the
                Southern New England region. To test the transferability of the EBI to the Mid-Atlantic
                Region, we sampled habitat quality and fish communities in the lower Western Shore of
                Chesapeake Bay (26 sites within 5 subestuaries) in summer 1995. Several of the metrics and
                the EBI itself are correlated with habitat quality and are successful at separating low quality
                sites from all other sites, but they were unable to discern differences between the fish
                communities in medium and high quality (nearly pristine) sites in the Chesapeake Bay. The
                EBI as originally developed was not directly transferable to the Mid-Atlantic Region but
                required modification in the selection of the EBI metrics and their classification levels.
                Further development with regards to the scale of sampling, aggregation of data, and analyses
                are required to standardize the EBI for use among regions.

                       Several of the metrics that comprise the EBI varied with anthropogenic stress in
                Chesapeake Bay in the same manner as for Southern New England estuaries. Fish abundance,
                biomass, and number of species declined with increased stress in both regions. Of the original
                set of eight metrics, the number of species, nursery species, resident species, and spawners,
                and the proportion benthic by number of individuals were correlated with habitat quality and
                were higher for medium than for poor quality sites in Chesapeake Bay. However, the total
                number of species and the number of species for these life history strategies were the same in
                moderately degraded sub-estuaries as in pristine sub-estuaries.

                       Because the Mid-Atlantic Region has a wider range of anthropogenic stress and an
                intrinsically more diverse and abundant fish community compared to Southern New England,
                we anticipated that other metrics may be more useful than the original metrics in discriminating
                between sites of differing habitat quality in the Mid-Atlantic. For example, we found very few
                specialized feeders in the Southern New England region so trophic metrics did not differ with
                habitat quality (most species were benthic invertivores). Chesapeake Bay fishes exhibit a
                broader array of food web position and feeding strategies and we would expect specialized
                feeders to decline with increased stress. In fact, the number of invertivores did better than most
                other potential metrics in discerning habitat quality and was included in the Chesapeake Bay
                index.


                       Further modifications were made to the EBI. In Chesapeake Bay as well as in New
                England, the count of individuals and total catch biomass for low quality sites were less than
                for medium quality sites. However, in both regions, the species dominance (number of species
                that comprise 90% of the catch) did not distinguish sites by quality, and there were few
                abnormalities among the individual fishes (less than 0. 1 %), and so these two metrics were not

                                                               62









             included in the calculation of the modified EBI for the Chesapeake Bay. In general, metrics
             based on the number of species in functional groups did better than those based on counts of
             individuals or biomass. Because the number of benthic species differed between medium- and
             high-quality sites, as well as between low- and medium-quality sites, it was added to the EBI.
             We elected to retain the proportion of benthic (count or number) because it was a good
             separator of low- and medium-quality sites.

                     Surprisingly, the absolute number of species in any habitat quality was not higher in the
             Mid-Atlantic Region compared to Southern New England, rather it was lower for several of
             the metrics, including the number of resident species. Although there were many more species
             in the total sampling catch in the Chesapeake Bay than in the Southern New England catch
             (about 59 versus about 35), the number of species that nurse, spawn, or reside permanently
             within the estuary was lower in trawl catches in Chesapeake than in New England. However,
             the number of species within a sampling site within the Chesapeake was sometimes higher than
             that in Southern New England. There was a much higher diversity among trawls within a site
             and within an embayment in Chesapeake Bay than in the Southern New England bays, but the
             EBI did not reflect this diversity. The EBI as originally developed apparently reflects the fish
             community diversity at a very local level perhaps due to the scale of sampling, and the manner
             in which the data were aggregated and analyzed. (The metrics were scored and the EBI
             calculated for each trawl and then averaged across trawls at each site.) Furthermore, the
             overall patterns in the metrics in relation to quality differences did not differ between riverine
             estuaries (York, James, and Rappahannock Rivers) and subembayments (Lynnhaven and
             Mobjack) within Chesapeake Bay. By calculating an index that integrates the habitat quality
             and quantity over each embayment and then throughout the subregion (lower Western Shore of
             Chesapeake Bay), we may have a more effective measure of habitat quality within Chesapeake
             Bay.



                                                         References


             Chun, N. K., M. J. Weaver, and L. A. Deegan 1996. Assessment of fish communities in New
                     England embayments: application of the Estuarine Biotic Integrity Index. Biological
                     Bulletin 191 (October).

             Deegan, L. A., J. T. Finn, S. G. Ayvazian,, C. A. Ryder 1993. Feasibility and application
                     of the index of biotic integrity to Massachusetts estuaries (EBI). Final Project Report.
                     Massachusetts Executive Office of Environmental Affairs, Department of
                     Environmental Protection, North Grafton, Massachusetts.

             Deegan, L. A., J. T. Finn, S. G. Ayvazian,, C. A. Ryder-Kieffer, and J. Buonaccorsi.
                      Development and validation of an estuarine biotic integrity index. Estuaries: in press.




                                                              63










                4.7       OHIO'S LAKE ERIE AND LACUSTUARY MONITORING PROGRAM


                                             Roger F. Thoma and Chris 0. Yoder

                                                  Ohio EPA, Columbus, OH

                       In 1993 the Ohio EPA began a project designed to develop numerical biological criteria
                for shoreline waters of Lake Erie and areas of tributary streams affected by lake levels,
                referred to as lacustuaries. The term lacustuary is a combination of lacustrine and estuary.
                Lacustuary is defined as a transition zone in a river that flows into a freshwater lake and is the
                portion of river affected by the water level of the lake. Lacustuaries begin where lotic
                conditions end in the river and end where the lake proper begins. They have hydrologic
                conditions similar to estuaries in that they are affected by tides (primarily wind driven,
                occasionally barometric) and are lentic habitats. Lacustuaries differ from estuaries since their
                chemical properties are less saline with salinity gradients going from higher upstream to lower
                at the lake interface (Brant and Herdendorf 1972). It is felt that the term lacustuary is needed
                to avoid confusion of terms and concepts that ensue when estuary is used for freshwater
                systems. Though there are some similarities, estuaries and lacustuaries differ in numerous
                important functions and should not be confused with each other.

                       This IBI project was conducted in the following steps: 1) sampling of the general
                habitat types found in the Lake Erie ecosystem using various sampling methodologies; 2)
                evaluation of sampler type efficiency and selection of the method to be used in each habitat
                type; 3) continued sampling using the selected methodology; 4) evaluation of potential metrics;
                5) selection and calibration of IBI metrics; 6) continued sampling; 7) calculation of Lake Erie
                shoreline and lacustuary IBI scores; 8) evaluation of environmental conditions in Lake Erie
                and associated lacustuary areas. This study was built on data collected since 1982.

                       Ninety sites (324 individual collections) were sampled in Lake Erie from 1993 through
                1996. Site selection reflected the habitat types found in the lake's nearshore areas and
                provided a thorough coverage (approximately one site for every 5 miles or 8 kilometers) of the
                area investigated. Sites were located along harbor breakwaters, sand/gravel beaches, the
                shores of the Lake Erie Islands, bedrock cliffs and modified shore lines with numerous types of
                structures designed primarily to prevent shoreline erosion. Wetland/bay-like habitats were
                sampled in Sandusky Bay, East Harbor State Park, and Presque Isle PA (11 sites).
                Lacustuaries were sampled at 125 sites (593 individual collections) from 1982 through 1996.
                Sites were located at the mouth, head, and midsection of each lacustuary.

                       All fish were collected using a 5.8-meter modified V-hull john boat. Lacustuary
                habitats were sampled during daylight and lake sites were sampled at night. Fish were
                identified to species, enumerated, examined for external anomalies, and either returned to the
                lake or preserved as voucher specimens. Weights were taken on a representative sub-sample if
                more than 15 individuals of a species were captured. All fish were weighed if 15 or fewer

                                                              64









          individuals of a species were captured. Each sampling site was 500 meters long and within 1
          meter of the shore.


          Metrics
                 A large number of metrics were examined to determine the metrics best suited for use
          in a Lake Erie IBI and lacustuary IBI. Examination of metrics for lake and lacustuary sites
          indicated the relative abundances and percent composition of fish in the two types of habitat
          should be evaluated separately. When metrics were selected, an effort was made to use
          groupings that maximized the range of values possible. Metrics with low breadth can result in
          a yes-no, present-absent evaluation instead of the intended strongly - moderately - little
          deviation assessment. Comments on selected metrics are listed below (LE- Lake Erie, L-
          lacustuary). A complete list is presented in Table 4.7. 1.

                       Number of benthic species (LE, Q
                 Ls metric is thought to primarily respond to environmental disturbance from excess
                 sedimentation and secondarily to toxicity and low oxygen levels. It comprises darters,
                 sculpins, and madtoms. Other benthic species of generally greater environmental
                 tolerance, such as bullheads and suckers, were excluded in order to maintain
                 sensitivity.

                       Number of sunfish species (LE, Q
                 Ls metric includes sunfish and species of the genera Pomoxis and Micropterus.

                       Number of cyprinid species (L)
                 @yprinid species were historically a prominent community component that could be
                 found in all lacustuary habitats, and several highly sensitive species (now apparently
                 extirpated in Ohio) were primarily associated with Lake Erie near shore areas. This
                 metric can accommodate future changes in the ecosystem if environmental conditions
                 improve to the point that locally extirpated species become reestablished.

                       Number ofphytophilic species (LE)
                 @ariations in this metric are associated with increases in submerged aquatic vascular
                 plants (especially Potamogeton and Vallisneria) which are found in high quality, clear,
                 low-polluted waters which is an ecological parameter of substantial historical
                 prominence.

                 0     Percent lake individuals (LE)
                 This metric reflects a species guild that has proven to be sensitive to environmental
                 disturbances in Lake Erie. Because sufficient numbers of lake-associated species still
                 exist and much room for improvement is possible, this metric is ideal for measuring the
                 long term trends of Lake Erie fish communities.




                                                   65









                                Percent phytophilic individuals (L)
                        As with percent lake species, this metric is highly sensitive to slight environmental
                        change. Historically, lacustuaries exhibited high numbers of phytophilic species and
                        very high numbers of individuals. Though numerous phytophilic species have
                        disappeared from Lake Erie's lacustuaries, many species still subsist at very low
                        numbers in almost all areas. As even the most polluted sites generally have the same
                        phytophilic species, we decided to use the number of individuals as sites of higher
                        environmental quality exhibited much higher abundances than degraded sites. This
                        allows discrimination between the very bad sites and fair sites. If lacustuary habitats
                        should improve in the future, this metric may to be converted to a number of species
                        metric.


                        0      Percent top carnivores (LE, L)
                        Only those species that at an adult size feed on fish or crayfish more than 80% of the
                        time are considered top carnivores. Species such as crappie and channel catfish that
                        have a more plastic feeding behavior and can convert to other forms of food resources
                        under sub-optimal conditions are excluded.

                        0      Percent non-indigenous species (LE, L)
                        Non-indigenous species have been found, in this study, to increase in areas of higher
                        disturbance, especially that associated with extensive urban development. Only species
                        which were present in the system originally (pre 1700s) are considered indigenous.

                               Percent diseased individuals (LE, L)
                        Is metric is a measure of the percent of individuals that have externally observable
                        deformities, eroded fins, lesions or tumors.

                Scoring Considerations and Attainiment Criteria
                Setting the Nine1y-fifth Percentile Line
                        Because the fish community of Lake Erie has experienced pervasive negative impacts
                (Hartman 1972, Regier and Hartman 1973, Trautman 1        '981, Van Meter and Trautman 1970,
                and White et al. 1975), the selection of reference sites and 95 percent lines is problematic. If
                one sets expectations at levels thought to be equivalent to the historic potential of Lake Erie, all
                sites would score so low that it would not be possible to distinguish highly, moderately, and
                slightly polluted areas. Alternatively, if a straightforward ninety-fifth percentile line is
                employed it becomes possible that sites will score in the exceptional range. This prospect is
                unacceptable in light of the present condition of Lake Erie. The intent of the IBI is to measure
                integrity and Lake Erie presently exhibits very little integrity. A score of exceptional would be
                construed as an indication that Lake Erie is approaching full recovery, which it is not.

                        The approach employed in this IBI effort has been to use a modification of the ninety-
                fifth percentile methodology. When drawing the ninety-fifth percentile line, the line was
                always drawn between the ninety-fifth percent value and the next value point. This

                                                                66








            acknowledges the fact that if greater integrity existed the ninety-fifth percent value would be
            more stringent while keeping scoring criteria at a level that allows discrimination of the present
            conditions. Using this methodology, none of the sites sampled in this study have scored fifty
            or higher.

                    Karr (in press) proposes the use of ecological dose-response curves to devise scoring
            criteria for IBI metrics. Such an approach may prove to be the best methodology to score Lake
            Erie's fish communities because of the extensive disturbances experienced and the lack of
            reference conditions. Future work on the Lake Erie and lacustuary IBI will examine ecological
            dose-response curves.

