Unitod State*
Environmental Protection Agency
Office of Water &
Office of Research and
Development
Office of Science and Technology
Health and Ecological Criteria Div.
Washington, D.C. 20460
EPA-822-R-93-014
September 1993 .
Sediment Quality Criteria
for the Protection  of
Benthic Organisms:
PHENANTHRENE

-------

-------
                                  CONTENTS
 Foreword	;    ft
 Acknowledgments	   iv
 Tables .'...'	   vi
 Figures   . .  .	     vii

 Introduction	     1-1
 Partitioning	     2-1
 Toxicity of Phenanthrene: Water Exposures . . .	     3-1
.Toxicity of Phenanthrene (Actual and Predicted): Sediment Exposures ......   4-1
 Criteria Derivation for Phenanthrene  .	        5-1
 Criteria Statement	     6-1
 References . .  . . . ..,...•			7-1
 Appendix A: Summary of Acute Values for Phenanthrene for Freshwater and ...
             Saltwater Species	   A-l
 Appendix B:  Summary of Data from Sediment Spiking Experiments with
             Phenanthrene	B-l

-------
                                      FOREWORD
                         ' ,                                          :
        Under the Clean Water Act (CTWA) the U.S. Environmental Protection Agency (U.S.
 EPA) and the States develop programs for protecting the chemical, physical, and biological
 integrity of the nation's waters.  Section 304(a)(l) directs the Administrator to develop and
 publish "criteria" reflecting the latest scientific knowledge on: (1) the kind and extent of effects
 -on human health and welfare, including effects on plankton, fish, shellfish, and wildlife, which
 may be expected from the presence of pollutants in any body of water, including ground water,
 (2) the concentration and dispersal of pollutants, or their byproducts, through biological, physical
 and chemical processes, and (3)  the effects of pollutants on biological community diversity,
. productivity, and stability. Section 304(a)(2) directs the Administrator to develop and publish
 information on, among other things, the factors necessary for the protection and propagation of
 shellfish, fish, and wildlife for classes and categories of receiving waters.

        To meet this objective, U.S. EPA has periodically issued ambient water quality criteria
 (WQC) guidance beginning with the publication of "Water Quality Criteria 1972" (NAS/NAE,
 1973).  All criteria guidance through late 1986 was summarized in an U.S. EPA document
.entitled "Quality Criteria for Water, 1986" (U.S. EPA, 1987). Additional WQC documents that
 update criteria for selected chemicals and provide new criteria for other pollutants have also been
 published. In addition to the development of WQC and to continue to comply with the mandate
 of the CWA,  U.S. EPA has conducted efforts to develop and publish sediment quality criteria
 (SQC) for some of the 65 toxic pollutants or toxic pollutant categories. Section 104 of the CWA
 authorizes the administrator to conduct and promote research into the causes, effects,  extent,
 prevention,  reduction and elimination of pollution, and to publish relevant information.  Section
 104(n)(l) in particular provides for study of the effects of pollution, including sedimentation in
 estuaries, on aquatic life, wildlife, arid recreation. U.S. EPA's efforts with respect to sediment
 criteria are also authorized under CWA Section 304(a).

       Toxic contaminants in bottom sediments of the nations's lakes, rivers, wetlands, and
 coastal waters create the potential for continued environmental degradation even  where water
 column contaminant levels meet established WQC. In addition, contaminated sediments can lead
 to water quality impacts, even when direct discharges to the receiving water have ceased. EPA
 intends SQC be used to assess the extent of sediment contamination, to aid in implementing
 measures to limit or prevent additional contamination, and to identify and implement appropriate
 remediation activities when needed.                                                  ,

       The criteria presented in this document are the U.S.  EPA's best recommendation of the
 concentrations of a substance that may be present  in sediment while still protecting benthic
 organisms from the-effects of that  substance.  These criteria are applicable to  a variety of.
 freshwater and marine  sediments  because  they are based  on the biologically available
 concentration  of the substance in sediments.  These  criteria do not protect against additive,
 synergistic or antagonistic effects of contaminants or bioaccumulative effects to  aquatic  life,
 wildlife or human health.
                                          11

-------
       The criteria derivation methods outlined in this document ace proposed  to provide
protection of benthic organisms from biological impacts from chemicals present in sediments.
Guidelines and guidance are being developed by U.S. EPA to assist in the application of criteria
presented in this document, in the development of sediment quality standards,  and in other
water-related programs of this Agency.

       These criteria are being issued in support of U.S. EPA'S regulations and  policy
initiatives. This document is Agency guidance only. It does not establish or affect legal rights
or obligations. It does not establish a binding norm and is not finally determinative of the issues
addressed. Agency  decisions in any particular case will be made by applying the law and
regulations on the basis of the specific facts.
                                           111

-------
                             ACKNOWLEDGEMENTS
 Principal Author

       "David J. Hansen

 Coauthors

        Walter J. Berry


        Dominic M. Di Toro


        PaulR.Paquin


        Laurie D. De Rosa


       Frank E. Stancil, Jr.

       Christopher S.  Zarba

Technical and Clerical Support

      Heinz P. Koffig

      Glen B. Thursby


      Maria R. Pamta


      Dinalyn Spears

      BettyAnne Rogers
 U.S. EPA, Environmental Research Laboratory,
 Narragansett, RI
 Science Applications International Corporation,
 Narragansett, RI

 Manhattan College, Bronx, NY;
 HydroQual, Inc., Mahwah, NT

 HydroQual, Inc.,
 Mahwah, NJ

 HydroQual, Inc.,
 Mahwah, NJ

 U.S. Environmental Research Laboratory, Athens, GA

 U.S. EPA Headquarters, Office of Water, Washington, DC
U.S. Environmental Research Laboratory, Athens, GA

Science Applications International Corporation
Narragansett, RI

NCSC Senior Environmental Employment Program
Narragansett, RI

Computer Science Corporation, Narragansett, RI

Science Applications International Corporation
Narragansett, RI
                                       IV

-------
      Persons who have made significant contributions to the development of the approach and
supporting science used in the derivation of sediment criteria for non-ionic organic contaminants
are as follows:
      Herbert E. Allen

      Gerald T. Ankley


      Christina E. Cowan

      Dominic M. Di Toro


      David J. Hansen


      Paul R. Paquin

      Spyros P. Pavlou
           » •  " -  '
      Richard C. Swartz


      Nelson A. Thomas
University of Delaware, Newark, DE

U.S. EPA, Environmental Research Laboratory,
Duluth, MN

Battelle, Richland, WA

HydroQual, Inc., Mahwah, NT;
Manhattan College, Bronx, NY

U.S. EPA, Environmental Research Laboratory,
Narragansett, RI

HydroQual, Inc., Mahwah, NJ

Ebasco Environmental, Bellevue, WA

U.S. EPA, Environmental Research Laboratory,
NewportjOR

U.S. EPA, Environmental Research Laboratory,
Duluth, MN
      Christopher S. Zarba
U.S. EPA Headquarters, Office of Water, Washington, DC

-------
 Table 2-1.    Phenanthrene measured and estimated log10KoW values.

 Table 2-2.    Summary of log10KoW values for phenanthrene measured by the U.S. EPA,
              Environmental Research Laboratory, Athens, GA.

 Table 3-1.    Chronic sensitivity of freshwater and saltwater organisms to phenanthrene.
              Test specific data.

 Table 3-2.    Summary of freshwater and saltwater acute and clhronic values, acute-chronic
              ratios, and derivation of final acute values, final acute-chronic ratios and final
              chronic values for phenanthrene.
•»
 Table 3-3.    Results of approximate  randomization test for the equality of freshwater  and
              saltwater FAV distributions forphenanthrene and approximate randomization test
              for the equality of benthic and combined benthic and water column (WQC) FAV
              distributions.

 Table 4-1.    Summary of tests with phenanthrene-spiked sediment.
            i " . e              *                                    .               •

 Table 4-2.    Water-only and sediment LCSOs used to test the applicability of the equilibrium
              partitioning theory for phenanthrene.
                                                          i"
 Table 5-1.    Sediment quality criteria for phenanthrene.

 Table 5-2.    Analysis of variance for derivation of sediment quality criteria confidence limits
              forphenanthrene.

 Table 5-3.    Sediment quality criteria confidence limits for phenanthrene.

 Appendix A. - Summary of acute values for phenanthrene for freshwater and saltwater species.
                                                                    i
 Appendix B. - Summary of data from  sediment spiking experiments with phenanthrene. Data
              from these experiments were used to calculate KQC values (Figure 2-1) and' to
              compare mortalities of amphipods with pore water toxic units (Figure 4-1) and
              predicted sediment toxic units (Figure 4-2).
                                          VI

-------
                                       FIGURES
                         s
                          :                                        -                  '  - •
Figure 1-1.   Chemical structure and physical-chemical properties of phenanthrene.

Figure 2-1.   Organic carbon-normalized  sorption  isotherm for  phenanthrene (top)  and
              probability plot of KQC (bottom) from sediment toxicity tests conducted by Swartz
-   .           (1991). The line in the top panel represents the relationship predicted with a log
                    4.46, that is CS(OC = KQC •  CD.
Figure 3-1.    Genus mean acute values from water-only acute toxicity tests using freshwater
              species vs. percentage rank of their sensitivity.  Symbols representing benthic
              species are solid, those representing water column species are open.  Asterisks
              indicate greater than values.  J = juvenile, L = larvae, X = unspecified life
              stage.

Figure 3-2.    Genus mean acute values from water-only acute toxicity  tests using saltwater
              species vs. percentage rank of their sensitivity.  Symbols representing benthic
              species are solid, those representing water column species are open.  Asterisks
              indicate greater than values.  A = adult, J = juvenile.

Figure 3-3. •   Probability distribution of FAV difference statistics to compare water-only data
              from freshwater vs saltwater (upper panel) and benthic vs. WQC (lower panel)
              data.

Figure 4-1.    Percent mortality of amphipods  in  sediments  spiked with acenaphthene or
              phenanthrene (Swartz, 1991), endrin (Nebeker et al., 1989; Schuytema et al.,
              1989), or fluoranthene (Swartz et al.,  1990; De Witt et al., 1992) and midge in
              sediments spiked with dieldrin (Hoke, 1992) or kepone (Adams et al., 1985)
              relative to  pore water  toxic  units.   Pore water toxic units  are ratios of
              concentrations  of chemicals measured in  individual treatments  divided by the
              water-only LC50 value from water-only tests.  (See Appendix B in this SQC
              document, Appendix B in the endrin, dieldrin, fluoranthene and acenaphthene
              SQC documents, and original references for raw data.)

Figure 4-2.    Percent mortality of amphipods  in  sediments  spiked with acenaphthene' or
              phenanthrene (Swartz, 1991), dieldrin (Hoke and Ankley, 1991), endrin (Nebeker
              et al., 1989; Schuytema et al., 1989) or fluoranthene  (Swartz et al., 1990; De
              Witt et al., 1992) and midge in dieldrin spiked sediments (Hoke, 1992) relative
              to "predicted sediment toxic units." Predicted sediment toxic units are the ratios
              of measured treatment concentrations for  each chemical in sediments  (/tg/goc)
              divided by the predicted LC50 0*g/goc) in sediments (KQC x Water-only LC50
              0*g/L) x 1 Kgoo/ljOOOgoc).  (See Appendix B in this document and Appendix B
              in the dieldrin, endrin, fluoranthene, and acenaphthene SQC documents for raw
              data).


                                          vii

-------
 Figure 5-1.   Comparison between SQC concentrations and 95% confidence intervals, effect
             concentrations from benthic organisms exposed to phenanthrene-spiked sediments.
             and sediment concentrations predicted to be chronically  safe in  fresh water
             sediments.  Concentrations predicted to be chronically safe (Predicted Genus
             Mean Chronic Values, PGMCV) are derived from tike Genus Mean Acute Values
             (GMAV) from water-only 96-hour lethality tests, Acute Chronic Ratios (ACR)
             and KQC values. PGMCV = (GMAV -r- ACR)!^. Symbols for PGMCVs are
             A for arthropods, O for fishes and D for other invertebrates.  Solid symbols are
             benthic genera; open symbols water column genera. Arrows indicate greater than
             values.   Error bars around  sediment LC50 values indicate observed range of
             LC50s. -        .

Figure 5-2.   Comparison between SQC concentrations and 95% confidence intervals, effect
             concentrations from benthic organisms exposed to phenanthrene-spiked sediments
             and sediment concentrations predicted  to  be chronically safe in salt  water
             sediments.  Concentrations predicted to be chronically safe (Predicted Genus
             Mean Chronic Values, PGMCV) are derived from the Genus Mean Acute Values
             (GMAV) from water-only 96-hour lethality tests, Acute Chronic Ratios (ACR)
             and KQC values. PGMCV =  (GMAV -5- ACR)Koc. Symbols for PGMCVs are
             A for arthropods, O for fishes and D for other invertebrates.  Solid symbols are
          >  benthic genera; open symbols water column genera. Arrows indicate greater than
             values.   Error bars around sediment LC50 values indicate observed range of
             LC50s.

Figure 5-3.   Probability distribution of concentrations of phenanthrene in sediments from
             streams,  lakes and  estuaries in the United States from 1986 to  1990, from the
             STORET (U.S. EPA, 1989b) database, compared to the phenanthrene SQC values
             of 18 ug/g in freshwater sediments  having TOC =  10% .and  1.8 jig/g in
             freshwater sediments having TOC = 1% and compared  to SQC values for
             saltwater sediments of 24 pg/g when TOC = 10 % and 2.4 ug/g when TOC=1 %.
             The upper dashed line on each figure represents the SQC value when TOC =
             10%, the lower dashed line represents the SQC when TOC  = 1 %.

Figure 5-4.   Probability distribution of concentrations of phenanthrene in sediments from
             coastal and estuarine sites from 1984 to 1989 as measured by the National Status
             and Trends Program (NOAA, 1991).  The horizontal line is the saltwater SQC
             value of 240
                                       viu

-------
      This report has been reviewed by the Health and Ecological Criteria Division, Office of
Science and Technology, U.S. Environmental Protection Agency, and approved for publication.
Mention  of trade  names or  commercial products does  not constitute endorsement or
recommendation for use.
                             AVATTJVRTTTTV MOT[CE

      This document is available to the public through the National Technical Information
Service (NTIS), 5285 Port Royal Road, Springfield, VA 22161.  NTIS Accession Number
xxxx-xxxxxx.
                                       IX

-------

-------
                                     SECTION 1


                                  INTRODUCTION


 1.1  GENERAL INFORMATION:


       Under the Clean Water Act (CWA) the U.S. Environmental Protection Agency (U.S.


 EPA) is responsible for protecting the chemical, physical and biological integrity of the nation's


 waters. In keeping with this responsibility, U.S. EPA published ambient water quality criteria


 (WQC) in 1980 for 64 of the 65 toxic pollutants or pollutant categories designated as toxic in


 the CWA.  Additional water quality documents that update criteria for selected consent decree


 chemicals and new criteria have been published since 1980.   These WQC are numerical


 concentration limits that are the U.S. EPA's best estimate of concentrations protective of human


 health and the presence and uses of aquatic life. While these WQC play an important role in


 assuring a healthy aquatic environment, they alone are not sufficient to ensure the protection of


 environmental or human health.                                            .


       Toxic pollutants in bottom sediments of the nation's lakes, rivers, wetlands, estuaries and


 marine coastal waters create the potential for continued environmental degradation even where


 water-column concentrations comply with established  WQC.   In addition, contaminated


 sediments can be a significant  pollutant source that may cause water quality degradation  to


persist, even when other pollutant sources are stopped.  The absence of defensible sediment
                                          r\

 quality criteria (SQC) makes it difficult to accurately assess the extent of the ecological risks  of


 contaminated sediments and to identify, prioritize and implement appropriate clean up activities
                                         1-1

-------
 and source controls.  As a result of the need for a procedure to assist regulatory agencies in
                          ;
                           J                                         !

 making decisions concerning contaminated sediment problems, a U.S. EPA Office of Science




 and Technology,  Health and  Ecological Criteria Division (OST/HEC) research team was




 established to review alternative approaches (Chapman, 1987).  All of the approaches reviewed




 had both strengths and weaknesses and no single approach was found to be applicable for SQC




 derivation in all situations (U.S. EPA, 1989a).  The equilibrium partitioning (EqP) approach was




 selected for non-ionic organic chemicals because it presented the greatest promise for generating

°T


 defensible national numerical chemical-specific SQC applicable across a broad range of sediment




 types. The three principal observations that underlie the EqP method of establishing SQC are:




       1.     The concentrations of non-ionic organic chemicals in sediments, expressed on an




            • organic carbon basis, and in pore waters correlate to observed biological effects




             on sediment dwelling organisms across a range of sediments.




       2.     Partitioning models can relate  sediment concentrations for non-ionic  organic




             chemicals on an  organic carbon basis to freely dissolved concentrations in pore



             water.




       3.     The distribution of sensitivities of benthic and water column organisms to




             chemicals are similar; thus, the currently established WQC final chronic values




             (FCV) can be used to define the acceptable effects concentration of a chemical




             freely-dissolved in pore water.




     , The EqP approach, therefore, assumes that: (1) the partitioning of the chemical between




sediment organic carbon and interstitial water is at equilibrium; (2) the concentration in either




phase can be predicted using appropriate partition coefficients and the measured concentration
                                         1-2

-------
in the other phase; (3) organisms receive equivalent exposure from water-only exposures or from
                         ;

any equilibrated phase: either from pore water via respiration, sediment via ingestion, sediment-


integument exchange, or from a mixture of exposure routes; (4) for non-ionic chemicals, effect


concentrations in sediments on an organic carbon basis can be predicted using the organic carbon


partition coefficient (Koc) and effects concentrations in water; (5) the FCV concentration is an


appropriate effects concentration for freely-dissolved chemical in interstitial water; and (6) the


SQC (pg/goc) derived as the product of the KOC and FCV is protective of benthic organisms.


SQC concentrations presented in this document are expressed as fig chemical/g sediment organic


carbon and not on an interstitial water basis because: (1) pore  water is difficult to adequately


sample; and (2) significant amounts of the dissolved chemical may be associated with dissolved


organic carbon; thus, total chemical concentrations in interstitial water may overestimate


exposure.


        The data  that  support the EqP approach  for deriving SQC for non-ionic organic


chemicals are reviewed by Di Toro et al. (1991) and in the SQC guidelines (U.S. EPA, 1993a).


Data supporting these observations for phenanthrene are presented in this document.


       SQC generated using the EqP method are suitable for use in providing guidance to


regulatory agencies because  they are:


       1.  numerical values;                                                          ,


       2.  chemical specific;


       3.  applicable to most sediments;


       4.  predictive of biological effects; and


       5.  protective of benthic organisms.
                                          1-3

-------
 As is the case with WQC, the SQC reflect the use of available scientific data to: (1) assess the

                          '•
 likelihood of significant environmental effects to benthic organisms from chemicals in sediments;


 and (2) to derive regulatory requirements which will protect against these effects.


        It should be emphasized that these criteria are intended to protect benthic organisms from


 the effects of chemicals associated with sediments.  SQC are intended to apply to sediments


.permanently inundated-with water, intertidal sediment and to sediments inundated periodically


 for durations sufficient to permit development of benthic assemblages.  They do not apply to


 occasionally inundated soils containing terrestrial organisms. These criteria do not address the


 question of possible contamination of upper trophic level organisms or the synergistic, additive


 or antagonistic effects of multiple chemicals.  SQC addressing these issues may result in values
                                   X

 lower or higher than those presented in this document.  The SQC presented in this document


represent the U.S. EPA's best recommendation at this time of the concentration of a chemical


in sediment that will not adversely affect most benthic organisms. SQC values may be adjusted


to account for future data or site specific considerations.