            InteRrijy classifications.
                    Integrity ranges of exceptional ( > 50), good ( > 42), fair ( > 3 1), poor ( > 17), and very
            poor (< = 17) have been set for Lake Erie and its lacustuaries. The predicament of setting
            specific integrity ranges for Lake Erie and its lacustuaries is difficult because all sites sampled
            have been affected to some degree by dramatic ecological changes. One approach has been to
            use the IBI value that occurs at the 25 percentile of the reference sites selected as representative
            of a habitat type as the level at which the "good" classification begins. It is incumbent in the 25
            percentile approach that the reference site data base is composed of sites that very nearly
            approach biological integrity. In the Huron-Erie Lake Plan (HELP) ecoregion, where most
            sites have been impacted and do not display ecological integrity, the Ohio EPA elected to use
            the 90th percentile of all sites sampled to derive attainment criteria. Because the Lake Erie
            system displays pervasive negative environmental effects, an approach like the HELP
            ecoregion strategy is desirable. This work differs from the previous HELP effort by using only
            the least impacted sites to set the 90 percentile instead of all sites. Use of a 25 percentile in
            Lake Erie waters would result in a criteria that accepts environmental degradation while the 90
            percentile of least impacted sets a goal that the data have demonstrated can be attained in a
            reasonable time frame with some environmental amelioration (even in light of pervasive
            impacts). Once the good attainment point was set, exceptional, fair, poor and very poor
            integrity ranges were set based on an understanding of species composition at differing IBI
            levels.


                    The potential for this scoring system to change is great, as Lake Erie is currently in a
            state of dynamic flux. New non-indigenous species are invading at increasing rates (Mills et
            al. 1993) and phosphorus levels are decreasing (Bertram 1993, Makarewicz and Bertram 1991)
            and the two are interacting in unpredictable ways to create considerable uncertainty.
            Continued monitoring will be required to track changing community conditions, and attainment
            criteria will need to be reviewed in light of future changes.

            Application examples:
            Lacustuary assessments
                    Four examples are provided that demonstrate the effectiveness of the IBI to identify
            areas with no improvement, improving conditions and gradients of impact, which can be

                                                             67








                related to site-specific anthropogenic activities. Multiple examples exist in each type of
                situation.


                        Index of biotic integrity scores from 4 years of biological monitoring in the Ottawa
                River have consistently remained in the poor to very poor range (Figure 4.7. 1). Numerous
                combined sewer overflows, urban runoff, leaking landfills and contaminated sediments
                combine to suppress communities to extreme low levels. Over the 10 year period of
                monitoring, none of the impact sources have been addressed and consequently no changes are
                detectable in fish communities. Restoration potential for this lacustuary is good because depths
                are still shallow enough to allow reestablishment of aquatic macrophyte communities, a factor
                critical to fish community integrity.

                        The Black River lacustuary was sampled in 1992 and 1994. Scores for the IBI were
                consistently poor in 1982 and poor to mostly fair in 1992 (Figure 4.7.2). Community
                improvements over the 10 year period were due to upgrades at the upstream Elyria waste water
                treatment plant that reduced loading to the Black River and its lacustuary. Removal of
                contaminated sediments after 1992 probably will lead to further fish community improvements.
                Presently the lacustuary is limited by nutrient enrichment primarily from upstream nonpoint
                pollution, both urban and rural. Very little submerged aquatic vegetation exists in the ,
                lacustuary although habitat structure is suitable. With further reductions of pollutant loads and
                a resurgence of plant life, fish communities in the Black River lacustuary should recover and
                attain exceptional conditions.

                        Seven sites have been sampled since 1989 in the 2.5 mile (4 km) length of the
                Ashtabula River lacustuary (Figure 4.7.3). Downstream of river mile (RM) 2.3, much of the
                waterway was lined with sheet piling and boat docks. A ship channel extended from the river
                mouth to RM 0.7. Fields Brook joins the Ashtabula River at RM 1.6. Sediment
                contamination has been documented downstream of Fields Brook. In 1989, fish community
                sampling was conducted to evaluate the degree of impact associated with chemical degradation
                originating from Fields Brook and habitat alteration of the lacustuary. It was concluded that
                shoreline development was the principal factor impacting fish communities in the lower
                Ashtabula River with a lesser effect from chemical pollutants. In general, IBIs were good to
                fair in upper reaches, fair to poor near Fields Brook, and fair to very poor in the ship channel
                area.


                        The Conneaut Creek lacustuary extends for 2.2 miles (3.5 km) upstream from the
                mouth. A total of 6 sites have been sampled since 1989 (Figure 4.7.4). Very little
                environmental deterioration was seen in the lotic portions of the system and extensive areas of
                the basin are wooded. The lower 0.5 mile of the stream was a ship channel with deep sheet
                piling lined banks while upstream from RM 0.5, the channel was shallower and at least
                partially vegetated along the banks; most of this reach was relatively narrow with moderate
                accumulations of silt and sediment. An area of thick silt and sediment with a large expanse of
                emergent and submergent vegetation was present at RM 1.0. Upstream from the ship channel,

                                                               68








             in the area of vegetation, IBI scores were in the good range while ship channel sites (RMs 1.3
             and 0.6) had IBI scores in the poor to fair range. The dichotomy of good and poor community
             conditions found in this lacustuary illustrate the strong effect that habitat alterations can have
             on biological conditions even in areas where no impacts from water column chemistry exist.

             Lake Erie
                    In Lake Erie, three factors affect fish community structure; lake-wide trophic changes
             as a result of nutrient enrichment, habitat loss primarily in the form of wetland destruction or
             diking and shoreline modifications, and localized environmental impacts from industrial and
             municipal discharges. Of principal significance is the predominant effect of lake-wide trophic
             changes and associated species losses. These changes have resulted in most sites scoring as
             fair with few good and no exceptional values attained (Figure 4.7.5). Four of the nine sites
             that clearly fall into the good range are from the shorelines of the Lake Erie Islands. Island
             sites score better, in part, due to their distance from lacustuaries and associated impacts.
             Habitat was also an important factor for island sites. The principal habitat type encountered
             around the islands was boulder - rubble strewn shorelines with high levels of substrate texture.
             It was observed in this study that the greater the habitat texture the greater the relative
             abundance and number of species. Breakwater sites, at the mouths of lacustuaries, had habitat
             textures similar to island sites, but failed to reach the levels attained at island sites. This was
             due to lacustuaries experiencing environmental stress from higher loads of pollutants. Beaches
             were the area of lowest substrate texture and tended to score lower than other habitat types (in
             the absence of other environmental stresses). Examples of localized pollution impacts were
             found in the Maumee Bay and Cuyahoga River at Cleveland, areas where in spite of the fact
             that habitats were highly textured breakwaters, IBI values remained in the poor range. The
             only site in this study that fell in the very poor classification was just east of the Maumee Bay
             area. This site was a riprapped beach in an area where extensive settling of organic debris and
             urban waste was occurring. The dominant species at this site was goldfish, a highly tolerant
             fish.


             General
                    None of the lake or lacustuary sites attained an integrity level of exceptional and only a
             few attained the good level. This was reflective of the widespread and pervasive nature of
             environmental impacts in the region. Many species were missing (Trautman 1981, Hartman
             1972) and trophic dynamics were radically changed (Regier and Hartman 1973, Stoermer et al.
             1987). Five of the 20 most abundant species were non-indigenous species. Ninety three
             species were recorded and the average relative abundance of individuals (number per
             kilometer) was 687.

                    At the good-fair integrity interface, similarities between Lake Erie and its lacustuaries
             begin to diverge. In the lake proper, environmental impacts are more widely dispersed and less
             intense, whereas in lacustuaries they can be very intense and are always more concentrated. In
             the lake, only 73 species were recorded and the average relative number of individuals
             (number per kilometer) was 934. Integrity levels of fair dominated the lake results (59%), poor

                                                              69








                to very poor (24%) comprised the next largest classification, and good (17%) the least. In the
                lacustuaries 87 species were recorded and the average relative number of individuals (number
                per kilometer) was 552. Poor to very poor IBI scores dominated the results (71 %) while fair
                comprised 23 % and good equaled only 6 %.


                                                          References


                Bertram, P.E. 1993. Total Phosphorous and Dissolved Oxygen Trends in the Central Basin
                       of Lake Erie, 1970-1991. J. Great Lakes Res. 19:224-236.

                Brant, R.A. and C.E. Herdendorf. 1972. Delineation of Great Lakes Estuaries. Proceedings,
                       15th Conference, International Association for Great Lakes Research 15:710-718.

                Hartman, W.L. 1972. Lake Erie: Effects of Exploitation, Environmental Changes and New
                       Species on the Fishery Resource. J. Fish. Res. Board Can. 29:899-912.

                Karr, J.R., (in press). Rivers as Sentinels: Using the Biology of Rivers to Guide Landscape
                       Management.

                Makarewicz, J.C. and P. Bertram. 1991. Evidence for the Restoration of the Lake Erie
                       Ecosystem. BioScience vol. 41, no. 4.

                Mills, E.L., J.H. Leach, J.T. Carlton, and C.L. Secor. 1993. Exotic species in the Great
                       Lakes: a history of biotic crises and anthropogenic introductions. J. Great Lakes Res.
                       19:1-54.


                Regier, H. A. and W. L. Hartman. 1973. Lake Erie's Fish Community: 150 Years of
                       Cultural Stress. Science 180: 1248-1255.


                Stoermer, E. F., J. P. Kociolek, C. L. Schelske, and D. J. Conley. 1987. Qualitative
                       Analysis of Siliceous Microfossils in the Sediments of Lake Erie's Central Basin.
                       Diatom Research 2:113-134.


                Trautman, M.B. 1981. The Fishes of Ohio. Ohio State Univ. Press. Columbus, OH, 782 pp.

                Van Meter, H.D. and M.D. Trautman. 1970. An annotated list of the fishes of Lake Erie and
                       its tributary waters exclusive of the Detroit River. Ohio J. Sci. 70(2):65-78.

                White, A.M., M.B. Trautman, E.J. Foell, M.P. Kelty, and R. Gaby. 1975. Water quality
                       baseline assessment for the Cleveland area-Lake Erie. V 2. The fishes of the Cleveland
                       metropolitan area including the Lake Erie shoreline. U.S. E PA, Chicago, 111. 181pp.



                                                               70









            Table 4.7.1 Metrics used in Ohio EPA's two IBIs developed to evaluate Lake Erie nearshore
            ecosystems and lacustuaries.

                     Lake Erie Metrics                                   Lacustuary Metrics


                                          Species number metrics
            # Species                                            # Species
            # Sunfish species                                    # Sunfish species
            # Phytophilic species                                #Cyprinid species
            # Benthic species                                    # Benthic species
                                          Behavior/trophic guild metrics
            % Lake assoc. individuals                            % Phytophilic individuals
            % Top carnivores                                     % Top carnivores
            # Intolerant species                                 # Intolerant species
            % Omnivore individuals                               % Omnivore individuals
            % Non-indigenous ind.                                % Non-indigenous ind.
            % Tolerant individuals                               % Tolerant individuals
                                           Communi1y health metrics
            % DELT*                                              % DELT*

            Relative numbers"                                    Relative numbers"


              Externally observable deformities, eroded fins, lesions, and tumors.
               Includes non-indigenous species and excludes gizzard shad.















                                                            71


















                               60

                                            Exceptional
                               50                                                                                          . .. ...

                                           Good
                               40

                                             Fair
                               30           . .............. ......................  ..................................... .................................................. . ................................................. ...... . ...

                                            Poor
                               20                                     X..                                           E 1990
                                                       :.............
                                               ..........  wmn@- . . .... .. ................. ......... 1996 ................................ ...... . ...................... .......................................
                               10      - Very Poor                            ..M   x 1992

                                 0 1                                                                                         1
                                    10 9              8        7       6        5        4       3        2        1        0
                                                                      River Mile





                   Figure 4.7.1 Ottawa River IBI scores for 1986, 1990, 1992, and 1996. Exceptional, good, fair,
                   poor, and very poor classifications are delimited by dashed lines.









                                                                               72


















                                   60

                                                   Exceptional
                                   50

                                                   Good
                                   40         . ........... . ..... ---     .. ...... .. .... .. .
                                                    Fair                                                                              West bank 1992
                       im- 30                                ....................................        ........................... ........ ...............................
                                                                                East bank
                                   20              Poor                         1992

                                                ............................................... ........... .......................... ............ ............................. ................................................. ..........
                                   10              Very Poor

                                       0     1-                                                                                                     ____j
                                            7               6               5               4               3               2                1               0
                                                                                      River Mile





                    Figure 4.7.2 Black River IBI scores for 1982 and 1992. Exceptional, good, fair, poor, and very
                    poor classifications are delimited by dashed lines.
                                                                                                                                       7:7@j










                                                                                                 73


















                                 60

                                             Exceptional
                                 50                                1995                     . . ....... ... ..

                                             Good
                                 40                                              . ............... .. ....... ...... ..................................................
                                              Fair                        1989
                                                                                                              1993
                                 30                     ....................................................... .I. . .... I. ............... . ........ ..............................

                                             Poor
                                 20

                                                ....................................... ................. ............. ............... .............................. ............. . .....
                                 10          Very Poor

                                    0

                                        3                            2                            1                           0
                                                                        River Mile





                    Figure 4.7.3 Ashtabula River IBI scores for 1989, 1993, and 1995. Exceptional, good, fair,
                    poor, and very poor classifications are delimited by dashed lines.









                                                                               74


















                              60
                                           Exceptional
                              50       .. . .. ..... . ........ .. .... .