       SQC values may also need to be adjusted because of site specific considerations. In spill


situations, where chemical equilibrium between water and sediments has not yet been reached,


sediment chemical concentrations less than SQC may pose risks to benthic organisms. This is


because for spills, disequilibrium concentrations in interstitial and overlying water may ,be
  *                            •

proportionally higher relative to sediment concentrations. Research has shown that the source


or "quality"  of TOC -in the  sediment does not effect chemical binding (De Witt et al., 1992).


However, the physical form of the chemical in the sediment may have an effect.  At some sites


concentrations in excess of the SQC may not pose risks to benthic organisms, because the
                                         1-4

-------
 compound may be a component of a paniculate, such as coal or soot, or exceed solubility such
                         /

 as undissolved oil. In these situations, the national SQC would be overly protective of benthic



 organisms and should not be used unless  modified  using the procedures  outlined  in  the


 "Guidelines for Deriving Site-specific Sediment Quality Criteria for the Protection of Benthic



 Organisms" (U.S. EPA, 1993b). The SQC may be underprotective where the toxicity of other



 chemicals are additive with the SQC chemical or species of unusual sensitivity occur at the site.



       This document presents the theoretical basis and the supporting data relevant to  the



 derivation of the SQC for phenanthrene.  An understanding of the  "Guidelines for Deriving



 Numerical National Water Quality Criteria for the Protection of Aquatic Organisms and Their



 Uses" (Stephan et al., 1985), response to public comment (U.S. EPA, 1985) and "Technical



 Basis for Deriving Sediment Quality Criteria for Nonionic  Organic  Contaminants for the



 Protection of Benthic Organisms by Using Equilibrium Partitioning " (U.S. EPA, 1993a)  is



 necessary in order to understand the following text, tables and calculations. Guidance into the



 acceptable use of SQC values is contained in "Guide for the Use and Application of Sediment


 Quality Criteria for Nonionic Organic Chemicals"  (U.S. EPA, 1993c).






 1.2 GENERAL INFORMATION:  PHENANTHRENE



    Phenanthrene is a member of the polycyclic aromatic hydrocarbon (PAH) group of organic



compounds. Phenanthrene is produced by fractional distillation of high-boiling coal-tar oil and



the subsequent purification of the crystalline solid (Hawley, 1981). Some uses of phenanthrene


are in  the manufacturing  of dyestuffs and  explosives,  in the synthesis of drugs and  in



biochemical research (Verschueren, 1983).  Some PAHs are of environmental concern because
                                        1-5

-------
they are known to be carcinogens and/or mutagens (Brookes, 1977).  With an increase in fossil
                         s                                     ,
                                                                    \
fuel consumption in the United States an increase in emissions of PAHs to the environment can


be expected over the next several decades (Eadie et al., 1982).


    Phenanthrene has a three ring structure and exists as colorless leaflets (Figure 1-1).  It has


a solubility in water at 25 °C of 1.18 mg/L and is a solid at room temperature (melting point of


10p.85°C) (Miller et al., 1985).  .Phenanthrene has a reported vapor pressure of 69.3 - 110.6


mPa at 25 °C (Bidleman, 1984).  Two significant processes which can influence the  fate of


phenanthrene in the sediment are sorption and biodegradation (U.S.  EPA, 1980). Sorption of


phenathrene onto solids in the water column and subsequent settling, as well as partitioning onto


organics in the sediment, can significantly affect phenanthrene transport. Bioaccumulated PAHs
                  _
with 4 rings or less are rapidly metabolized. Therefore,  long-term partitioning into biota is not


considered a significant fate process (U.S. EPA,  1980).  Other processes found to have little or


no effect on the fate of phenanthrene in the sediment are oxidation, hydrolysis and volatilization


(U.S. EPA, 1980).


     The acute toxicity  of phenanthrene ranges from 96 to  >  1150 ug/L for freshwater and


21.9 to 600 pg/L for saltwater organisms (Appendix A). Differences between phenanthrene


concentrations causing acute lethality and chronic toxicity in invertebrates are small; acute-


chronic ratios range from 1.2 to 3.3 for two species.  The only available acute-chronic ratio for


a fish, rainbow trout, is 59 (Table 3-3). Although phenanthrene bioaccumulates in aquatic biota,

                                           •>                     •  !           •
the associated health or ecological risks are unknown.




1.3  OVERVIEW OF DOCUMENT:
                                          1-6

-------
             MOLECULAR FORMULA
             MOLECULAR WEIGHT
             DENSITY
             MELTING POINT
             PHYSICAL FORM
             VAPOR PRESSURE
178.22
1.179 g/cc (25°C)
100.85°C
Colorless leaflets
69.3 - 110.6 mPa (25°C)
             CAS NUMBER:
             CHEMICAL NAME:
85-01-8
Phenanthrene
FIGURE 1-1.  Chemical structure and physical-chemical properties of phenanthrene.
                               1-7

-------
 1.3 OVERVIEW OF DOCUMENT:
                         t
                          :                                        '
       Section 1 provides a brief review of the EqP methodology, and a summary of the

 physical-chemical properties and aquatic toxicity of phenanthrene. Section 2 reviews a variety
                                                                  I

 of methods and data useful in deriving partition coefficients for phenanthrene and includes the

 KOC recommended for use in the derivation of the phenanthrene SQC. Section 3 reviews aquatic
                                                                  I
 toxicity data contained in the phenanthrene WQC document (U.S. EPA, 1980) and new data that

 were used to  derive the FCV  used in this document to derive the SQC concentration. In
'i
 addition, the comparative sensitivity of benthic and water column species is examined as the

 justification for the use of the FCV for phenanthrene in the derivation of the SQC.  Section 4

 reviews data  on the toxicity of phenanthrene  in sediments,  the need for organic carbon

 normalization  of phenanthrene sediment concentrations and the accuracy of the EqP prediction

 of sediment toxicity using KQC and an effect concentration in water. Data from Sections 2, 3

 and 4 are used in Section 5 as the basis  for the derivation of the SQC for phenanthrene and its

 uncertainty. The SQC for phenanthrene  is then compared to STORET (U.S. EPA, 1989b) and

 National Status and Trends (NOAA, 1991) data on phenanthrene's environmental occurrence in

 sediments.  Section 6 concludes with the criteria statement for phenanthrene.  The references

 used in this document are listed in Section 7.
                                        1-8

-------
                                    SECTION 2

                                 PARTITIONING

2.1 DESCRIPTION OF THE EQUILIBRIUM PARTITIONING METHODOLOGY:

      Sediment quality criteria (SQC) are the numerical concentrations of individual chemicals

which are intended to be predictive of biological effects, protective of the presence of benthic

organisms and applicable to the range of natural sediments from lakes,  streams, estuaries and

near coastal marine waters. As a consequence, they can be used in much the same way as water

quality criteria (WQC); ie., the concentration of a chemical which is protective of the intended

use such as aquatic life protection.  For non-ionic organic chemicals, SQC are expressed as ng

chemical/g organic carbon and apply to sediments having ^ 0.2% organic carbon by dry

weight.  A brief overview  follows of the concepts which underlie the equilibrium partitioning

(EqP) methodology for deriving SQC.  The methodology is discussed in detail in the "Technical

Basis for Deriving Numerical Sediment Quality Criteria for Non-ionic Organic Contaminants for

the protection of Benthic Organisms by Using Equilibrium Partitioning" (U.S. EPA, 1993a),

hereafter referred to as the SQC Technical Basis Document.

      Bioavailability of a chemical at a particular sediment concentration often differs from one
                                                                               J
sediment type to another. Therefore, a method is necessary for determining a SQC based on the

bioavailable  chemical fraction in a  sedimenj:.   For non-ionic organic chemicals,  the

concentration-response relationship for the biological effect of concern can most often be

correlated with the interstitial water (i.e., pore water) concentration (jug chemical/liter pore

water) and not to the sediment chemical concentration (jig chemical/g sediment) (Di Toro et al.,


                                        2-1

-------
 1991).  From a purely practical point of view, this correlation suggests that if it were possible
                         *:
 to measure the pore water chemical  concentration, or predict it from the total sediment

 concentration and the relevant sediment properties, then that concentration could be used to

quantify the exposure concentration for an organism.  Thus, knowledge of the partitioning of

chemicals between' the solid and liquid phases  in a sediment is a necessary component for

establishing SQC.  It .is for this reason that the methodology described below is called the

equilibrium partitioning (EqP) method.

       It is shown in the SQC Technical Basis Document (U.S. EPA, 1993a) that the final acute

values (FAVs) in the WQC documents  are appropriate for benthic species for a wide range of

chemicals.  (The data showing this for phenanthrene are presented in Section 3). Thus, a SQC

can be established using .the final,chronic value (FCV) derived using the  WQC Guidelines

(Stephan et al., 1985) as the acceptable effect concentration in pore or overlying water (see
                                                         *-    •     '   '
Section 5),  and the partition coefficient can be used to relate the pore water concentration to the

sediment concentration via the partitioning equation. This acceptable concentration in sediment

is the SQC.

       The calculation is as follows:  Let FCV (/ug/L) be the acceptable concentration in water

for the chemical  of interest; then compute the SQC  using the  partition coefficient, (Kp)

Ck/Kg»edimcBi)> between sediment and water:                                           ,
  *                             .                                             ,
             SQC = KpFCV                                                 (2-1)
                                          -»
This is the fundamental equation used to generate the SQC.  Its utility  depends upon the

existence of a methodology for quantifying the partition  coefficient, Kp.

       Organic carbon appears to be the dominant sorption phase for nonionic organic chemicals
                                         2-2

-------
in naturally occurring sediments and thus controls the bioavailability of these compounds in

sediments.  Evidence for this can be found in numerous toxicity tests, bioaccumulation studies

and chemical analyses of pore water and sediments (Di Toro et al., 1991).  The evidence for

phenanthrene is discussed in this section and section 4.  The organic carbon binding of a

chemical in sediment is a function of that chemical's organic carbon partition coefficient (Koc)

and the weight fraction of organic carbon in the sediment (foe). The relationship is as follows:
             KP  =  foe KOC                                                  (2-2)

      It follows that:


             SQCoc = KocFCV                                              (2-3)

          '•>,''•"


where SQCoc is the sediment quality criterion on a sediment organic carbon basis.

      KOC is not usually measured directly (although it can be done, see section 2.3).


Fortunately, KOC is closely related to the octanol-water partition coefficient (KoW) (equation 2-5)


which has been measured for many compounds, and can be measured very accurately. The next

section reviews the available information on the KoWfor phenanthrene.




2.2 DETERMINATION OF KOW FOR PHENANTHRENE:


    Several approaches have been used to determine KQW for the derivation of  SQC,  as
                                         T
discussed in the SQC Technical Basis Document.  At the U.S. EPA, Environmental Research

Laboratory at Athens, GA (ERL,A) two methods were selected for measurement and two for


estimation of KOW.  The measurement methods were shake-centrifugation (SC), generator column



                                        2-3

-------
 (GCol) and the estimation methods were SPARC (SPARC Performs Automated Reasoning in

 Chemistry; Karickhoff et al.,  1989) and CLOGP (Chou and Jurs, 1979).  Data were also

 extracted from the literature.  The SC method is a standard procedure in the Organization for

-Economic Cooperation and Development (OECD) guidelines for testing chemicals, therefore it

 has regulatory precedence.

 . .  In the examination  of the literature data primary references were found listing measured

 logioKoW values for phenanthrene ranging from 4.28 to 4.63 (Table 2-1).  Primary references

 were found in the literature for estimated log10KoW values ranging from 4.44 to 4.64 (Table 2-1).

 Although the range of reported values for phenanthrene is significantly lower than the range of

 values for some other compounds, it is relatively large, and we were not able to determine from

 studying the primary articles that any value was more likely to be accurate than any other.

  TABLE 2-1.  PHENANTHRENE MEASURED AND ESTIMATED IX)G10KOW VALUES.
METHOD
Measured
Measured
Measured
Measured
Measured
Estimated
Estimated
Estimated ,
Estimated
Estimated
Estimated
LOG10KOW
4.28
4.46
4.56
4,57
4.63
4.44
4.45
4.49
4.58
4.63
4.64 ->
REFERENCE
Haky and Young, 1984
Hansch and Fujita, 1964
DeBruijn etal., 1989
Karickhoff etal., 1979
Bruggemanetal., 1989
Kamletetal., 1988
Mabeyetal., 1982
CLOGP1
SPARC"
Mackay etal., 1980
Yalkowsky et al., 1983
aCLOGP is an algorithm that is included in the database QSAR located at the U.S. EPA,
Environmental Research Lab., Duluth, MN (Chou and Jure, 1979).
bSPARC is from SPARC Performs Automated Reasoning in Chemistry,  (Karickhoff et al.,
1989).
                                        2-4

-------
   KQW values for SPARC and CLOGP are also included in Table 2-1. SPARC is a computer
                        j                                    •


 expert system under development at ERL,A, and the University of Georgia, at Athens. The



 CLOGP algorithm is included in the database QSAR located at EPA's Environmental Research



 Laboratory (ERL,D) at Duluth, Minnesota. For more information on SPARC and CLOGP see



 U.S. EPA (1993a).  The SPARC estimated log10K
-------
  TABLE 2-2.  SUMMARY OF  LOG10KoW VALUES FOR PHENANTHRENE MEASURED
    BY THE U.S. EPA, ENVIRONMENTAL RESEARCH LABORATORY, ATHENS, GA.
SHAKE-
CENTRIFUGATION
4.29
4.25
. " 4.33
4.33
4,30*
GENERATOR
COLUMN
4.47
4.41
4.46
4.24
4.40*
SLOW STIR
FLASK
4.57
4.53
4.50
4.54'
•t            ALogio of mean of measured values.



 2.3  DERIVATION OF KQC FROM ADSORPTION STUDIES:

       Several types of experimental measurement of the KQC are available. The first type

 involves experiments which were designed to  measure the partition coefficient  in particle

 suspensions.  The second  type of measurement is from sediment toxicity tests in which
                        i                              '
 measurements of sediment phenanthrene, sediment organic carbon  (OC), and non-dissolved

 organic carbon (DOC) associated phenanthrene dissolved in pore were used to compute KQC.
 2.3.1 KOC FROM PARTICLE SUSPENSION STUDIES:


      Laboratory studies to characterize adsorption are generally conducted using particle


 suspensions. The high concentrations of solids and turbulent conditions necessary to keep the


 mixture in suspension make data interpretation difficult as a result of a particle interaction effect.


 This effect suppresses the partition coefficient relative to that observed for undisturbed sediments

 (Di Toro,  1985; Mackay and Powers, 1987).


      Based on analysis of an extensive body of experimental data for a wide range of
                                      2-6

-------
 compound types and experimental conditions, the particle interaction model (Di Toro, 1985)
                         ;

 yields the following relationship for estimating KP:




                          foe KOC
 -  ',     •           KP = _ _ _

                                                                              (2-4)
                    1 +


where m is the particle concentration in the suspension (kg/L), and i^ =  1.4, an empirical
constant.                                                                         .


In this expression the KOC is given by:

             logioKoc = 0.00028 +  0.983 log10Kow                                (2-5)

       A sorption isotherm experiment that demonstrates the effect of particle suspensions was


found in a comprehensive literature search for partitioning information for phenanthrene (Table

2-2) (Magee et al.,  1991).  The experiment showed  an observed KP of 12.9 L/kg for a


phenanthrene solution and  sand with 0.11% organic carbon content. Calculated KP using K^

(Equation 2-5) and f^ is 21 L/kg.  The difference between the observed and calculated KP can


be explained by particle interaction effects.   Particle interaction results in  a lower observed

partition coefficient.  The particle interaction model (Equation 2-4) predicts KP of 8.29 L/kg,

which is in agreement with the observed KP.   LogIOBCoc computed from observed KP and f^ is

4.07.  This value is lower than KQC from laboratory measurements due to particle interaction

effects. This data is presented as an example of->particle interaction effects only, as 100 percent

reversibility is assumed in the absence of a  desorption study and an actual KQC can not be

computed.


       In the absence of particle effects, KQC is related to KQW via Equation 2-5. For logio

= 4.36 (ERL,A, mean measured value), this expression results in an estimate of


                                         2-7

-------
 4.29.
          TABLE 2-3.  SUMMARY OF KOC VALUES FOR PHENANTHRENE
            DERIVED FROM LITERATURE SORPTION ISOTHERM DATA.
             Observed           n           Solids
                                               (g/L)      References
             4.07               1              100       Magee et al., 1991



2.3.2 KOC FROM SEDIMENT TOXICTTY TESTS:

      Measurements of KOC aiQ available from sediment toxicity tests using phenanthrene
            *    *              -          "                         !    '    .
(Swartz, 1991). These tests are from three marine sediments having a range of organic carbon

contents of 0.82 to 3.6 percent (Table 4-1; Appendix B).  Phenanthrene concentrations were

measured in the sediment and pore waters providing the data necessary to calculate the partition

coefficient for an undisturbed bedded sediment.

      The 'upper panel  or Figure 2-1 is a plot of the organic carbon-normalized sorption

isotherm for phenanthrene, where the sediment phenanthrene concentration Otg/g
-------
      100000
                           PHENANTHRENE
 I-   10000
            =	1—I I I UNI	Illl Mill	Illl Mill	1 J<> I I III


            I  	r QC •= KOC <= d

                  (Log  K QC " 4-46)
   t>
   O
«
       1000
        100
CO
              ,___LEGEND_	

              - Swortz. 199?
         10 I	iiii inn	tin inn	1  i t i M...    ii.,....
           1            10          100         1000        10000


              PORE WATER CONCENTRATION (ug/L)
        e.0
 O
 O
So
So
O OB
is
        &o
iniiiii  i 11 mm
                             T—T-T
             T-I—r
        SJO I I  I Illllll—I  i nnm  Iiii
          0.1
                                 mini i i—inn 1111
                                                   J_J—Illllll 11
10  20     60     80  00


   PROBABILITY
                                                      09   9M
 Figure 2-1.   Organic carbon-normalized  sorption isotherm for  phenanthrene (top) and
            probability plot of KQC (bottom) from sediment toxicity tests conducted by Swartz
            (1991). The line in the top panel represents the relationship predicted with a log
                  .46, that is Cs>oc = KQC • CD.
                                    2-9

-------
of 4.54 using Equation 2-5.
                         '••                                               ,
2.4 SUMMARY OF DERIVATION OF KOC FOR PHENANTHRENE:      .

      •The KOC selected to calculate the sediment quality criteria for phenanthrene is based on

the regression of log10Koc to log10Kow (Equation 2-5), using the phenanthrene log10KoW of 4.54

recently measured by ERL, A. This approach rather than the use of the KOC from the toxicity

test was adopted because the regression equation is based on the most robust data set available

that spans a broad range of chemicals and particle types, thus encompassing a wide range of KOW

and foe. The regression equation yields a log10Koc of 4.46. This value is in agreement with the

            .33 measured in the sediment toxicity tests.
                                        2-10

-------
                                   SECTION 3

               TOXICITY OF PHENANTHRENE: WATER EXPOSURES

3.1 TOXICITY OF PHENANTHRENE IN WATER: DERIVATION OF PHENANTHRENE
      WATER QUALITY CRITERIA:

      The equilibrium partitioning (EqP) method for derivation of sediment quality criteria

(SQC) uses the phenanthrene water quality criterion (WQC) Final Chronic Value (FCV) and

partition  coefficients  (Koc)  to estimate the maximum  concentrations of nonionic  organic

chemicals in sediments, expressed on an organic carbon basis, that will not cause adverse effects

to benthic organisms.  For this document, life stages of species classed as benthic are either

species.that live in the sediment (infauna) or on the sediment surface (epibenthic) and obtain

their food from either the sediment or water column (U.S. EPA, 1989c).  In this section (1) the

FCV from the phenanthrene WQC document (U.S. EPA, 1980) is revised using new aquatic

toxicity test data; and  (2) the use of this FCV is justified as the effects concentration for SQC

derivation.