                                          Good
                                                                1989                      . .. . . .. .... ... .... . . . ................................. . ......... ............
                              40                                  13 -,
                                                                                   1995
                                            Fair                                .......... .. .........
                              30       .. .... . ... . .......................................................... ---  ....................-........................ .... ....................... .............................. 1993

                                           Poor
                              20 .-

                                            .......... .....................................-................ .... ...... .................. ...... ....... . .......... -- ............................. .. ......................................... ...........................
                              10       - Very Poor
                                 0    1
                                     3                               2                               1                               0
                                                                         River Mile





                 Figure 4.7.4 Conneaut Creek IBI scores for 1989, 1993, and 1995. Exceptional, good, fair,
                 poor, and very poor classifications are delimited by dashed lines.
                                                                                     777@@@







                                                                                  75
























                             60

                                           Exceptional
                             50   ----------------------------------------------
                                           Good           9                         0
                                                          *-                 - - - - - - - - - - - - - - - -%- - - -
                             40     ------------

                                      Fai
                                        r


                                                                                                - - - - - - - - - -
                             30


                                     Poo
                             20

                                                    --------------            -------------


                                                      Sandusky Bay
                             10      Very Poor        and Lake Erie                                Presque Isle PA
                                                         Islands          Cleveland
                                            Maumee Ba@                      area
                                0      . I   I I -'-                        @ I --, I I I I . I     I I I I ,
                                1350         1300         1250         1200         1150         1100          1050

                                                              Lake Shore Mile





                  Figure4.7.5 IBI scores for all Lake Erie sites. Habitats include Sandusky Bay, Bass islands
                  area, and miscellaneous shore types (rocky and sandy beaches). Lake shore miles are measured
                  from east to west. Exceptional, good, fair, poor, and very poor categories are delimited by dashed
                  lines.













                                                                        76










             4.8           AN INDEX OF BENTHIC CONDITION TO DETERMINE THE
                                    MAGNITUDE OF ENVIRONMENTAL STRESS


                                           Kevin Summers and Virginia Engle

                                      US EPA, Gulf Ecology Div., Gulf Breeze, FL

                    The Environmental Monitoring and Assessment Program for Estuaries (EMAP-E) in
             the Louisianian Province has collected data from 644 stations in four years (1991-1994). One
             of the objectives of EMAP is to develop and test indicators of environmental quality and to use
             these indicators to determine the status of, and trends in environmental condition over large
             geographical areas. A core response indicator that has been developed for EMAP-E is the
             benthic index. The benthic index is a useftil and valid indicator of estuarine condition that is
             intended to provide environmental managers with a simple tool for assessing the health of
             benthic macroinvertebrate communities. It represents the response of the benthic
             macroinvertebrate community to environmental stressors.

                    The benthic index was developed by first choosing a set of test sites that represent
             extreme degraded and reference conditions based on a priori guidelines for dissolved oxygen,
             sediment toxicity, and sediment contamination. These test stations were also chosen to
             represent both the range of natural habitat conditions found in the province and the entire
             geographic area included in the province. We compiled a suite of parameters that represent
             indicators of benthic community health including species richness and diversity, overall
             abundance, and the proportional abundance of major taxonomic and trophic groups of benthos.
             Parameters that showed a high degree of correlation with natural habitat conditions (e.g.,
             salinity or sediment grain-size) were adjusted accordingly. Stepwise and canonical
             discriminant analyses were used to determine which subset of the benthic parameters best
             discriminated between the degraded and reference test sites and to assign coefficients or
             weighting factors to each of the parameters.

                    We originally developed a benthic index using data from the 1991 demonstration
             project in the Louisianian Province. That benthic index combined the Shannon-Wiener index
             (adjusted for salinity) and the percentages of total abundance represented by tubificids (Family:
             Tubificidde) and bivalves (Class: Bivalvia). This original index successfully discriminated
             between reference sites and sites that were degraded with respect to sediment contaminants,
             sediment toxicity, and hypoxia. However, when this benthic index was applied to an
             independent set of data from the Louisianian Province (EMAP's 1992 sampling of 159 new
             sites), validation of the index was unsuccessful. This was partly the result of 1992 sites that
             had benthic conditions that were substantially more degraded than the original test sites used to
             develop the index. A new, revised benthic index was developed using test sites from 1991 and
             1992 that represented a broader set of environmental conditions.

                    The revised benthic index that was developed for EMAP-E in the Louisianian Province

                                                            77








                is a linear combination of 1) the proportion of expected diversity, 2) mean abundance of
                tubificid oligochaetes, 3) the percent of total abundance represented by capitellid polychaetes,
                4) percent bivalves, and 5) percent amphipods. The weights on each of the independent
                variables were determined empirically based on the data. This benthic index successfully
                delineates benthic communities that have characteristics similar to those found in areas known
                to be degraded, from benthic communities that are similar to those found in known, reference
                areas. The difference in benthic community structure indicated by our benthic index is more
                likely to be due to anthropogenic stress than to natural habitat variability.

                       Validation of the benthic index was accomplished by using an independent set of data
                from two subsequent years, 1993 and 1994, as well as data from special study sites
                representing between-year and within-year replicates. Validation of the benthic index
                consisted of three steps: assessment of the correct classification by the index of an independent
                set of degraded and reference sites, comparison of the cumulative distribution function of the
                index among four years, and correct classification of replicate sites by the index. The revised
                benthic index was validated successfully using the independent data and was then
                retrospectively applied to all of the data collected from Gulf of Mexico estuaries during 1991-
                1994.


                       The benthic index is intended to be used as an indicator of the ecological health of
                estuaries by ranking and classifying the conditions of benthic invertebrate communities over
                large geographical areas. It can also be used successfully to classify specific areas of a single
                estuary as degraded or reference with respect to benthos. We can then try to identify what
                possible stressors may exist only in the degraded areas. This provides a clue to what
                environinental impacts may be affecting the benthic communities at the degraded areas.

                       Monitoring ecological indicators of condition on a regional scale can produce
                information that is useful to resource managers. EMAP's probabilistic sample design and
                standardized methodologies allowed for the collection of data that can be used in preforming
                assessments across the region with a quantifiable level of confidence. Benthic index estimates
                for the estuaries of the Gulf of Mexico based on the 1991-1994 monitoring indicate that
                23 ï¿½ 6 % of the estuarine area in the Louisianian Province had degraded benthic resources based
                on low benthic index scores.


                       Using the benthic index as an indicator of benthic condition, we explored the spatial
                distribution of degraded benthic communities in individual estuaries, Pensacola Bay, FL and
                Mobile Bay, AL. These estuaries were sampled as part of a regional EMAP effort to
                characterize ecological conditions on a smaller geographic scale. We also investigated
                statistical associations between various environmental indicators and the benthic index in these
                estuaries.


                       Pensacola Bay, an estuary in northwest Florida, has a history of sedimentation
                problems due to poor flushing and locally high inputs of suspended sediments which are

                                                             78









             generally retained within the system. However, the sediment and biological quality of
             Pensacola Bay have deteriorated since the 1950s and recovery is improbable without
             substantial intervention. The benthic index identified 12 degraded sites that were located
             primarily in the mainstem of Pensacola Bay and in the three bayous proximal to the city of
             Pensacola (Bayous Chico, Grande, and Texas). Pensacola Bay has severely contaminated
             sediments with as many as 40 chemicals at concentrations greater than ER-L guidelines,
             especially in the bayous. The benthic community is impoverished throughout the bay, but
             severely so in the areas with low sediment quality.

                    The benthic communities of Mobile Bay are more affected by hypoxia and nutrient
             enrichment than by toxic sediments. Although hypoxia in Mobile Bay is primarily driven by
             salinity stratification and the timing and duration of wind events, the severity and extent of
             hypoxic bottom waters may be exacerbated by nutrient enrichment. In this case the dominant
             benthic taxa at degraded sites are small, tube-dwelling polychaetes indicative of a stressed
             environment.


                    We have successfully synthesized benthic community information into a benthic index
             of ecological condition that provides environmental managers with an easy way to assess the
             status of the health of benthic communities over large geographical areas. A response
             indicator like the benthic index provides a numerical quantification of the response of the
             benthic communities to environmental stresses. Because the benthic index is scalable and the
             criteria for determining the classification of degraded or reference are numeric, the application
             of the benthic index to other estuaries is straightforward. The application of the benthic index
             to data from an independent sampling program in Pensacola Bay illustrates this point.























                                                            79










               4.9         A BENTHIC INDEX FOR ESTUARIES OF THE SOUTHEASTERN
                                                    UNITES STATES

                             Robert F. Van Dolah', Jeffrey L. Hyland', A. Frederick Holland,
                                         Jeffrey S. Rosen' and Timothy R. Snoots'
                             'SC Marine Resources DiYision, P.O. Box 12559, Charleston, SC
                           2NOAA Carolinian Province Office, P.O. Box 12559, Charleston, SC
                                   'TPMC, Mill Wharf Plaza, Suite 208, Scituate, MA


               Introduction
                      We have developed and validated a benthic index for southeastern estuaries using data
               from the joint EPA-NOAA Environmental Monitoring and Assessment Program (EMAP) in
               the Carolinian Province (Cape Henry, VA-St. Lucie Inlet, FL). Our approach follows methods
               developed by Weisberg et al. (1997) to characterize the condition of infaunal assemblages in
               Chesapeake Bay. This approach differs from the one used in previous EMAP estuarine
               surveys of the Virginian Province (Weisberg et al. 1992) and Louisianian Province (Engle et
               al. 1994), which produces an index derived from multivariate stepwise and canonical
               discriminant analysis. The approach we have adopted here is a modification of the Index of
               Biotic Integrity (IBI) developed originally for freshwater systems (Karr et al. 1986, Karr
               1991). Though there are similarities to the latter IBI approach, one major difference is the
               way in which scoring criteria for selected biological attributes were established.

                      Our goal was to develop an index that characterizes the quality of estuarine habitats
               based on the condition of resident benthic infaunal assemblages. Additionally, the index
               should be:


               1 .    suitable for use throughout the region;

               2.     applicable to all habitat types;

               3.     easy to understand and interpret; and

               4.     effective in discriminating between degraded and undegraded habitats.


               Methods
                      Results of the EMAP survey completed in 1994 indicated that several natural abiotic
               factors (salinity, latitude, silt-clay, and TOC) had strong influences on infaunal variables
               (Hyland et al. 1996). The approach used here attempted to produce an index of integrated
               benthic response variables independent of these abiotic factors. The basic steps used to
               develop the index involved:

               1 .    selecting a test data set (75 stations sampled in the summer 1994 from the NC/VA


                                                            80









                    border to the southern end of Indian River Lagoon, FL);

            2.      defining major habitat types based on classification analysis of the benthic species test
                    data and evaluation of the physical attributes associated with the resulting site groups;

            3.      comparing various candidate benthic attributes between reference sites and degraded
                    sites for each of the major habitat types;

            4.      selecting the attributes that best discriminated between reference and degraded sites for
                    inclusion in the index (key criteria considered were whether differences were in the
                    right direction and statistically significant);

            5.      establishing scoring criteria (thresholds) for the selected attributes based on the
                    distribution of attribute values at reference sites;

            6.      deriving a combined index value for each sample by assigning an individual score for
                    each attribute, based on the scoring criteria, and then averaging the individual scores;
                    and

            7.      validating the index with an independent data set (96 stations sampled during the
                    summer 1993 and 1995).

                    Several criteria were used to classify stations as degraded or undegraded on the basis of
            chemistry and toxicity data. Stations were considered to be degraded if:

            1 .     sediments were contaminated (i.e., three or more contaminants in excess of lower,
                    threshold ER-L/TEL sediment bioeffect guidelines, or one or more contaminants in
                    excess of higher ER-M/PEL probable effect guidelines);

            2.      laboratory sediment bioassays showed toxicity (@: 2 hits using amphipods, seed clams,
                    and/or Microtox); or

            3.      there was low dissolved oxygen observed in the water column (< 0.3 mg/L for any
                    observation, < 2.0 mg/L for 20% or more of observations, or < 5.0 mg/L for all
                    observations over a 24-hr time series). ER-L and ER-M values are from Long et al.
                    (1995) and Long and Morgan (1990); TEL and PEL values are from MacDonald
                    (1994).

                    Forty benthic infaunal attributes were considered and statistically compared within each
            of four habitat groups. These groups were oligohaline-mesohaline stations (:!@ 18 ppt) from all
            latitudes, polyhaline-euhaline stations (> 18 ppt) from northern latitudes (> 34.5' N),
            polyhaline-euhatine stations from middle latitudes (30-34.5' N) and polyhaline-euhatine
            stations from southern latitudes (< 30' N). The initial list of attributes included various

                                                             81








                measures of diversity, abundance, dominance, and presence of indicator species (e.g.,
                pollution sensitive vs. tolerant species, surface vs. subsurface feeders). A subset of six
                candidate metrics that best discriminated between reference and degraded sites was identified
                for possible inclusion in the index. Scoring criteria for each of these metrics were developed
                based on the distribution of values at undegraded sites (score of 1, if value of metric for sample
                being evaluated was in the lower 10th percentile of corresponding reference-site values; score
                of 3, if value of metric for sample was in the l0th-50th percentile of reference-site values; or
                score of 5, if value of metric for sample was in the upper 50th percentile of reference-site
                values). Scoring criteria were determined separately for each metric and habitat type.