3.2 ACUTE TOXICITY - WATER EXPOSURES:

      Fourteen standard acute toxicity tests with phenanthrene have been conducted on 9

freshwater species from 8 genera (Appendix A^  Overall genus mean acute values (GMAVs)

range from 96 to > 1,150 /tg/L.  The acute values for all species tested, except for fathead

minnows, differed by only a factor of 5; 96 to 490 /tg/L. Three tests on three benthic species

from three genera are contained in this database (Figure 3-1; Appendix A).  Benthic organisms

were similar to water column species in sensitivity to phenanthrene; GMAVs range from 126
                                       3-1

-------
   10000
                 A Arthropods

                 D Other Invertebrates

                 O Fishes
 §1000
 HI


 1
 UJ
z
I
§
ul   100
   .       :
       10
                                                                      Pimephales *(J)
                                                       Oncorhynchus (J)


                                               Lepomis (J)
                                      Gammarus (X)
                    Hydra (X)
                             Daphnia (X)
                       20
                                    40
60
                                                                80
                                                                              100
                     PERCENTAGE RANK OF FRESHWATER GENERA
Figure 3-1.
             Genus mean acute values from water-only acute toxicity tests using freshwater

             species vs. percentage rank of their sensitivity. Symbols representing benthic
             species are solid, those representing water column species are open.  Asterisks

             indicate greater than values.  J = juvenile, L  = larvae, X = unspecified life
             stage.
                                         3-2

-------
to 490 /xg/L.  One epibenthic species was tested, the amphipod, Gammanis pseudolimnaeus
                        !                          ,    '
                                                             -   '       ,

(LC50 = 126 ftg/L).  Infaunal species tested included the annelid, Lumbriculus variegatus



(LC50 = >419 Atg/L) and the midge, Chironomus tentans (LC50 = 490 jtg/L).  The Final



-Acute Value (FAV) derived from the overall GMAVs (Stephan et al.f 1985) for freshwater



organisms is 59.63 figfL (Table 3-2).



      Fourteen acute tests have been conducted on 11 saltwater species from 11 genera



(Appendix A).  Overall (GMAVs) range'from 21.9 to 600 /xg/L, similar to the range for



freshwater genera. Fish and crustaceans were the most sensitive. Within this database there are



results from thirteen tests on benthic life-stages of nine species from nine genera (Figure 3-2;



Appendix A).  Benthic organisms were among both the most sensitive,  and most resistant,



saltwater genera to phenanthrene. The most sensitive benthic species is the mysid, Mvsidopsis



bahia. with an average flow-through 96 hour LC50 of 21.9 ftg/L based on two tests with



measured concentrations.  Other benthic species for which there are data appear less sensitive;



GMAVs range from 145 to 600 /*g/L. The FAV derived from the overall GMAVs  (Stephan et



al, 1985) for saltwater organisms is 16.61 /tg/L (Table 3-2).



3.3 CHRONIC TOXICITY - WATER EXPOSURES:



      Chronic  toxicity tests  have been  conducted with  phenanthrene  using a  freshwater



cladoceran (Daphnia magma1) and rainbow troufo(Oncorhynchus mykiss) and a saltwater mysid



fMysidopsis bahia'),  (Table 3-1).  The  D. magma and Q.  myMss were  tested in life-cycle



exposures.  Q. mykiss embryos, sac fry and swim-up benthic (intergravel) stages were tested



in an early life-stage toxicity test.



      Call et al. (1986) conducted both freshwater tests. 2. magma exposed 21 days to a mean
                                         3-3

-------
    1000
a
HI
3
B
<
1
I
UJ
O
     100
       10
A Arthropods
D Other Invertebrates
O Fishes
                                                                         Neanthes (A)
                                                                 Cyprinodon (J)
                                                           Mya *(A)
                                                 Mytilus '(A)     |
                                  (Leptocheirus (A)
                           *D//70p/7//US(J)
                     "Papums(A)
              * Palaemonetes (A)
                  . Mysidopsis (J)
                        20            40            60            80

                      PERCENTAGE RANK OF SALTWATER GENERA
                                                                 100
  Figure 3-2.   Genus mean acute values from water-only acute toxicity tests using saltwater
                species vs. percentage rank of their sensitivity.  Symbols representing benthic
                species are solid, those representing water column species are open.  Asterisks
                indicate greater than values. A = adult, J = juvenile.
                                             3-4

-------
  U
  H
  hi


  g
  Oi
  co
co
H
O

a,

g
TBR
S
CO

o
           a
           S

           s
        C AJ
        o u
        0) o
        O HI

        Cb H
                     ^J    JJ
                     «    «

                     I-IVD  i-HO
                     iH 00 ; rl 00

                     S3  S3
   a

   OK
   a eg
          y
         M
         0.
          ii
         . o
         B 0
         Q) *W

         ««
        Is
                    «a  os
                    OBM  ....
                    at eo  f> »
66 JB

**
00
      -1
       «S
       4j m
                 i!
                 a
                 03
                    in
                                   em
EI  U zco z
                    si
                                  5.!
                                  I'd
                                                    5  |
                                                    §  "S
         I
         u
         a
         o
                                                     u
                  u
                  •
                                                     «  "
 i  !?
    •H
    •S
                                                 •  £
                                                 <
                                               U

                                                 >  **
                                            5*  • 2  "o

                                           3  *•
            ^ -

        -  *^
        -H  *J "3


        !li
        01  "S*
^  ?
h  -H
u  At
                                           g   ;
                                                '  S  S
                                                ,-  5  8


                                           «   e
                                                       I
                                                   '  a
                                                   o  o<
                                                   0
                                                   H
                                          3-5

-------
phenanthrene concentration of 163 /tg/L experienced 98% reduction in reproduction and 83%
                          t                                           i
                          t                                                        -

reduction in survival relative to controls (Table 3-1). There was no statistically significant effect



on survival or reproduction of daphnids in phenanthrene concentrations from 46 to 57 /tg/L.



O. mykiss exposed to phenanthrene for 90 days in an early life-stage toxicity test were not



affected in 5 /xg/L.  Duration of incubation  and hatching success were not affected in any



treatment.  However, the percentage of abnormal and dead fry at  hatch was significantly



increased at the highest exposure (66 /tg/L). Sac fry were underdeveloped from hatching until



test termination and swim-up delayed in .>. 14 pg/L.  At test termination, wet weights and



standard lengths were reduced in _>. 32 pg/L.  Survival was reduced in _> 8 pg/L.



       M. bahia exposed to phenanthrene in a  life-cycle toxicity test (Kuhn and Lussier, 1987)



were affected at phenanthrene concentrations similar to those affecting the Q. mykiss (Table 3-



1).  Survival, growth and  reproduction were not affected in _<. 5.5 /ig/L.   At the highest



concentration of phenanthrene  (11.9 /*g/L) all  mysids died.



       Derivation of the FCV for phenanthrene is complicated because Acute-Chronic Ratios



(ACR) differ in the three species tested by a factor of almost 50 (Table 3-2).  The final ACR,



therefore, can not be the mean of these three values (Stephan et al.,  1985).   The difference



between concentrations of phenanthrene acutely and chronically toxic to invertebrates is small.



ACRs are 1.214 for the freshwater (D. magnal  and 3.333 for the saltwater M- bahia. mean ratio



of 2.012.  The ACR of 59.29 for Q. mykiss (Call et al., 1986) probably should not be used to



derive  the final ACR or chronic values for untested fishes because (1)  it is over 10 x the ratio



for tested invertebrates,  (2) the Q. mykiss 96 hr LC50 of 375 jtg/L would be 50 /*g/L if based



on immobilization (Call et al., 1986), thus the ACR would be 7.905 and (3) the chronic value
                                          3-6

-------
u.
1
9

H
         O
        •H
         0)  O
         -U-H
              *
              3
              I
           H)


           ^
           o
                             M
                                  Ok
                                  0\
                                  in
in
                         to
                         01
                                                CS
                         1-1
                         H
                                     in
                                     r»
                                     ro
              r-
              M
                                                                       I VD
                                                              1
                                                              O
                                                                                     14

                                                                                     O
                                                                                     col
                                                       3-7

-------
  may be conservative based on tests with other fish species. In non-standard chronic exposures,
                          *                                          i
                           r
  sensitivities of early life-stages of largemouth bass (Micropterus salmoides) and Q. mykiss

  (Black et al., 1983; Milleman et al., 1984) were less than observed by Call et al. (1986).  These

 jchronic exposures lasted from fertilization to four days after hatching, about 7 days for bass and

 27 days for trout- "Hatching and survival of Q. mykiss were reduced in 38 /tg/L but not in 31

 /ig/L;  in contrast to the effect concentration of 8 /tg/L was observed by Call et al. (1986).  The

 LC50  for these tests was 40 /tg/L for CT. mykiss and 180 /tg/L for bass (Black et al., 1983;
"t
 Milleman et al., 1984).  Because the most acutely  sensitive species to phenanthrene were

 invertebrates, the FAV, 59.63 /tg/L for freshwater and 16.61 /tg/L for saltwater, was divided

 by the invertebrate  mean ACR of 2.012 to derive an initial estimate of the FCV. These initial

 FCVs  were- 29.64 /tg/L for freshwater and 8.255 /tg/L for saltwater aquatic life.  The initial

 freshwater FCV was lowered to 6.325 /tg/L the chronic value from the Q.  mykiss early life-
                  *                                          s
 stage test with intergravel benthic embryonic and sac-fry life stages of this important species.