                        Forty different combinations of the six candidate benthic metrics were further evaluated
                to determine which represented the best combined index. For each, a combined index value
                was calculated by assigning a score for each component metric (based on the individual scoring
                criteria) and then averaging the individual scores. A combined score < 3 was used to suggest
                the presence of a degraded benthic assemblage (very unhealthy to some apparent level of
                stress). The metric combination that produced the highest percentage of correct classifications
                (i.e., agreement with predictions of sediment bioeffects based on the various exposure
                measures) was then selected to represent the final index.


                Results
                        The final index was the average score of four metrics: total abundance, number of
                species, 100 % - % abundance of the two most dominant taxa, and % abundance of pollution-
                sensitive taxa. Percent pollution-sensitive taxa consisted of the percent of total faunal
                abundance represented by Ampeliscidae + Haustoriidae + Hesionidae + Tellinidae +
                Lucinidae + Cirratulidae + Cyathura polita and C burbanki.

                        This combined benthic index correctly classified stations 93 % of the time in the
                developmental data set and 75 % of the time in the independent validation data set (Table 4.9. 1
                and Figure 4.9. 1). Figure 4.9. 1 further illustrates that stations with index values below 3
                (suggestive of some apparent stress to highly degraded conditions) usually coincided with sites
                considered to be degraded based on a combination of chemistry and toxicity data, and that
                stations with scores of 3 or higher usually coincided with undegraded sites. Agreement was
                the highest at the two ends of the scale. Thus, the evaluation of sediment quality based on the
                benthic index appears to agree reasonably well with predictions of sediment bioeffects based on
                the combined exposure data. Additional comparisons revealed that the benthic index detected a
                higher percentage of samples where bioeffects were expected (based on contaminant bioeffect
                exceedances) than did any of the four individual sediment bioassays (Figure 4.9.2).



                                                            References


                Engle, V.D., J.K. Summers, and G.R. Gaston. 1994. A benthic index of environmental
                        condition of Gulf of Mexico estuaries. Estuaries, 17 (2): 372-384.


                                                                82










           Hyland, J.L., T.J. Herrlinger, T.R. Snoots, A.H. Ringwood, R.F. Van Dolah, C.T.
                  Hackney, G.A. Nelson, J.S. Rosen, and S.A. Kokkinakis. 1996. Environmental
                  Quality of Estuaries of the Carolinian Province: 1994. Annual Statistical Summary for
                  the 1994 EMAP-Estuaries Demonstration Project in the Carolinian Province. NOAA
                  Technical Memorandum NOS ORCA 97. NOAA/NOS, Office of Ocean Resources
                  Conservation and Assessment, Silver Spring, MD.

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

           Karr, J.R., K.D. Fausch, P.L. Angermeier, P.R. Yant, and I.J. Schlosser. 1986. Assessing
                  biological integrity in running waters: A method and its rationale. Special Publication
                  5. Illinois Natural History Survey, Champaign, Illinois.

           Long, E.R., D.D. MacDonald, S.L. Smith, and F. D. Calder. 1995. Incidence of adverse
                  biological effects within ranges of chemical concentrations in marine and estuarine
                  sediments. Envir. Man., 19: 81-97.

           Long, E.R. and L. G. Morgan. 1990. The potential for biological effects of sediment-sorbed
                  contaminants tested in the National Status and Trends Program. NOAA Technical
                  Memorandum NOS OMA 52. U.S. Department of Commerce, National Oceanic and
                  Atmospheric Administration, National Ocean Service, Rockville, MD.

           MacDonald, D.D. 1994. Approach to the assessment of sediment quality in Florida coastal
                  waters. Vols. I-IV. Report prepared for Florida Department of Environmental
                  Protection.


           Weisberg, S.B., J.B. Frithsen, A.F. Holland, J.F. Paul, K.J. Scott, J.K. Summers, H.T.
                  Wilson, R. Valente, D.G. Heimbuch, J. Gerritsen, S.C. Schimmel, and R.W. Latimer.
                  1992. EMAP-Estuaries Virginian Province 1990  'demonstration project report. U.S.
                  EPA Environmental Research Laboratory, Narragansett, R.I. EPA/600/R-92/100.

           Weisberg, S.B., J. A. Ranasinghe, D.M. Dauer, L.C. Schaffner, and J.B. Frithsen. 1997.
                  An estuarine benthic index of biotic integrity (B-IBI) for Chesapeake Bay. Estuaries, 20
                  (1): 149-158.










                                                       83









               Table 4.9.1 Classification efficiencies of the Carolinian Province benthic index.


                                       1994 "development" data     1993/95 "validation" data
                                                    % Correct                   % Correct
               Habitat        Sites     Classifications              Sites      Classifications


               Oligo. - Mesohaline,       20            90            46            78
               All Latitudes


               Poly. - Euhaline,          24            92            13            85
               Northern Latitudes


               Poly. - Euhaline,          22            95            27            74
               Middle Latitudes


               Poly. - Euhaline,            9           100           10            50
               Southern Latitudes


               Overall                    75            93            96            75




































                                                             84













                                    50

                                                                        Habitat
                                    40                                 Undegraded
                                                                       Degraded

                                                                              77@eMWF
                              0     30
                                                                                T',         /1,1/h, . . . . . .
                                                                              W,

                                                                              rgft",
                              4--
                                                                              01.
                                                                              'Re,
                                                                              gg/              A
                              0
                                    20
                              C


                                                                              01,
                                                                                           00/11/1 i
                              a)                                              "EM,
                                    10                                        K


                                                                              -0/1
                                                                                               MOMIN

                                                                                                                  0
                                       0
                                                1.0-1.5       2.0-2.5         3.0-3.5       4.0-4.5           5.0
                                                Unhealthy        Some                         Healthy
                                                 Benthos -@--Stress                          Benthos

                                                                          Index Score



                              Figure 4.9. 1. Frequency distribution of index scores for undegraded vs.
                              degraded stations in 1993/1995 "validation" data set.




















                                                                              85









                                                                          Benthic Index vs. Bioassays

                                                        Percent of Contaminated Stations Showing Bioeffects
                                                 100



                                                  80



                                          a)
                                          0       60


                                          a)
                                                  40
                                          D
                                          a
                                          X
                                          Ui
                                          -0
                                                  20



                                                     0
                                                                  Amphipod        Amphipod        Microtox      Seed Clam        Benthic
                                                                  Ampefisca       Ampefisca        Vibrio       Mercenaria        Index
                                                                    abdita          verrilli       fischeri     mercenaria


                                          Figure 4.9.2. Comparison of the percent of expected bioeffects detected
                                                                                                            a
                                          with the benthic index vs. sediment bioassays. Percent expected bioeffects
                                          = # stations (1995 core & supplemental) where an effect was detected
                                          stations with @! I ER-M/PEL or @! 3 ER-L/TEL exceedance.






















                                                                                             86










             4.10     CHESAPEAKE BAY BENTHIC COMMUNITY RESTORATION GOALS

                      Ananda Ranasinghe', Stephen Weisberg2     , Daniel Dauer 3, Linda Schaffher4,
                                           Robert Diaz4 and Jeffrey Frithsen'

                                              'Versar Inc. Columbia, MD
                        2Southern California Coastal Water Research Project, Westminster, CA
                                          3Old Dominion Univ., Norfolk, VA
                                4Virginia Inst. of Marine Science, Gloucester Point, VA

                   Benthic macroinvertebrate assemblages have been an integral part of the Chesapeake
            Bay monitoring program since its inception due to their ecological importance and their value
            as biological indicators. The condition of benthic assemblages reflects an integration of
            temporally variable environmental conditions and the effects of multiple types of environmental
            stresses. As such, benthic assemblages provide a useful complement to more temporally
            variable chemical and water quality monitoring measures.

                   While assessments using benthic monitoring data have been useful for characterizing
            changes in environmental conditions at individual sites over time, and for relating the condition
            of sites to pollution loadings and sources, the full potential of these assessments for addressing
            larger management questions, such as "What is the overall condition of the Bay?" or "How
            does the condition of various tributaries compare?" has not yet been realized. Regional-scale
            assessments of ecological status and trends using benthic assemblages are limited by the fact
            that benthic assemblages are strongly influenced by naturally varying habitat elements, such as
            salinity, sediment type, and depth. Such natural variability confounds interpretation of
            differences in the benthic community as simple responses to anthropogenic environmental
            perturbations. An additional limitation is that different sampling methodologies used in
            various programs often constrain the extent to which the benthic data can be integrated for a
            unified assessment.


                   The objective of this project was to develop a practical and conceptually sound
            framework for assessing benthic environmental conditions in Chesapeake Bay that would
            address the general constraints and limitations just described. This was accomplished by
            standardizing benthic data from several different monitoring programs to allow their
            integration into a single, coherent data base. From that data base a set of measures
            (Chesapeake Bay Benthic Restoration Goals) was developed to describe characteristics of
            benthic assemblages expected at sites having little evidence of environmental stress or
            disturbance (CBP 1994, Weisberg et al. 1997). Using these goals, benthic data from any part
            of the Bay could be compared to determine whether conditions at that site met, were above, or
            were below expectations defined for reference sites in similar habitats.

                   The approach used to develop these restoration goals was similar to that used by Karr et
            al. (1986) to develop an index of biological integrity for freshwater fish. A set of candidate

                                                           87








               attributes believed to have properties that differentiate high and low quality assemblages were
               first identified, and reference sites believed to be "minimally impacted" were designated.
               Properties of the biotic assemblages at these sites were then compared to assemblage properties
               at all other sites. Properties that differed significantly between these two groups of sites were
               selected as metrics to be included in the restoration goals. An index was developed to assist
               managers in identifying the extent to which these restoration goals were being achieved. The
               Restoration Goals Index (RGI) is calculated as the average score of metrics, after each metric
               is scored as 5, 3, or 1, depending on whether its value at an individual site approximated,
               deviated slightly, or deviated strongly from its value at the best reference sites.

                      The restoration goals were developed based on available data from seven benthic survey
               projects: the Maryland and Virginia Chesapeake Bay Benthic Monitoring Programs, U.S.
               EPA's Environmental Monitoring and Assessment Program (Holland et al. 1990), the
               Maryland and Virginia Biogenics studies, a James River study, and a study in the Wolf Trap
               area of the Chesapeake Bay. These seven projects were selected for several reasons: each
               provided data readily available on electronic media; collectively they provided sample
               representation in all salinity habitats of Chesapeake Bay; and all used a 0.5 mm. sieve in sample
               processing, which was a critical aspect of the study, since the numbers and types of organisms
               collected depend on the mesh size used to sieve the sediment.

                      The attributes incorporated into the restoration goals included metrics from each of the
               following five categories:

               1 .    benthic biodiversity measures

               2.     measures of assemblage abundance and biomass

               3.     life history strategy measures

               4.     measures of activity beneath the sediment surface

               5.     feeding guild measures

                      Restoration goals were developed independently for eight habitat classes defined by
               salinity and sediment type to ensure that natural differences in benthic communities related to
               these habitat factors did not confound interpretation of the indices. The eight habitat classes
               were determined by cluster analysis of the composite data set.

                      Restoration goals were developed using data from only the summer period, July 15th
               through September 30th. This restriction avoided seasonal variation that would confound
               interpretation of benthic, community responses to environmental degradation. The summer
               sampling period was common to six of the seven benthic survey projects. Using data from a
               different season would have reduced the data available because the various programs differed

                                                             88







            substantially in the extent of sampling during other seasons of the year. An index developed
            for summer was desirable because benthic communities are expected to show the greatest
            response to pollution stress during the summer.

                    Three approaches were used to validate the goals and the accompanying index. First,
            the RGI was computed for all samples taken from each reference site to test whether
            expectations of RGI values greater than three were met. This test indicated a high degree of
            correct classification; classification efficiency was more than 95 % in five of the seven habitat
            classes. The lowest correct classification efficiency for reference sites was 92.3 % in the high
            mesolialine mud habitat class. Second, RGI values were computed for all samples taken from
            degraded habitats to test whether expectations of RGI values less than three were met. This
            test used data that had been excluded from development of the RGI; therefore, it was an
            independent validation test. A high level of classification efficiency was observed in this test;
            classification efficiency was 85 % or better for degraded sites in five of the six habitat classes
            in which data from degraded sites were available. The one habitat class that did not validate as
            well was tidal freshwater. For the third validation test, sites that were sampled more than once
            during the summer of any year were identified, and the RGI was computed for each visit. RGI
            values at each site were evaluated for differences in status between visits within each year to
            ascertain the stability of the index. Instability of the index would indicate an unacceptable
            signal-to-noise ratio in the attributes. The results indicated that the RGI index was relatively
            stable. The correlation between RGI values for the first and second visits exceeded 80% for
            all habitats.


                    The validation results indicate that these preliminary restoration goals are effective for
            distinguishing between sites of high quality and those of lower quality in six of the seven
            habitats for which data were available for goal development. The only habitat class for which
            the restoration goals did not validate well was tidal freshwater. Although restoration goals
            validated well, additional analysis and development of goals appears to be appropriate before
            the goals are applied rigorously for environmental management purposes. Steps for further
            goal development are recommended.



                                                         References


            Holland, A.F. (ed.). 1990. Near coastal program plan for 1990: Estuaries. EPA 600/4-
                    90/033. U.S. EPA, ERL, ORD, Narragansett, RL

            Karr, J.R., K.D Fausch, P.L. Angermeier, P.R. Yant and I.J. Schlosser. 1986. Assessing
                    biological integrity in running waters: A method and it's rationale. 111. Nat. Hist. Surv.,
                    Pub #5. Champaign, Ill.

            Weisberg, S.B., J.A. Ranasinghe, D.M. Dauer, L.C. Schaffner, R.J. Diaz and J.B. Firthsen.
                    1997. An estuarine benthic index of biotic integrity (B_IBI) for Chesapeake Bay.
                    Estuaries, 20: 149-158.

                                                             89








               4.11      A PRELIMINARY STUDY OF THE USE OF MARINE BIOCRITERIA
                              SURVEY TECHNIQUES TO EVALUATE THE EFFECTS OF
                            OCEAN SEWAGE OUTFALLS IN THE MID-ATLANTIC BIGHT


                                                      George Gibson

                      US EPA, Office of Water, Office of Science and Technology, Washington, D.C.

               Objectives
                      This project investigates the practical, low cost application of marine biological
               community measurements and the nearfield/farfield survey technique, for use by coastal States
               as a water resource quality management tool. The methods applied here were derived from
               work reported by Pearson and Rosenberg (1978) and Mearns and Word (1982) with
               modifications.


               Study Methods
                      The study area was a 10 mi coastal reach between Bethany Beach, Delaware and Ocean
               City, Maryland (Figure 4. 11. 1). These are nearly adjacent resort communities on the Mid-
               Atlantic seaboard between Delaware Bay and Chesapeake Bay. Each has a secondary
               treatment municipal sewage discharge site about 1.5 nautical miles (nm) offshore. Discharge
               in both cases is through a diffuser at a water depth of approximately 40 ft (12 in). The
               Bethany Beach sewage treatment plant discharges about 14 mgd and Ocean City about 32 mgd.

                      A series of nine north-south trending stations were installed parallel to the coast at
               .intervals of about 1 nin, in about 40 ft depth of water and over medium to fine sandy bottoms
               to obtain as similar a habitat as possible. The stations were labeled "A" through "I", with
               station "C" at the Bethany Beach outfall and station "G" at the Ocean City outfall.

                      The variables measured were benthic fish and macroinvertebrate communities as
               reflected in indexes and metrics incorporating number of taxa and number of individuals per
               taxa. Fish surveys were made with a 20 ft (16 ft effective opening), I inch mesh otter trawl.
               Tows were made parallel to the shoreline at 2 knots over 0.5 mn with the station coordinates
               located at the mid-point of the tow. Benthic macroinvertebrate samples were collected with a
               0.1 m2 Smith-McIntyre grab or with a 0. 1 ni@ Young grab, and three replicates were taken for
               each sample at each station site.

                      Sampling surveys have been conducted twice a year in July-September and January-
               February since 1993 to determine if multiple season indexing is necessary or appropriate.
               While the mid-Atlantic area is considered to have four discrete seasons, benthic. communities
               are expected to be in flux during spring and fall and to be most stable in summer and winter
               (U.S. EPA 1994).

                      To make comparisons between the sample sites, habitat control in the survey design
               was maintained as well as possible by attention to four major variables; 1) sediment grain size,

                                                           90








            2) water depth, 3) water quality ( conductivity, temperature, depth, dissolved oxygen, pH,
            transmissivity, and 4) salinity. At the beginning of the project, sediment samples were
            collected from all nine stations and analyzed for heavy metals and a for a standard array of
            toxic contaminants. All results were insignificant, suggesting no other sources of biotoxicity or
            impairment indigenous to the immediate area.

                   In keeping with the objective of low cost, applications of standard, but robust
            taxonomic indexes were applied to the biological community data for impact detection, The
            underlying premise for the indices is that once the raw data for species and numbers of
            individuals per species are compiled, the investigator's primary question is whether or not
            there is a detectible impact. More refined indices and indicators can later be applied or
            developed as needed. In this regard, the treatments selected for this project were: total number
            of individuals, total number of taxa (species), evenness index, Simpson's dominance index,
            Margalef's taxa richness index, and the Shannon index of general diversity.


            Results
            Fish Survey Dat
                   Analysis of the fish data showed no significant differences in trawl data between the
            stations in summer or winter collections for either number of taxa or numbers of individuals.
            Qualitatively, taxa and number of individuals overall shifted considerably between summer and
            winter surveys at the nine stations. Greater numbers of both species and individuals (excepting
            winter runs of striped anchovy, Anchoa hepsetus) occur in the summer surveys.


            Benthic Macroinvertebrate Data
                   Benthic macroinvertebrate results have been much more promising, but the same
            seasonal trend observed with fish for number of taxa and number of individuals has prevailed.
            Summer measurements were much more indicative of the condition of the benthic
            macroinvertebrate assemblages. The data in this instance was for three replicates at each
            station twice a year for three years. Significant differences are evident between each of the
            outfall sites and the other stations in the summer data. The number of individuals show a
            gradient from high to low, proceeding from north to south, with an increase in the vicinity of
            the Ocean City outfall station. This suggests enhanced and or enriched conditions perhaps from
            the Delaware Bay discharge, and at the Ocean City outfall site.

                  When numbers of species were compared, a negative trend in outfall impact was evident,
           especially for the Bethany Beach outfall station (Figure 4.11.2). A similar pattern occurred at
           Ocean City, but was not as strong. Ludwig and Reynolds (1988) state that a simple count of the
           number of species present, for samples of equal size, avoids some of the problems of using
           indices which combine and may confound a number of variables that characterize community
           structure. However, in this instance, it appears that at least some indices enhance the
           measurement of outfall perturbations. Box plots of Margalef richness index (Figure 4.11.3)
           over the three years of summer data provide strong indications of the negative effect of both


                                                       91








              discharges on the benthic macroinvertebrate community. Simpson's dominance index and the
              Shannon index of general diversity reveal a similar effect.

              Discussion and Conclusions
                     The nearfield/farfield survey design for biological surveys, together with basic indices of
              community structure, appears to work equally well on the west coast and in mid-Atlantic coast
              open water environments. Summer benthic macroinvertebrate data from stations "A" and "C"
              were significantly different, lending confidence to the conclusion that the wastewater discharges
              were having a measurable impact on the coastal marine environment. This is of particular
              interest because routine water quality and sediment investigations at the sites failed to detect
              change between the outfalls and the surrounding stations. The standard indices such as
              Margalef's richness index, Simpson's dominance index, and Shannon's diversity index are
              robust and were entirely appropriate for this survey.

                     For biocriteria development and site monitoring, it is important to account for
              seasonality. For the mid-Atlantic Bight, late June to early September appears to be a time of
              relatively high, stable community productivity and an optimal index period if once a year
              sampling is preferred. Because Bethany Beach and Ocean City are summer resort communities,
              their populations increase at least ten-fold in warm weather (Bethany Beach, DE, and Ocean
              City, MD Chambers of Commerce, personal communication, 1990). Their lower winter
              discharge rates, together with a natural cyclic depletion of the marine community, may account
              for the failure of our data to reveal sewage impacts in this season. This may not be the case
              with a year-round municipality of fairly large size. In any case, if the responsible agency can
              afford to sample at least occasionally in winter, that baseline biological data may prove
              invaluable in the event of oil spills or other marine accidents.



                                                        References


              Ludwig, J.A. and J.F. Reynolds. 1988. Statistical Ecology. J.A. Wiley & Sons, NY, NY.
                            337pp.

              Mearns, A.J. and J.Q. Word. 1982. Forecasting the effects of sewage solids on marine
                     benthic communities. In, Ecolo2ical Stress and the New York Bi2ht: Science and
                     Management. G.F. Mayer (ed.). Estuarine Research Federation, Columbia, SC. p 495-
                     512.


              Pearson, T.H. and R. Rosenberg. 1978. Macrobenthic succession in relation to organic
                     enrichment and pollution of the marine environment. In, Oceanography and Marine
                     Biology, an Annual Review, 16:229-3 11.

              U.S. EPA. 1994. Chesapeake Bay Benthic Community Restoration Goals. U.S. EPA,
                     Chesapeake Bay Program, Annapolis, MD. CBP/TRS 107/94. 88pp.

                                                            92




























                                                                      A







                                                                         Bethany Beach Ocean Ouffall




                                                                  D

             DELA WARE




             MAR YLA ND


                                                         0 F




                                                                       Ocean City Ocean Oulfall


                                                          H
















           Figure 4. 11. 1 Offshore sampling locations off the coasts of Delaware and Maryland during the
                         summers of 1992-1994.


                                                         93







                                      Total Macroinvertebrate Taxa


                       100 -


                         80   -------------------------------------------------------------


                         60                -----------------------       ---------------------

                   z     40  ----------------      --- E-@- - -

                         20  ----------------  T   -----------------------      ---------- T-
                                              Bethany                      @c -ean
                          0 1                i Beachi      i       i      ! city
                                 A      B       C      D      E       F      G      H        I
                                                    Sample Station


             Figure 4.11.2 Number of taxa of macroinvertebrates collected at nine stations off the coasts of
             Delaware and Maryland during summers of 1992-1994. Lines and bars show maximum,
             minimum, 75 %, 25 % and median values.

                                                        94







                                            Richness Index

                    12 -


                      10 ---------------------------------------------------------------


                                        ----------      --- -----    I------------------------
                 c    6  -----------------------        ---T    --- ---------
                 C"
                 U                                  T                              T
                      4  ------------------ ------------------     Y    --- ------------
                                                                           "7--,
                      2   ---------------------------------------------------------------
                                          B ethany                        Ocean
                                           Beach
                      0 -,        i       i
                             A       B       C       D       E       F      G       H        I
                                                  Sample Station


           Figure 4.11.3 Taxa richness of macroinvertebrate species collected at nine stations off the
           coasts of Delaware and Maryland during summers of 1992-1994. Lines and bars show
           maximum, minimum, 75 %, 25 % and median values.

                                                       95









               4.12         THE BENTHIC RESPONSE INDEX: MEASURING IMPACTS ON
                                 BENTHIC ASSEMBLAGES IN SOUTHERN CALIFORNIA

                           Mary Bergen', Bob Smith2, Steve Weisberg', Don Cadien3, Ann Dalkey4,
                                         Dave Montagne   3 ,Jan StU113 , and Ron Velarde5

                            'Southern California Coastal Water Research Project, Westminster, CA
                                                 2EcoAnalysis, Inc., Ojai, CA
                 3Technical Services Division, County Sanitation Districts of Los Angeles Co., Whittier, CA
                         4Environmental Monitoring Division, City of Los Angeles, El Segundo, CA
                          5Metropolitan Wastewater Department, City of San Diego, San Diego, CA

                      To determine the best course of action, managers need to know if a resource has been
               impacted and how much it has been impacted. They also need to know if the condition of the
               resource is improving or degrading. Ideally, the manager should be able to evaluate the status
               of the resource using an objective, quantitative measure. This measure should clearly
               discriminate impacted from reference sites, be easily quantified and communicated, and be
               insensitive to differences in habitat, seasons or other sources of natural variability. In addition,
               there should be breakpoints or thresholds for the measure that indicate meaningful points of
               change in the resource, such as the limit of reference condition or the initiation of severe
               degradation in the resource. In southern California, we have been able to develop a benthic
               index which meets these criteria.


                      To develop the index, we followed an approach used in other programs. We assembled a
               database including information from known reference and impacted sites. We identified 26
               measures, including measures of diversity, abundance, biomass, species composition and mode
               of feeding, and tested the ability of the measures to discriminate impacted and reference sites.
               Most measures were not informative. However, two measures based on species composition,
               the Infaunal Trophic Index (ITI) and an ordination score, provided meaningful discrimination.

                      While either ITI or ordination could be used as an index, each has limitations. For this
               reason, we decided to create a new index that would capture the information in the measures but
               that would be easy to calculate and communicate. The Benthic Response Index (BRI) is the
               abundance-weighted average of the pollution tolerance of the species in the samples:


                                                             Y_ Pi (',/Noi)
                                                      BRI =   ----------------
                                                             Y 'VNoi

               where P is the average position of the species on an impact gradient.



                                                               96








                  To determine P for the species, the impact gradient was defined in an ordination space.
           The P value was calculated from the distribution of the abundance of the species on the impact
           gradient. P values were calculated for 537 species in three depth zones: 10-35m, 25-120m, and
           110-324 m.


                  Thresholds were defined for reference and four response levels: (1) marginal deviation,
           (2) loss in biodiversity, (3) loss in community function and (4) defaunation. The threshold for
           the reference was the 90th percentile of the index value in samples from reference areas. The
           endpoints of the distributions of species on the impact gradient were used to define response
           levels. The threshold for loss in biodiversity was exclusion of 25 % of the species pool. Loss in
           community function was exclusion of 75 and 90% of the arthropods and echinoderms,
           respectively. Defaunation was exclusion of 90% of the species pool.

                  The index was validated with data from monitoring programs that had not been used in
           index calibration. Validation was based on the ability of the index to reproduce known spatial
           and temporal patterns as well as the ability of the index to discriminate impacted from reference
           sites across habitat types. The index was validated in every test case.

                  Based on the results, we believe that we have developed an effective index for soft-
           bottom assemblages from Point Conception to the Mexican border between 10 and 250 m. Our
           approach to index development could be used in other geographic areas and with other
           assemblages, particularly in areas where there is clear separation of the impact gradient from
           natural gradients.
