 The initial saltwater FCV of 8.255 /tg/L was not lowered because the chronic sensitivities of

 saltwater fishes is not known and should not be estimated using the ACR for trout which may

 not be appropriate for other fish species. The initial FCV for saltwater aquatic life is used as

 the FCV because it is 13 to 52 times  lower than acute values for tested saltwater fishes  and

 approximately equal to the chronic value of 8?129  /tg/L for the M-  bahia.   Although  this

 procedure to derive the FCV is complicated and does not follow exactly the WQC Guidelines

 (Stephan et al.,1985) for idealized databases,  the procedure is consistent with the guidelines

 requirement that the criterion be consistent with sound scientific evidence.
                                           3-8

-------
 3.4 APPLICABILITY OF THE WATER QUALITY CRITERION AS THE EFFECTS
    CONCENTRATION FOR DERIVATION OF THE PHENANTHRENE SEDIMENT   .
    QUALITY CRITERION:

       The use of the FCV  (the chronic effects-based WQC concentration) as the effects

 concentration  for calculation  of the EqP-based  SQC assumes  that benthic (infaunal and

 epibenthic) species, taken as a  group, have sensitivities similar to all benthic and water column

 species tested  to derive the WQC concentration.  Data supporting the reasonableness of this

 assumption over all chemicals for which there are published or draft WQC documents are

 presented in Di Toro et al. (1991), and the SQC Technical Basis Document (U.S. EPA, 1993a).

 The conclusion of similarity of sensitivity is supported by comparisons between (1) acute values

 for the most sensitive benthic and acute values for the most sensitive water column species for

 all chemicals;  (2) acute values for all benthic species and acute values for all species in the

 WQC documents across all chemicals after standardizing the LC50 values; (3) FAVs calculated

 for benthic species alone and FAVs calculated for all species in the WQC documents; and (4)

 individual chemical comparisons of benthic species vs. all species. Only in this last comparison

 are phenanthrene-specific comparisons in sensitivity of benthic  and all (benthic and water-

 column) species conducted.  The following paragraphs examine the  data on the similarity of

 sensitivity of benthic and all species for phenanthrene.

       For phenanthrene, benthic species account for 3 out of 8 genera tested in freshwater, and

 10 out of 11 genera tested in  saltwater (Figures 3-1, 3-2).  An initial test of the  difference

between the freshwater and saltwater FAVs for all species (water column and benthic) exposed

to phenanthrene was performed using the Approximate Randomization method (Noreen, 1989).

The Approximate Randomization method tests the significance level of a test statistic when
                                        3-9

-------
             TABLE 3-3. RESULTS OF APPROXIMATE RANDOMIZATION
             TEST FOR THE EQUALITY OF THE FRESHWATER AJND
             SALTWATER LC50 DISTRIBUTIONS FOR PHENANTHRENE
             AND APPROXIMATE RANDOMIZATION TEST FOR THE
             EQUALITY OF BENTfflC AND COMBINED BENTHIC
             AND WATER COLUMN (WQC) LC50 DISTRIBUTIONS.

            Compar-
            ison"    Habitat or Water Type*  AR Statistic1"    Probability6

            Fresh    Fresh (8)   Salt (11)      43.03          72
            vs Salt   •     '

            Benthic   Benthic (13) WQC (19)    7.35           80
            vs Water
            Column +
            Benthic (WQC)
            •Values in parentheses are the number of LC50 values used in the comparison.
            bAR statistic = FAY difference between original compared groups.
            "Probability that the theoretical AR statistic ^ the observed AR statistic given
             that the samples came from the same population.
compared to a distribution of statistics generated from many random subsamples.  The test

statistic in this case is the difference between the freshwater FAV, computed from the freshwater

(combined water column and benthic) species LC50 values, and the saltwater FAY, computed

from the saltwater (combined water column and benthic) species 1.C50 values (Table 3-1).  In

the Approximate Randomization method, the freshwater LC50 values and the saltwater LCSO
  *                         . *                                 !
values are combined into one data set.  The data set is shuffled, then separated back  so that

randomly generated "freshwater" and "saltwater" FAVs can be computed.  The LC50 values

are separated back such that the number of LCSO values used to calculate the sample FAVs are

the same as the number used to calculate the original FAVs.  These two FAVs are subtracted
                                     3-10

-------
 and the difference used as the sample statistic.  This is done many times so that the sample
                          ;           -.?


 statistics make up a distribution that is representative of the population of FAV differences



 (Figure 3-3).  The test  statistic is compared to this distribution to determine it's level of



 significance.  The null hypothesis is that the LC50 values that comprise the saltwater and



 freshwater data bases  are not different.  If this is  true,  the difference between the actual



 freshwater and saltwater FAVs should be common to the majority of randomly generated FAV



 differences.  For phenanthrene, the test-statistic falls at the 73 percentile of the generated FAV



 differences.  Since the probability is less than 95%, the hypothesis of no significant difference



 in sensitivity for freshwater and saltwater species is accepted (Table 3-3).



       Since freshwater and saltwater species showed similar sensitivity, a test of difference in



 sensitivity for benthic and all (benthic and water column species combined, hereafter referred



 to as "WQC") organisms combining freshwater and saltwater species using the Approximate



 Randomization method was performed. The test statistic in this case is the difference between



 the WQC FAV, computed from the WQC LC^ values, and the benthic FAV, computed from



 the benthic organism LCjo values. This is slightly different then the previous test for saltwater



 and freshwater species. The difference is that saltwater and freshwater species represent two



 separate groups. In this test the benthic organisms are a subset of the WQC organisms set. In



 the Approximate Randomization method for thisltest, the number of data points coinciding with



the number of benthic organisms are selected from the WQC data set.  A "benthic" FAV is



 computed.   The original WQC FAV and the "benthic"  FAV are then used to compute the



 difference statistic.  This is done many times and the distribution that results is representative



 of the population of FAV difference statistics.  The test statistic is compared to this distribution
                                         3-11

-------



0
2!
S3
Lu'cb
S~

.
Ii.



C.VW
100
120
80
An
^v»
0

-40
•80
^Ti^\^
•120

•16O
-200
i 1 1 1 inn i i 1 1 1 mi i i i i i i i mi 1 1 1 i i mini i i
- FRESHWATER VS SALTWATER ^o °~
- '.. rf*9 7
- _^,a^B5^^^ "
j^ssss^^^
o • -

- ^o ' - -
itrfflfflnDOBCfiflTC®*^^^
- O -
-O . _
- ' . -
i i ii mil i i 1 1 inn i i i i i i i HIM M i i mini i i
             0.1
10  20      6O      80   90
89     99.9




lit
o
•MM
5
65
Ii. CO
g3
>
Ii.

.


.
500

100
120
80

4O
0

-40
•80
-12O


-100
-2OO
i 1 1 nun i i 1 1 1 mi i i i i i i i in u M i i . mini i i
- BENTHJC VS WQC I
. —
•—
^
— ' _
	 • u—t.r-L 'j 	

— IUUUUUUUUUUUDUUUUUUUUU.T ^
— . _
^Q
" rssSSsP**^^ ~
~ cffST^ ->
O ~
-o

i i mini i i i HUH i t i i i i i nun i i i mini i i ~











1



0.1 1 10 20 60 80 90 99 99.9
                                  PROBABILITY

Figure 3-3.   Probability distribution of FAV difference statistics to compare water-only data
            from freshwater vs saltwater (upper panel) and benthic vs. WQC (lower panel)
            data.
                                     3-12

-------
to determine its level of significance. The probability distribution  of the computed FAV
                           j
                           •                                                  •  •

differences are shown in the bottom panel of Figure 3-3. The test statistic for this analysis falls



at the 80 percentile and the hypothesis of no difference in sensitivity is accepted (Table 3-3).



This analysis suggests that the FCV for phenanthrene based on data from all tested species is an



appropriate effects  concentration for benthic organisms.
                                           3-13

-------

-------
                                    SECTION 4


           TOXICITY OF PHENANTHRENE (ACTUAL AND PREDICTED):
                              SEDIMENT EXPOSURES



4.1 TOXICITY OF PHENANTHRENE IN SEDIMENTS:


       The toxicity of phenanthrene spiked into sediments has been tested with two saltwater


amphipod species,  Freshwater benthic species have not been tested in phenanthrene-spiked


sediments.  All concentrations of phenanthrene in sediments or interstitial water where effects


were observed in benthic species (Table 4-1) are  greater than SQC or FCV concentrations


reported in this document.  Details about exposure  methodology are provided because, unlike


aquatic  toxicity  tests,  sediment  testing  methodologies  have  not  been  standardized.


Generalizations across species or sediments are limited because of the limited number of


experiments. Therefore, insights into relative sensitivities of aquatic species to phenanthrene can


only be obtained from results of water-only tests (Section 3).  Data are available from many


experiments using both field and laboratory sediments contaminated with mixtures of PAHs and


other compounds which include phenanthrene. Data from these studies have not been included


here because it is not possible to determine the contribution of phenanthrene to the observed


toxicity.
                                                                        /

      Swartz  (1991) exposed  the amphipods Eohaustorius estuarius and  Leptocheirus


plumulosus to three phenanthrene-spiked sediments with  total organic carbon contents  (TOC)


ranging from 0.82 to  3.6%.  Sediments were rolled (1) for two hours in phenanthrene-coated


bottles; (2) stored at 4°C for 72 hours; (3) rolled for an additional two hours, and (4) then
                                        4-1

-------















t
1
g
CO
p
gj
M
Ct
co
i








g
i
s
0

i
S3
a
M
3
9
1







S3
m
S
B
S







d
d
f-i



o
§

o
j£
CO*

en"?-
w en
* •
fj W


g
H
V\ Ol

fl iH
^ ^
M M
4J It
«! a
^ ^
CO CO



VO o
o ^
«
o o
en o
•. «
^ 00





d in
w in
T-l d



e . e
in in

0 0
s s
eo en

o of.
in toot
d CO d


!- * .
en
g •&
i-H 9
CO O
& S
,_j B
1 £



S 9
5) C fflj CD
s — H 9 — 
-------
 stored for 7 days at 4°C. In some of these experiments the concentration of phenanthiene was
                         i             ~.             i

 not sufficient to cause 50%  mortality in  any of the concentrations tested.  In these cases


 additional experiments were performed with sediments from the same locations with similar TOC


 concentrations as were used in the original experiments, but with one or two treatments with


 higher phenanthrene concentrations and the appropriate controls (Table 4-1).  When there was


 a difference between the control mortality in one of the original experiments and in the follow


 up experiment with the corresponding sediment and species, Abbott's correction was performed


 on the data for each treatment separately using the appropriate control mortality. Then the data


 for both experiments were pooled.  The pooling of the data appears justified by the similarity


 of the dose-response relationships in the original and the follow up experiments (Appendix B).


 The 10-day LCSO's for both species increased with increasing organic carbon concentration when


the phenanthrene concentration was  expressed on a dry  weight basis, but decreased when


concentration was expressed on an organic carbon basis.  LCSO's normalized to dry weight


differed by a factor of 3.1 (39.2 to 122 /tg/g) for JL estuarius over a 3.3-fold range of TOC and


a factor of 2.8 (92.4 to 255 ;tg/g) for JU plumulosus over a 1.8-fold range of TOC. The organic


carbon normalized LCSO's for E. estuarius differed by a factor of 1.1 (3,820 to 4,050 /tg/goc)


while for L. plumulosus they differed by a factor of  1.3 (6,490 to 8,200 ftg/gbc)-

                                           ->
       Overall, the  need for organic carbon  normalization of the concentration of non-ionic


organic chemicals in sediments is presented in the SQC Technical Basis Document (U.S. EPA,


 1993a). The need for organic carbon normalization for phenanthrene is also supported by the


results of  spiked-sediment toxicity  tests  described above.   Although it is  important to


demonstrate that organic carbon normalization is necessary if SQC are to be derived using the
                                         4-3

-------
 EqP approach, it is fundamentally more important to demonstrate that KQC and water only effects
                         /                                   -


 concentrations can be used to predict effects concentrations for phenanthrene and other non-ionic



 organic chemicals on an organic carbon basis for a range of sediments.  Evidence supporting this



 prediction for phenanthrene and all other nonionic organic chemicals follows in Section 4.3.






 4.2 CORRELATION BETWEEN ORGANISM RESPONSE AMI) PORE WATER

  CONCENTRATION:



       One corollary of the EqP theory is that pore-water LCSO's for a given organism should



 be constant across sediments of varying organic carbon content (U.S. EPA, 1993a). Appropriate



pore-water LC50 values are available for two benthic species (Table 4-2). Swartz (1991) found



 10-day LC50 values based on pore-water concentrations varied by a factor of 1.1 (138 to 146
           1                                                           •

jig/L)  for IL estuarius and by a factor of 1.3 (306 to 387 /tg/L) for L- plumulosus.  This



variability  is somewhat less  than  that shown when  dry weight (factors of 3.1  and 2.8)



normalization is used to determine LCSOs based on phenanthrene concentration in sediments, but



similar to that shown when organic  carbon (factors of 1.1 and 1.3) normalization is used.
                                                                   !


     A. more detailed evaluation of the degree to which the response of benthic organisms can



be predicted from toxic units  of substances in pore water can be made utilizing results from



toxicity tests with sediments spiked with other substances, including  acenanphthene and

                                                                                i

phenanthrene (Swartz, 1991),  endrin (Nebeker et al., 1989; Schuytema et al.,  1989), dieldrin



(Hoke 1992),  fluoranthene (Swartz et al., 1990, De Witt et al., 1992), or kepone (Adams et al.,



1985) (Figure 4-1; Appendix B). The data included in this analysis come from tests conducted



at EPA laboratories or from tests which utilized designs at least as rigorous as those conducted



at the EPA laboratories.  Tests with acenaphthene and phenanthrene used two saltwater



                                        4-4

-------
e3
8
.P
TIONINO
H
f*
55


S
0
I
e
s
8
u
2
S
1
a
§1

H
15
8
o
M
8
CO
o
in
s.

c
H
i
SB
ILB 4-2: HATER-ONLY
PHBNANTHRB
S3



o
o
S
o
&

•o
o
.."So o
0 3 m-H
"^ O 2 8>,
5 o!

* o O)
3 SI*
M



5)y
: 2 o
P 43 I/I
ig-S AJ
'I'lSr--'.

'
8*
14
2« e
AJ in

£
££
e
t> o
8 _
%I
mon/Sci. Name Math
Duration

5

en
en

M
2
CO



t-
e



eo
t-
r»
rt



0
in
0
*


en

o et
rH C3

2 3


•
^^ *^
Ol rt
e
SB
hipod, .FT
lauBtoriua eatuariua
Qi«C
< M

ot

Swartz,



0



CO
rt*



e
en



r<
en

*
w

en
rt
H



1-1
rt
e
^
hipod, FT
austoriuB eBtuarius
B"5
< «

m

Swartz,



o



eo
t-
rt



e
Ct
eo
rt


M
n
1-1

rt en
• 4
^1 W

«O
iH



fH
rt
O
^
hipod, FT
auBtoriuB eatuarius
&*
Is

ert

Swartz,



in



in
rt
in



o
^3
<0


si

To «
en a
IH e

eo
rt



in
eo
<3
^
hipod, FT
tocheiruB Dlumulosui
fa
3

m

Swartz,



M
M



in
rt
rt
in



.0
ert
10


vo
1-1

o
in
.*"*

o
rt



in
eo
o
^«
e!
i-
m
f-
a
J±
t*"5
.1
A AJ
Q.C
Is

ert

*N
AJ
S
CO



in

'

in
rt
in*



o
e
n
CD


in
in
rt

«o en
m c*

o
rt



in
eo
e
£
hipod, FT
tocheiruB Dlumulosui
a c
II









.
"o
r4
•


i
ii
o
o
o
• S
8 U a 0
_, « 5
S h i AJ
D AJ H Ct
| § 5 4
u B S -H
! ? : '.
• Flow through, M •
tatmant- specific TOC
idicted LC50 (pg/gj
! of Bediment used i
< S S X
* & f1 ^
4-5

-------
  Sr
  O
    W UJ 3
    Q M G-
 I I' . I" I I  I

D < *> O > O
i  t  i  i  i  i  i  i  i  i
                           §
 I   8   8   S   8
                              CO
                                fldrS3
                                fc «o rO.

                                -2)*a §'


                                'S'3' .S &
                                  «^
                                       €
II
•si
                             g
                                ife»s
                                S $ Q «•


                                III |*II^
                                •sj^s
                                  § s
                                    Bl
                                S-s-x-g s

                                ftS Si
                                      •a
                                        §«
                             in
                                 I g
                                 Cw ^"^
                                 C! 'O\

                                 •&S1I
                                     J

                                         •D CO
                               4




                               I


                               I
            4-6
                 %

-------
 amphipods (L. plumulosus and IL estuarius) and marine sediments. Tests with fluoranthene used
                         ;
                         ;                                      •

 the saltwater amphipod rRhepoxynius abronius) and marine sediments.  Freshwater sediments




 spiked with endrin were  tested using the amphipod Hyalella  azteca: while the midge,




JChironomus tentans. was  tested using kepone-spiked sediments.   Figure 4-1 presents  the




percentage mortalities of the benthic species tested in individual treatments for each chemical




versus "pore water toxic units" (FWTUs) for all sediments tested. PWTUs are the concentration




of the chemical in pore water 0*g/L) divided by the water only LC50 Gig/L).  Theoretically,




50% mortality should occur at one interstitial water toxic unit.  At concentrations below one




PWTU.there should be less than 50% mortality, and at concentrations above one PWTU there




should be greater than 50% mortality.  Figure 4-1 shows that at concentrations below one




PWTU mortality  was generally low, and increased sharply at approximately one PWTU.




Therefore, this comparison supports the concept that interstitial water concentrations can be used




to predict the response of an organism to a chemical that is not sediment-specific.  This pore




water normalization was not used to derive SQC in this document because of the complexation




of nonionic organic chemicals with pore water DOC (Section 2) and the difficulties of adequately



sampling pore waters.






4.3  TESTS OF THE EQUILIBRIUM PARTITIONING PREDICTION OF SEDIMENT

       TOXICITY:




       SQC derived using the equilibrium partitioning approach utilize partition coefficients and




FCV  from WQC documents  to derive the SQC  concentration  for protection of benthic




organisms.  The partition coefficient (Koc)  is used to normalize sediment concentrations and




predict biologically available concentrations across  sediment types.  The data required to test the
                                        4-7

-------
 organic carbon normalization for phenanthrene in sediments arc  available  for two benthic
                          i
                           :                                          \

 species. Data from tests with water column species were not included in this analysis.  Testing


 of this component  of SQC derivation requires  three  elements:  (1)  a water-only effect


 concentration, such as  a 10-day LC50  value in /tg/L,  (2)  an  identical  sediment effect


 concentration on an  organic carbon basis, such as a 10-day UC50 value in ^eg/goo and (3) a


.partition coefficient for the chemical, KOC in Lflsgoc-  This section presents evidence that the


 observed effect concentration in sediments (2) can be predicted  utilizing the water effect


 concentration (1) and the partition coefficient (3).


       Predicted 10-day LC50 values from phenanthrene-spiked sediment with £. estuarius and


 L. plumulosus were  calculated (Table 4-2) using the log10Koc value of 4.46 from Section 2 of


 this document and the sediment LCSO's in Swartz (1991).  Ratios of actual to predicted LCSOs


 for phenanthrene averaged 1.05 (range 1.04 to 1.07) for IL estuarius and 1.42 (range 1.22 to


 1.54) for L. plumulosus.  The overall mean for both species was 1.22.


       A more detailed evaluation of the accuracy and precision of the EqP prediction of the

                             •  •          •             .            '   •      '     N
response of benthic organisms can be made using the results of toxicity tests  with amphipods


exposed to sediments spiked with acenaphthene, phenanthrene, dieldrin, endrin, or fluoranthene.