                                                           97








                                   5.0 Workshop Discussion Group Summaries

               5.1                              VEGETATED HABITATS


                      The vegetated habitat category included submerged aquatic vegetation (vascular plants
               and algae), emergent wetlands, mangrove and kelp habitats. The group established several basic
               concepts as ground rules in evaluating which biological attributes are most appropriate for the
               habitat type under consideration. For the purposes of using biological indicators as measures of
               the condition of habitat designated as EFH, a healthy habitat implies a healthy and sustainable
               fish population. This links an ecosystem approach (as called for by the Magnuson Stevens Act)
               to the management of individual populations. A second consideration was that there must be
               some way to connect human impact effects to all chosen attributes. In addition, reference
               habitats result from evolutionary and biogeographic processes. With extensive discussion the
               working group came up with the following attributes that should be considered when developing
               indices of biological integrity for submerged aquatic vegetation.

               Submerged Aquatic Vegetation (SAV)
               Plant Attributes:
                      Diversity:
                             Vascular plants vs. Macroalgae
                             Genetic diversity
                      Abundance:
                             Shoot density, SAVs
                             Patchiness, SAVs
                             Plant exotics
                             Macroalgal biomass
                      Function:
                             Blade width/length
                             Epiphyte biomass
                      Population/processes:
                             Plant age structure
                             Perennial vs. annual species
                             Runners vs. tuft rhizomes
                             Rhizome density
                             Number flowering
                      Tolerance:
                             Frequency of wasting disease
                      Physical:
                             Total Organic Carbon in sediments
                             Light
                             Temperature




                                                            98










           Animal Communities in SAV Communities:
                  Diversity:
                         Infauna; epifauna; fish
                  Abundance:
                         Presences of large bivalves/ invertebrates
                         Avoid biomass
                  Function:
                         Number of nursery fish species
                         Number of resident species
                         Number of spawning species
                         Number of shellfish species (as bioindicators of SAVs but varies geographically)
                  Tolerance:
                         Fish lesions
                         Number of tolerant species

                  In the time available the working group was able only to outline the attributes for
           submerged aquatic vegetation habitats. However, general discussions provided several
           observations on the other types of vegetated habitats. For example, emergent wetlands are
           usually characterized by greater plant species richness and diversity and this attribute should be
           properly captured. One might also want to emphasize the importance of exotic species such as
           Phragmites. On the other hand, kelp beds and mangroves represent much more limited plant
           diversity communities and are more "monocultures" like SAV communities.

























                                                         99










               5.2                                  BENTHIC HABITATS


                      The open water benthic habitat category included soft bottom, hard bottom, and live
               bottom substrates. Soft bottom habitats were defined as having unconsolidated sediment,
               including anything from silt-clay deposits to coarse sand or gravel. Hard bottom habitats
               included cobble, consolidated rock and other solid surfaces to which benthic organisms can
               attach, primarily in the near shore and intertidal zone or estuaries. Live bottom was considered
               to be those habitats in which the physical structure of the habitat was composed of, or built by,
               sessile organisms, and included oyster bars, coral reefs and offshore benthic assemblages with
               significant three-dimensional relief. For the purpose of discussion, a distinction was made
               between estuarine (including submerged and intertidal), coastal shore zone, and offshore, in
               addition to substrate type. The coast was also subdivided into regions based on large scale
               oceanographic and geological features.

                      The following regions were delineated in the Atlantic and Gulf of Mexico coasts:

               1 .    Canadian border to Cape Cod;

               2.     Cape Cod to Cape Hatteras;

               3.     a combination of the area from Cape Hatteras to Cape Canaveral and from Tampa Bay to
                      the Rio Grande; and


               4.     southern Florida and the Caribbean


                      On the Pacific coast the regions were:

               1 .    Mexico to Pt. Conception;

               2.     Pt. Conception to the Columbia River;

               3.     Columbia River to Canada and Alaska's Pacific coast; and


               4.     Alaska's Bearing Sea and Arctic Ocean coasts.

                      The major determinants of regions on the Atlantic and Gulf coasts were climate, ocean
               circulation patterns, and geology. A primary determinant for regional definition on the Pacific
               coast was temperature, driven by climate and circulation patterns. Habitats were included out to
               the limit of the continental shelf and/or the Exclusive Economic Zone (EEZ). Inshore, habitats
               were considered up to the limit of the tidal fresh zone, or upstream to salmon spawning grounds,
               where that was relevant.





                                                               100








                  For each of the regions, the value and practicality of a series of parameters were
           discussed as measures of the health of the benthic community. The discussions were limited to
           biological measurements only. Measures of physical and chemical parameters of the habitat were
           excluded. As discussions proceeded, it became clear that the types of proposed measurement
           categories and metrics were basically similar for all estuarine environments, regardless of
           region, but this was not always the case for near shore or deep ocean habitats. The categories of
           parameters listed below applied to virtually all habitat types. Additional selected habitat or
           regional specifics were identified.

                         Infauna- community structure, composition, number of organisms and biomass by
                         taxa;
                         Shellfish, epibenthic fish, benthic foraging fish- community structure,
                         composition, number of organisms and biomass by taxa;
                         Percent spatial extent of 3-dimensional refugia- SAV, mangrove, reef etc.;
                         Percent spatial extent of living refugia vs. total refugia;
                         Dominance by selected species (opportunistic vs equilibrium);
                         Changes in dominance; and
                         Biomass of fish food.


                  In addition, parameters which apply specifically to estuaries included measures of
           resident vs. migratory species, and functional parameters of selected species (e.g., filtering
           capacity). Parameters which reflected contaminant impact such as body burdens, incidence of
           disease or the dominance of pollution tolerant species, were considered to be useful on a site-
           specific basis and are applicable to all habitats. Contamination was not considered to be an issue
           in the offshore habitat of the southeast, unlike some other regions. In shoreline and offshore
           habitats, the benthic epifaunal community was considered more appropriate than the infaunal
           community, depending upon the bottom type. The age structure of selected species, particularly
           in live bottom areas, was included as a measure of physical disturbance and/or chemical impact.
           In some locations, inclusion of shorebirds in the community metrics may be appropriate.

                  Eight general contrasts between degraded and healthy biological communities were
           enumerated. These were considered to be the functional ecological consequences of habitat
           degradation, and would be quantified by measurement of the specific parameters reviewed
           above. These are applicable to any habitat type including benthic, water column and vegetated
           habitats although some are targeted toward contaminant impacts and may not apply to all site
           specific locations. These contrasts were considered to represent the extremes of a continuum
           between healthy and degraded habitats. It is important to recognize that this continuum may not
           be linear, and may contain threshold points at which a small change in habitat integrity results in
           a large change in signal.






                                                         101









                                  Degraded                                       HeLlthy

                                low diversity                                 high diversity

                     high dominance by selected species                       low dominance

                   high proportion of immature individuals                  stable age structure

                      high proportion of tolerant species            low proportion of tolerant species

                    high proportion of r selection species         high proportion of K selection species

                         high chemical body burdens                     low chemical body burdens

                        high disease/lesion incidence                  low disease/lesion incidence

                     low coverage by biological refugia*           high coverage by biological refugia*



                       *emergent or submerged vegetation, coral reef, live bottom, oyster bar




























                                                             102










           5.3                              WATER COLUMN HABITATS


                  This session included the open water column habitats of freshwater streams, estuaries,
           near shore and coastal waters. The session opened with a discussion on the use of IBIs for the
           water column. Should an IBI be developed for the EFH of managed species under the
           Magnuson-Stevens Fishery Conservation and Management Act? Should separate IBIs be
           developed for each principle habitat type -- fresh, estuarine, coastal? The objective is to protect
           ecological units required to support a sustainable fishery and the managed species' contribution
           to a healthy ecosystem. Are the right components present? Are the organisms which are
           present healthy? Is the habitat capable of supporting transient organisms? The discussion
           progressed to what types of organisms might provide information for developing water column
           IBIs and the potential value of water column indicators. The group considered general classes of
           organisms, plankton and pelagic nekton.

                  The group proceeded with enumeration of potential water column IBI metric
           measurements that might be involved. These should include metrics which are indicative of
           community or population impact(s) from chemical, physical, and biological parameters.
           Examples of chemical attributes included contaminants, excess nutrients, and paralytic shellfish
           poisons adversely affecting the habitat or species of concern. Examples of physical impacts
           included low dissolved oxygen, stressful temperatures or salinities, inadequate light penetration,
           and altered stream flow or tidal circulation. Biological attributes might include the presence of
           pathogens, exotic species, or species compositions indicating degraded conditions (e.g., harmful
           algal blooms). Any taxa and/or trophic levels could be included as a biologically mediated
           stressor.


                  The group returned to discussing the approach to assessment of habitat quality with the
           use of IBI sampling. The first step might be to develop a rapid, relatively inexpensive screening
           IBI to determine whether or not a problem exists with a particular area or region as a diagnostic
           tool. After potential problem areas have been identified, a follow-up assessment protocol, which
           would be more site specific, could be developed, based upon investigation of potential stressors
           causing the habitat degradation. This might be accomplished by examining extant and historical
           data (e.g., NOAA's Eutrophication Survey, Status and Trends data, EMAP data, state agency
           information, land use practices, etc.). After potential factors have been identified, a selection of
           the most important variables would be possible, and an IBI should be developed to address
           them. Ultimately, it would be desirable to focus on IBI metrics that would address habitat
           quality of populations managed under Fishery Management Plans. Also, IBI metrics that would
           be indicative of, or sensitive to, impacts on habitat conditions before they affect managed
           populations to a significant degree is the objective. Thus, sensitive species, growth indicators,
           disease condition factors and abnormalities may be desirable to include. Some of these
           characteristics may only be exhibited in adults. It would be desirable to develop some
           characteristics specific to juveniles as well. Other water quality metrics were discussed,
           including the presence of fecal coliforms and parasites, biological oxygen demand, and


                                                           103









               phytoplankton phaeophytin/chlorophyll ratios might be a useful combination of chemical-
               biological indicators.

                      For plankton, the group considered both phytoplankton and zooplankton. For example,
               as waters degrade (i.e., move toward eutrophic conditions) phytoplankton composition can
               change from larger cell sized diatoms easily grazed by fish to smaller celled, less nutritious
               green algae often difficult or impossible for many species of fish to feed upon. Presence of
               brown tides (i.e., algal blooms) or toxic blooms of algae (e.g., Pfiesteria) would represent a
               degraded water column habitat. However, the group was concerned that water column IBIs
               might not provide the best bang for the buck. The water column may be too variable, too
               dynamic, or too transient in quality. Concern was expressed over seeing a signal within
               appropriate temporal and spatial scales. The group concluded that a holistic approach to IBIs
               should be developed and that the inclusion of more than just the water column might be
               desirable (i.e., water column and benthic for coastal, estuarine, and fresh waters).



































                                                             104










             5.4                                        SYNTHESIS


                    In habitats that have a high degree of structural complexity, such as reefs and vegetated
             areas, all components of the various communities should be assessed in a single index. In some
             cases, this may include the physical characteristics of the biota (e.g., canopy density, areal
             extent of live bottom). This makes for a more complex index because it will involve both
             plants and animals, and benthic and free swimming animals. However, the degree of
             ecological integration between them in these types of habitats is more intimate than in other
             habitat types. The functional connection between benthic habitats and the overlying water
             column habitat in open water and/or offshore habitats is less direct, except for bottom feeding
             fish which may come and go. In these types of habitats, the benthic and pelagic communities
             should be assessed independently. Assessment parameters in soft bottom habitats in estuaries
             were consistent regardless of location. Regional differences in assessment metrics and/or the
             appropriate community to sample were identified in coastal and hard bottom environments.
             Strictly pelagic communities will be difficult to assess with the IBI approach, due to high
             spatial and temporal variability. However, planktonic communities are more easily sampled in
             a quantitative fashion than nekton. Plankton also includes many trophic guilds (algae,
             zooplankton, larval stages of larger organisms) which respond to differing types of stressors.
             The process and activities required to develop bioindicators will be separate from the process
             of application of bioindicator measurement for monitoring and assessment purposes.

                     Table 5.1 summarizes the recommendations of potential metrics from the three
             breakout groups. Many potential metrics are common to all or multiple habitat types. While
             there is no need for commonality of metrics between different habitats, it is logical that there
             would be similar types of measures of ecological condition regardless of habitat. There was
             considerable overlap in the metrics in the diversity, abundance and condition categories. It is
             instructive that there is very little overlap in functional metrics. Functional roles of a species in
             a habitat is much more site specific than other parameters. In the tables, tolerant and intolerant
             refer to pollution indicative and sensitive, respectively.

                     Tables 5.2 - 5.4 summarize metrics currently in use in the development programs for
             benthos and fish discussed in the technical presentations of this workshop. These correspond to
             benthic estuarine, and water column estuarine or vegetated habitats in Table 5. 1. The tables
             also contain metrics from additional programs which were not represented at the workshop,
             but are summarized here in collaboration with the investigators. Overall, the metrics which
             have been found to work in the field, do not cover as wide a range of metrics as presented in
             the recommendation table. Clearly, our knowledge of marine biological communities on the
             ground does not match our expectations based on ecological theory. Either the theoretical
             paradigms need to be tested or refined in the marine environment, or sampling techniques need
             to be refined. On the other hand, except for species-specific metrics of abundance and/or
             function, which are largely site-specific, virtually all the metrics which are in use are
             addressed in Table 5. 1. The range and specificity of metrics utilized in fish IBI projects are
             greater than those used in benthic invertebrate projects. This probably reflects a greater

                                                              105








               relationships and simpler community structure of fish assemblages relative to benthic
               invertebrate communities. Several of the metrics could be placed in more than one category as
               they could represent more than one attribute (e.g.,%# tubificid oligochaetes reflects abundance
               and function).














































                                                             106







             Table 5.1          Summary of recommendations of potential metrics for marine IBI development.


                                                      VEGETATED                                            WATER COLUMN                                                  BENTHIC

               CATEGORY                    PLANTS                     ANIMALS               ESTUARINE                   COASTAL                     ESTUARINE                   COASTAL/LIVE

                                    diversity                   diversity (infauna,         diversity                   temporal diversity          diversity (infauna,         diversity (infauna -
                                    (marsh, macroalgae)         epifauna, fish)             (phytoplankton,             (phytoplankton              epifauna- incl. fish)       soft bottom, epifauna-
                                                                                            zooplankton, fish)          zooplankton)                                            incl. fish)

               DIVERSITY              species (marsh only)      # species                   # species                   # species                     species                   #species

                                    dominance (marsh            dominance                   dominance                   dominance                   dominance                   dominance
                                    only)

                                    shoot density (SAV)         abundance                   u plankton abundance        jA plankton abundance       abundance                   abundance
                                                                                            (temporal)                  (temporal)

                                    biomass                     biomass/taxa                  biomass                     biomass                   biomass/taxa (=age          biomass/taxa (=age
                                                                                                                                                    structure)                  structure)

               ABUNDANCE            chlorophyll                                             phytoplankton               phytoplankton
                                                                                            chlorophyll                 chlorophyll

                                    patchiness (SAV)

                                                                large bivalves

                                                                                            fish metrics

                                                                                            phytoplankton cell size     phytoplankton cell size

                                                                                            zooplankton size            zooplankton size

               FUNCTION                                                                     %zooplankton by             %zooplankton by
                                                                                            trophic guild               trophic guild

                                                                                            % larvae by trophic         % larvae by trophic
                                                                                            guild                       guild




                                                                                                     107









                Table 5.1 (Cont.)

                CATEGORY                                VEGETATED                                               WATER COLUMN                                                     BENTHIC

                                                                                                 % diatoms, green,            % diatoms, green,
                                                                                                 blue green                   blue green

                                                                                                                                                           % 3-D refugia                 % 3-D refugia

                                                                                                                                                           fish food biomass             fish food biomass

                                                                   # shellfish species

                                                                   # benthic species

                FUNCTION                                           # nursery species

                (cont.)                                            # resident species

                                                                   # spawning species

                                      blade area

                                      epiphyte biomass

                                      disease                      disease                       disease                      disease                      disease                       disease

                                      indicator species (r/K,      indicator species             indicator species            indicator species (r/K,      indicator species             indicator species (r/K,
                                      tolerant,                    (r/K, tolerant,               (r/K, tolerant,              tolerant,                    (r/K, tolerant,               tolerant,
                                      opport/equilib, exotic)      opport/equilib, exotic)       opport/equilib, exotic)      opport/equilib, exotic)      opport/equilib, exotic)       opport/equilib, exotic)

                                      age structure                                                                                                        age structure of live         age structure of live
                                                                                                                                                           refugia                       refugia

                                      tissue burdens                                                                                                       tissue burdens                tissue burdens

                                                                                                 freq, duration, timing,      freq, duration, timing,
                CONDITION                                                                        extent of blooms             extent of blooms

                                                                                                 freq, duration, timing,      freq, duration, timing,
                                                                                                 extent of toxic blooms       extent of toxic blooms




                                                                                                          108







             Table 5.1 (Cont.)

             CATEGORY                         VEGETATED                                     WATER COLUMN                                        BENTHIC


                                                                              coliform count          coliform count

                                                                              freq, duration, timing, freq, duration, timing,
                                                                              extent of fish kills    extent of fish kills


                                                                                                                              % live vs dead          % live vs dead
                                                                                                                              refugia area            refugia area

             CONDITION         rhizome density

             (cont.)             flowering

                               runners vs tuft
                               rhizomes




































                                                                                      109






                     Table 5.2         Metrics utilized in current benthic estuarine and coastal IBI development projects.

                      CATEGORY                 GoM Estuary           S.E. Ad. Estuary Chesapeake Bay             S. Calif. Coastal      EMAP Virginian          NY/NJ Harbors f
                                                                                                                                        Prov. t

                                               Shannon-Wiener                               Shannon-Wiener                              Gleason's D*

                      DIVERSITY                                      # species                                                                                    species

                                                                     %dominance
                                                                     (1/# top 2
                                                                     species)

                                                                     abundance              abundance                                                           abundance

                                                                                            biomass                                                             biomass
                      ABUNDANCE                                                             %taxa >5cm

                                               %# tubificid                                                                               tubificid
                                               oligochaetes                                                                             oligochaetes*

                                                                                                                                          spinoids*

                                               %# bivalves


                                                                                            % biomass
                                                                                            > 5cm deep*

                      FUNCTION                                                              %carnivores &                                                       % # carnivores &
                                                                                            omnivores*                                                          omnivores

                                                                                            % deep deposit
                                                                                            feeders*


                                                                     % sensitive taxa       % tolerant taxa,                                                    % tolerant taxa

                      CONDITION                                                             % intolerant                                                        % intolerant taxa
                                                                                            taxa*


                                                                                                                 cumulative taxa-
                                                                                                                  specific tolerance
                     *Salinity specific/normalized
                     t Strobel et.al. 1995. Statistical Summary: EMAP-Estuaries Virginian Province, 1990-1993. EPA/620/R-94/026.
                                            I                                            L


                     tRanasinghe et.al. (in review). A benthic index of biotic integrity for the New York/New Jersey Harbor. J. Aq. Ecosystem Stress and Recovery.



                                                                                                        110









                Table 5.3           Metrics utilized in current vegetated habitat IBI development projects.


                  CATEGORY                  New Englandt SAV Chesapeake B.t
                                                                     SAV

                                             species                    species

                  DIVERSITY                 dominance (#
                                            species =90%)

                                             individuals                individuals
                                            (or biomass)             (or biomass)
                  ABUNDANCE
                                            # estuarine spawning     # estuarine spawning
                                            sp.                      sp.

                                            # resident sp.           # resident sp.

                                            % benthic                % benthic
                  FUNCTION                    or biomass)               or biomass)

                                             nursery sp.                nursery sp.

                                                                        benthic sp.

                                                                        invertevore sp.

                  CONDITION                 % diseased
                tDeegan et al. (in preparation),








                    Table 5.4           Metrics utilized in current water column habitat IBI development projects.

                      CATEGORY                 Chesapeake B.           Lake Erie                Texas NRCC t            Chesapeake B.
                                               water column                                     (seine*, trawl +)       Plankton@

                                               # species                 species                # taxa*+

                      DIVERSITY                dominance                                        dorninance*
                                               species = 90 %)

                                                trawl species

                                                                         sunfish species

                                                                       #phytophilic
                                                                       species

                                                                       # benthic species

                                                                                                                        Margalef
                                                                                                                        (zooplankton)

                                                individuals W/O        # individuals W/O          individuals*          mesozooplankton
                                               menhaden                gizzard shad                                     #/m 2
                      ABUNDANCE                                                                                         Microzooplankton
                                                                                                                        #/in,

                                                anadromous
                                               species

                                                                                                                        mesozooplankton
                                                                                                                        biomass

                                                                                                                        microzooplankton
                                                                                                                        biomass

                                                                                                % penaeid* +

                                                                                                % shad or
                                                                                                anchovy*+



                                                est. spawners
                                               residents

                                               % benthivores

                      FUNCTION                 % carnivores            % top carnivores
                                               %planktivores           % omnivores

                                                                                112









                 Table 5.4 (Cont.)


                 CATEGORY                 Chesapeake Bay           Lake Erie               Texas NRCCt               Chesapeake Bay
                                          water column                                     (seine*, trawl+)          Planktont

                                                                   % lake assoc.
                                                                   individuals

                                                                                                                     %microzooplankton:
                                                                                                                     mesozooplankton
                 FUNCTION

                 (cont.)                                                                                             %calanoid:
                                                                                                                     cyclopoid +
                                                                                                                     cladocerans

                                                                                                                     % Bosmina sp.

                                                                     intolerant species

                 CONDITION                                         % tolerant
                                                                   individuals


                                                                   % exotic
                                                                   individuals


                                                                   % diseased                                        A% mean abundance
                tGuillen, G.J. 1996. Development of a Rapid Bioassessment Method and Index of Biotic Integrity for Tidal
                Streams and Bayous located along the Northwest Gulf of Mexico. 1996. TNRCC Special Report. Houston, Texas.
                t Alden et al. (in preparation) Long-Term. Trends in Zooplankton Indices of Environmental Health in the
                Chesapeake Bay and its Tributaries, Draft Report, Ches. Bay Prog.
























                                                                            113








                                       6.0 Workshop Consensus and Conclusions

                        Based on knowledge gained from preliminary studies, the IBI approach will be useful
                for assessing habitat quality in two primary ways. It brings together multimetric information to
                describe the quality of the biological community in simple, yet quantitative terms, and can be
                used for technical ecological assessment or to formulate research hypotheses for testing. The
                approach was specifically designed to assess environmental harm resulting from anthropogenic
                stressors. It provides a more site-specific assessment of target habitats than EMAP, which
                provides a probabilistic assessment over an entire region. This will be essential for application
                to EFH quality assessment. The IBI approach addresses a broader range of habitats and
                stressors than the Status and Trends or Mussel Watch approach, which are specifically geared
                toward contaminant exposure.

                        In addition to the regulatory need for site specific biological measurements, it is useful
                to be able to represent the condition of complex ecosystems concisely, by means of composite
                indices or simple graphics, so that managers and non-specialists can readily evaluate and
                compare information, establish goals, and set priorities for remediation or protection. This
                requires the use of succinct, understandable statistics that are also meaningful, reproducible
                and technically valid. Indicators are essential for:

                1.      determining management priorities;

                2.      measuring the effectiveness of management actions;

                3.      measuring progress towards restoration goals;

                4.      developing the capability to predict environmental consequences of management
                        options; and

                5.      communicating to the general public.

                        Technical formulation and testing of an IBI for a specific habitat requires a logical
                accumulation of data on parameters specifically selected because they are considered to be
                symptomatic of the ecological consequences of anthropogenic degradation. Point and non-point
                source runoff, toxic contamination, hydrologic alteration of watersheds and overharvest all
                affect biological communities. However, direct, quantitative cause-and-effect relationships
                between specific activities and ecological consequences are difficult to assess due to the
                complex interactions between ecosystems and anthropogenic stressors. No single parameter
                such as diversity, richness, indicator taxa or abundance, has been identified which can reliably
                distinguish between degraded and undegraded habitats in disparate environments, or in
                response to different stressors. The underlying ecological principle of the IBI is that a healthy
                community requires a diversity of intact ecosystem functions and processes to persist.
                Confirmation of deleterious effects at the community level is an inherent confirmation that

                                                                114








            population level effects are occurring within that community. IBIs also provide a mechanism to
            support contrasts of similar habitats in different regions. Data gaps and method deficiencies
            will become readily apparent in the process. While the cumulative index may be a ranking
            based on a number of metrics, the quantitative behavior of those metrics in relation to each
            other, and our ability to assess them in relation to anthropogenic impacts is instructive. The
            detailed information from individual metrics is not lost in the process. The IBI approach
            provides a framework for assessing habitat quality with a consistent, technically defensible
            method. It has a demonstrated utility in fresh water environments as a technical assessment
            method and as a management tool.

                   One difficulty with the application of IBI techniques to complex marine systems has
            been the relative lack of intimate knowledge of the ecological roles and interactions of specific
            species and/or functional guilds, compared to fresh water systems. Therefore, a basic element
            of any future IBI development work is simple taxonomic and natural history documentation of
            the species selected for use as markers of stress. Data gaps in life histories of critical species,
            including the degree of natural variation, must be identified and resolved. While it is preferable
            that metrics can be related to known functional aspects of an ecosystem, this has not always
            been the case. For example, if the presence of a specific taxonomic group or trophic guild of
            organisms is shown to be sensitive to habitat degradation, a measure of their abundance may be
            used as a meaningful metric. The presence of species with demonstrated sensitivity to heavy
            metals contamination is one example. (Indicator species for specific stresses must be selected
            with great care because of the potential for differential sensitivity to different stressors, and/or
            in different environmental settings.) Alternatively, the presence of a species or guild which
            correlates with some known measure of habitat degradation, without any specific knowledge of
            cause and effect, has been used successfully. The important aspects are that the metric must be
            demonstrated to be correlated with habitat degradation, and that this relationship can be
            quantified. Formulation of management options is difficult unless this 'dose-response'
            relationship can be demonstrated. To overcome this problem, statistical methods to select
            metrics from an array of potential measures have been successfully utilized to correlate
            ecological responses with anthropogenic stressors. These methods have included discriminant
            analysis, correlation coefficients, cluster analysis and multiple regression techniques. A
            comparative assessment of these methods has not been performed.

                   A related problem is the definition of what constitutes a reference condition. A-priori
            selection of 'reference' sites based upon one set of parameters (e.g., contaminants) have not
            been tested for efficacy in habitats which may have been impacted by other stressors (e.g.,
            eutrophication). Also, the delineation of a pristine reference site is problematic in regions
            where anthropogenic impacts are ubiquitous.

                   It is assumed there is a continuum of biological response between degraded and
            undegraded conditions, although the biological response may not be linear, and there are
            probably thresholds beyond which dramatic ecological changes occur. Some metrics have been
            shown to be bimodal. Negative and positive signals must be selected carefully. For example,

                                                          115








                diatom blooms in New York bight might be considered as indicative of healthy conditions.
                However, if early diatom blooms, resulting from eutrophication, cause Si limitation, this may
                result in a subsequent Nanachloris bloom. Because this is a smaller species, menhaden cannot
                feed on them. Thus, in this case, an early diatom bloom results in a deleterious dominance
                shift in the phytoplankton community.

                               It is not necessary to sample all subunits of an ecosystem. This would not be
                possible in any case, as all gear is selective to some degree. Assuming the ecosystem is
                integrated at some level, assessment of specific habitat types and/or locations within a system
                should be adequate. The level of effort for a given location will have to be determined on a
                site-specific basis. Locations for follow-up programs will build upon existing efforts and will
                ftirther develop methods in a consistent approach. In addition, specific index periods will need
                to be established for specific habitats. This will vary in different parts of the country for the
                same habitat type but should consider a time of year when environmental conditions and the
                action of stressors are relatively stable. The process will also need to include consideration of
                sampling efficiency, expense, safety and complexity.































                                                              116








                                             7.0 Follow-up Actions

                   Under the provisions of the Magnuson-Stevens Fishery Conservation and Management
            Act, NMFS must describe and map essential fish habitat. This process will also involve
            characterizing habitat quality. NMFS should move forward to identify appropriate attributes
            that would constitute biological indicators of habitat quality for the following habitat types:
            SAV, riparian, estuarine benthic/water column, coastal benthic, and coral reef habitats.
            Ongoing activities around the nation that are involved in developing and applying biological
            indicators should be inventoried (Figure 7. 1). Estuarine fish bioindicators have been, or are
            being developed, in Massachusetts, Chesapeake Bay, North Carolina, Florida and Texas.
            Investigations on the transferability of fish community bioindicator metrics for submerged
            aquatic vegetation (SAV) developed for Cape Cod estuaries and tested in Chesapeake Bay, and
            from Chesapeake Bay pelagic habitats to coastal embayments have been instructive. The degree
            of modification to the metrics which was necessary to adapt the systems to different regions
            was relatively straight forward. Benthic indicator development projects have employed
            complex mathematical schemes to develop metrics, due to the more complex and less
            understood biological communities associated with benthic invertebrate communities. Current
            development projects in the New York/New Jersey harbors, the Virginian province,
            Chesapeake Bay, SE Atlantic, and Gulf of Mexico rely heavily on EMAP data. Chemical
            contamination data has been used extensively to guide definition of reference sites, and
            therefore metric selection. Coastal benthic efforts on the Atlantic and Pacific continental
            shelves have taken divergent approaches from estuarine studies due to the more diffuse nature
            of impacts in those habitats. However, gradients of habitat degradation can be identified and
            quantified.

                   NMFS must develop partnerships with other Federal, state, university and private
            research institutions that are involved, or interested in developing and applying indices of
            biological integrity. Maximum use of ongoing programs should be made. A coordinated IBI
            development effort for South Atlantic estuarine benthic habitats is already well under way
            between the South Carolina Department of Natural Resources, NOAA/NOS, and U.S. EPA.
            NMFS should participate in that effort, and consider adopting the existing system. A benthic
            invertebrate IBI for northwest Pacific salmon spawning streams has been proposed, based on
            data gathered in Washington and Oregon streams. That assessment approach should also be
            evaluated for utility in EFH quality assessment. In addition U.S. EPA is developing a
            program for a mid-Atlantic integrated assessment of water and benthic habitats in estuaries and
            streams, which NMFS should become involved in.

                   A program should be devised to initiate the necessary data collection efforts addressing
            both evaluation of existing data, and the design of a systematic sampling and research program
            to fill data gaps, and generate new data to develop fish community IBI metrics. The major
            steps in the process should include:

            I .    Prioritization of habitat types and identification of habitat delineation parameters;

                                                         117











                2.     An inventory of existing data residing in state and Federal agencies which would be
                       appropriate for metric development in selected habitats;

                3.     Development of a research program to derive quantitative relationships between
                       potential metrics and anthropogenic stress, including contaminants, eutrophication, and
                       physical habitat alteration. The response pattern and mechanism(s) of cause and effect
                       are particularly important if the IBIs are to have direct management utility. In addition,
                       conduct research, where necessary, to illuminate the ecological role of potential
                       indicator species, including their natural variation and limiting factors;

                4.     Identification of spatial and temporal gradients, and gradients between habitat types that
                       will influence the distributional pattern of biota in a natural setting. This step, together
                       with item #3 will be essential for designation of reference sites and derivation of
                       regional vs. site-specific IBIs;

                5.     Initiate sampling and/or assessment of existing data sets to identify potential metrics,
                       gear requirements, sample requirements and statistical treatment;

                6.     Metric evaluation in independent validation sites;


























                                                               118








                        Current Marine Bioindicator Development Progr


                                                         Fish Community                                CT DEP
                                                         Benthic Community               MD DNR    EPA

                                                         Phytoplankton Community
                                                  A      Zooplankton Community                                     M
                                                 ,mm,,   Regional Benthic Community
                                                                                                                   D
                                                  0      Coastal Benthic Community                               EPA

                        NOAA                                                                                     CBF
                      CA DWR                                                                                    NC D
                      & SF

                        SCCWRP                                                                                NOAA/E
                           PERL




                                                                                                           FL DEP




                                                                      TX NRCC   EPA






                  Figure 7.1   Map of locations of site-specific and regional marine IBI development programs.

                                                                             119







                                                     Appendix 1

                                                        AGENDA
                                                    July 14-15, 1997

               Day 1- Introduction and Presentations of Existing Approaches and Applications

                            I. Hartwell/D. Brown (NMFS) - Welcome and introduction

                            J. Karr (Univ. Wash) - Attaining environmental goals

                            C. Linder (MD DNR) - An estuarine IBI for Chesapeake Bay

                            M. Weaver (Woods Hole Mar. Biol. Lab.) - Estuarine biotic integrity index

                            K. Summers (US EPA) - An index of benthic condition

                            M. Bergen (S. Cal. Coastal Res. Proj.) - The benthic response index

                            A. Ranasinghe (VERSAR Inc.) - Chesapeake Bay benthic restoration goals

                            G. Gibson (US EPA) - Marine biocriteria survey techniques

                            J. Hyland (NOS) - A benthic index for estuaries of the S.E. US

                            R. Thoma (OH EPA) - Ohio biological monitoring program
                                                  - Lake Erie and lacustuary monitoring program

                            F. Holland (SC DNR)- Assessment of watershed development on tidal creeks

                            M. Monaco/P. Orlando (NOS) - Spatial framework for EFH data collection

              Day 2 Assessment of Metrics and Index Derivation Methods

              A)     Morning breakout groups to assess matrix of regional metrics types by syste

                     Starting point.
                            For a given habitat group, what are the parameters, or types of parameters,
                     necessary to assess habitat condition? How would those parameters be measured? How
                     would you combine the measurements to arrive at a conclusion?

                     Group I - Vegetated {SAV (vascular plants & seaweed), emergent marsh, mangrove,
                            kelp)

                                                          120









                   Group 2 - Open water benthic (soft bottom, hard bottom, live bottom (oyster bar, coral
                          reef, offshore benthic assemblage), deep vs shallow & intertidall

                   Group 3 - Water column {tidal-fresh, estuarine, near-shore, coastal)

            B)     Reconvene to compare and critique metric parameters and approache

            Q      Afternoon breakout groups to continue discussions about parameters based on morning
                   sessions.


            D)     Final session to formulate a consensus statement on a bioindicator framework for EFH
                   habitat quality assessment, identify research needs and, potential pilot 12roRram
                   locations/data bases.













































                                                          121







                                                      Appendix 2



                                                      PARTICIPANTS


               Robert Boyles               (803) 727-2078       [email protected]
                      S.C. Sea Grant
                      Univ. of South Carolina
                      287 Meeting St.
                      Charleston, SC 29401


               Mary Bergen                 (714) 894-2222       [email protected]
                      Southern California Coastal Water Research Project
                      7171 Fenwick Lane
                      Westminster, CA 92683


               Dail Brown                  (301) 713-2325       [email protected]
                      NOAA/NMFS F/HC
                      1315 East West Hwy
                      Silver Spring, Md 20910

               Tracey Collier              (206) 860-3312       [email protected]
                      NOAA/NMFS NWFSC
                      2725 Montlake Blvd. E.
                      Seattle, WA 98112


               David Dow                   (508) 495-2249       [email protected]
                      NOAA/NMFS NEFSC
                      166 Water St.
                      Woods Hole, MA 02543-1026


               David. W. Evans             (919) 728-8752       dave.evans anoaa.gov
                      NOAA/NMFS SEFSC
                      101 Pivers Island Rd
                      Beaufort, NC 28516

               Gary Fitzhugh               (904) 234-6541       [email protected]
                      NOAA/NMFS SEFSC
                      3500 Detwood Beach Rd.
                      Panama City, FL 32408





                                                           122








            Michael Fulton               (803) 762-8576       mike. fulton@noaa. gov
                   NOAA/NMFS SEFSC
                   219 Ft. Johnson Rd.
                   Charleston, SC 29412

            George Gibson                (410) 573-2618       [email protected]
                   US EPA
                   839 Bestgate Rd
                   Annapolis, Md 21401

            Ian Hartwell                 (301) 713-2325       [email protected].
                   NOAA/NMFS F/HC
                   1315 East West Hwy
                   Silver Spring, Md 20910

            Fred Holland                 (803) 762-5107       [email protected]. state. sc.us
                   S.C. Dept. of Natural Resources
                   P.O. Box 12559
                   Charleston, SC 29422-2559

            Jeff Hyland                  (803) 762-5415       [email protected]
                   NOAA Carolinian Province Office
                   217 Ft. Johnson Rd.
                   P.O. Box 12559
                   Charleston, SC 29422-2559

            James R. Karr                (206) 685-4784       jrkarr@u. washington. edu
                   Univ. of Washington
                   Box 357980
                   Seattle, WA 98195

            Pete Key                     (803) 762-8596       [email protected]
                   NOAA/NMFS SEFSC
                   219 Ft. Johnson Rd
                   Charleston, SC 29412


            Cecilia Linder               (302)645-4384        [email protected]
                   Univ. Delaware
                   700 Pilottown Rd.
                   Lewes, DE 19958





                                                         123








               Mark Monaco                 (301) 913-3005      mark. monaco@noaa. gov
                      NOAA/NOS ORCA
                      1305 East-West Hwy
                      Silver Spring, Md 20910

               Paul Orlando                (301) 713-3000      porlando@seamail. nos. noaa. gov.
                      NOAA/NOS ORCA
                      1305 East-West Hwy
                      Silver Spring, MD 20910

               Paul Pennington             (803) 762-8620      [email protected].
                      NOAA/NMFS SEFSC
                      219 Fort Johnson Rd.
                      Charleston, SC 29412

               Ananda Ranasinghe           (410) 740-6085      [email protected]
                      Versar Inc.
                      9200 Rumsey Rd.
                      Columbia, MD 21045

               Casimere Rice               (425) 743-3307      [email protected]
                      NOAA/NMFS NWFSC
                      10 Park Ave..
                      Mukilteo, WA 98275

               Robert Stickney             (409) 845-3854      [email protected]
                      Texas Sea Grant
                      Texas A&M Univ.
                      1716 Briarcrest Dr., Suite 702
                      Bryan, TX 77802

               Kevin Summers               (904) 934-9244      summers. [email protected]. gov
                      US EPA, Gulf Ecology Division,
                      Sabine Island Dr.,
                      Gulf Breeze, FL 32561

               Roger F. Thoma              (216) 963-1141      roger. thoma@epa. state. oh. us
                      Ohio EPA
                      2110 E. Auroa Rd
                      Twinsburg, OH 44087





                                                          124









                     Jim Thomas                                          (301) 713-2325                        [email protected]
                                  NOAA/NMFS F/HC
                                  1315 East-West Hwy.
                                  Silver Spring, Md 20910

                     Robert VanDolah                                     (803)762-5048                         [email protected]. state. sc. us
                                  S.C. Dept. of Natural Resources
                                  Marine Resources Division
                                  P.O. Box 12559
                                  Charleston, SC 29422

                     Melissa Weaver                                      (423) 974-3065                        [email protected]
                                  Dept. of Ecology & Evolutionary Biol.
                                  Univ. of Tennesee. 569
                                  Dabney Hall
                                  Knoxville, TN 37996-1610

































                                                                                                      125






                                                                                                                                      U.S. GOVERNMENT PRINTING OFFICE: 1998-693-362    62088 Region No. 10





















































































                                                                                                  I
























                                                                                           -
                                                                       3 6668 14100 5233