The data included in this analysis come from tests conducted at EP'A laboratories or from tests

                                            ->
which utilized designs at least as rigorous as those conducted at the EPA laboratories. Data from
  4                             .                     •.

the kepone experiments are not included because a measured KQW obtained using the slow-stir


flask method  is not available. Swartz (1991) exposed the saltwater amphipods E.. estuarius and


L. plumulosus to acenaphthene in three marine sediments having organic carbon contents ranging


from  0.82 to 4.2%  and to phenanthrene in three  marine sediments  having  organic carbon
                                         4-8

-------
 contents ranging from 0.82 to 3.6%.  Swartz et al. (1990) exposed the saltwater amphipod |L
                          t.
 abronius to fluoranthene in three marine sediments having 0.18,0.31 and 0.48 % organic carbon.


 Hoke and Ankley (1991) exposed the amphipod IL azteca to three dieldrin-spiked freshwater

 sediments having 1.7, 3.0 and 8.5% organic carbon, and Hoke (1992) exposed the midge C.

 tentans to freshwater dieldrin spiked sediments having 2.0 and 1.5% organic carbon.  Nebeker


 et al. (1989) and Schuytema et al. (1989) exposed H. azteca to three endrin-spiked sediments

 having 3.0,  6.1 and 11.2%  organic carbon.  Figure 4-2 presents the percentage mortalities of


 amphipods in individual treatments of each chemical versus "predicted sediment toxic units" for


 each sediment treatment.  PSTUs are the concentration of the chemical in sediments fttg/goc)


 divided by the predicted LC50 Qtg/goc) in sediments (the product of KQC and the 10-day water-

 only LC50).. ,In this normalization, 50 % mortality should occur at one PSTU. At concentrations


below one PSTU mortality was generally low, and increased  sharply at one PSTU.  The means


of the LC50s for these tests .calculated on a PSTU basis were 1.90 for acenaphthene, 1.16 for

dieldrin, 0.44 for endrin, 0.80 for fluoranthene and 1.22 for phenanthrene. The mean value for

the five chemicals is 0.99. This illustrates that the EqP method can account for the effects of


different sediment properties and properly predict the effects concentration in sediments using

the effects concentration from water only exposures.
                                         4-9

-------
 I I I I  I


o < *- o o
   T	1	T
  T	1	T
                  T	1	T
 O O



 B



 1

 a
  r<*%
  i

 o|

r I
          *! °
               <^t
           °%D°^d
               = 9
                  I-
e

I
               >i^
                ^<^>
MB
I

,.
i i i r i
1 -^
: o DJfe
i
1 O
1 	 1 i i i tPOi
^
-

—
,
                          I
§   8   3
                       §|i

                       f&!

                                       "
                               &•!
«ii!iJSi5

I*f*filM«




I -§ f "8«5 * .9 •§ *
i @/o e g«s ^ g i
a « S'-O * g tS g .
    S «  .2 ^b<

     fill'
                      •*&°.g



                      fill
                       BO «nH vBrf T3
                      •a as " ^
                                 •5 o 5 §8  B
                      -1^1 II
                      § I r*l §1

                 4-10

-------
                                    SECTIONS

                 CRITERIA DERIVATION FOR PHENANTHRENE

5.1 CRITERIA DERIVATION:

       The water quality criteria (WQC) Final Chronic Value (FCV), without an averaging

period or return frequency (See section 3), is used to calculate the sediment quality criteria

(SQC) because it is probable that the concentration of contaminants in sediments are relatively

stable over time, thus exposure to sedentary benthic species should be chronic and relatively

constant.  This is m contrast to the situation in the water column, where a rapid change in

exposure and  exposures of  limited durations  can occur due  to fluctuations  in effluent

concentrations dilutions in receiving waters or the free-swimming or planktonic nature of water

column organisms. For some particular uses of the SQC it may be appropriate to use the area!

extent and vertical stratification of contamination of a sediment at a site in much the same way

that averaging periods or mixing zones are used with WQC.

       The FCV is the value that should protect 95% of the  tested species included in the
                                         ->
calculation of the WQC from chronic effects of the substance. The FCV is the quotient of 'the

Final Acute Value (FAV), and the final Acute Chronic Ratio (ACR) for the substance.  The

FAV is an estimate of the acute LC50 or EC50 concentration of the substance corresponding to

a cumulative probability of 0.05 for the genera from eight or more families for which acceptable

acute tests have been conducted on the substance.  The ACR is the mean ratio  of acute to
                                        5-1

-------
 chronic toxicity for three or more species exposed to the substance 'that meets minimum database
                         ;

 requirements.  For more information on the calculation of ACRs, FAVs, and FCVs see the


 National Water Quality Criteria Guidelines (Stephan et al., 1985).  The FCV used in this


 document differs from the FCV in the phenanthrene WQC document (U.S. EPA, 1980) because


 it incorporates recent data not included in that document, and omits some data which does not


 meet the data requirements established in the WQC Guidelines (Stephan et al., 1985).


        The equilibrium partitioning (EqP)~method for calculating SQC is based on the following

it
 procedure.  If FCV G*g/L) is the  chronic concentration from the WQC for the chemical of


 interest, then the SQC (pg/g sediment), is computed using the partition coefficient, Kp (L/g


 sediment), between sediment and pore water:


            ,  SQC=KPFCV                                                 (5-1)


        Since organic carbon is the predominant sorption phase for nonionic organic chemicals


 in  naturally  occurring sediments,  (salinity,  grainsize and other sediment  parameters have


 inconsequential roles in sorption,  see  sections 2.1 and  4.3) the organic  carbon  partition


 coefficient, (Koc) can be substituted for KP.  Therefore, on a sediment organic carbon basis, the


 SQCoc Gtg/goc), is:


              SQCOC = KOCFCV                                               (5-2)
                                           ->
 Since (Koc) is presumably independent of sediment type for non-ionic organic chemicals, so also


 is SQCoc. Table 5-1 contains the calculation of the phenanthrene SQC.


     ,   The organic carbon normalized SQC is applicable to sediments with an organic carbon


 fraction of foe ==  0.2%.  For sediments with foe <  0.2%, organic carbon normalization and


 SQC may not apply.
                                         5-2

-------
        TABLE 5-1. SEDIMENT QUALITY CRITERIA FOR PHENANTHRENE
Type of
Water Body
Fresh Water
Salt Water

(L/kg)
4.54
4.54
LogioKoc
(L/kg)
4.46
4.46
FCV
0*g/L)
6.32
8.26
SQCoc
Otg/goc)
180*
240b
     •SQCoc  = (104-46 L/kgoc)«(10-3 kgoc/goc)«(6.32 /tg phenanthrene/L)  =  180 /tg
phenanthrene/goc.


     "SQCoc  = (104-46 L/kgoc)»(10-3 kgoc/goc)-(8.26 /tg phenanthrene/L)  =  240 jig
phenanthrene/goc


       Since  organic carbon is the factor controlling the bioavailability of nonionic organic

compounds in sediments, SQC have been developed on an organic carbon basis, not on a dry
           *•••'.'"
weight basis.  When the chemical concentrations in sediments are reported  as  dry weight

concentration and organic carbon data  are available,  it is best to  convert the  sediment

concentration to /tg chemical/gram organic carbon. These concentrations can then be directly

compared to  the  SQC value.   This facilitates comparisons between  the SQC and field

concentrations relative to  identification  of  hot spots and  the degree to which  sediment

concentrations do or do not exceed SQC values.  The conversion from dry weight to organic

carbon normalized concentration can be done using the following formula:
                                                                                i
       /tg Chemical/goc =  fig Chemical/gDRYWT ^- (% TOC -s-  100)

                      = /tg Chemical/gDRYWT  • 100 •*• % TOC

       For example, a freshwater sediment with a concentration of 6.00 /tg chemical/gDRY ^

and 0.5 % TOC has an organic carbon-normalized concentration of 1,200 /tg/goc (6.00 /tg/goRYwr

• 100  -^ 0.5  = 1,200 /tg/goc) which exceeds  the SQC of 180  /tg/goc-  Another freshwater


                                       5-3

-------
 sediment with the  same concentration  of phenanthrene (6.00 Mg/goRY WT)  but a TOG
                         j                                         :


 concentration of 5.0% would have an organic carbon normalized concentration of 120 Mg/goc



 (6.00 ftg/gDRYwr • 100 •*• 5.0 = 120 /tg/goc). which is below the SQC for phenanthrene.



       In situations where TOC values for particular sediments sine not available, a range of



 IDC values may be used in  a "worst case" or "best case"  analysis. In this case, the organic



.carbon-normalized SQC values  (SQCoc) may be "converted" to dry weight-normalized SQC



 values (SQCnRy wr.).  This "conversion" must be done for  each level of TOC of interest:



             SQCDRYWT =  SQCoc(ftg/goc) • (% TOC + 100)



 where  SQCj>RYwr is the dry weight normalized SQC value. For example,  the SQC value for



 freshwater sediments with 1 % organic carbon is 1.8 ftg/g:



           '   SQCDRYWT; =  180 /tg/goc • 1% TOC -5- 100 = 1.8 fig/gom-wr



 This method is used in the analysis of the STORET data in section 5.4.







 5.2 UNCERTAINTY ANALYSIS:



    Some of the uncertainty in the calculation of the phenanthrene SQC can be estimated from



 the degree to which the EqP model, which is  the basis for the criteria, can rationalize the



 available sediment toxicity data.  The EqP model asserts that (1) the bioavailability of nonionic

                                          ->                       :

 organic chemicals from sediments is equal on an organic carbon basis, and  (2) that the effects
  •»                             .                                   •           .


 concentration in sediment G*g/goc) can be estimated from the  product of the effects concentration



 from water only exposures G*g/L) and the partition coefficient KQC (L/kg).  The uncertainty



 associated with the SQC can be obtained from a quantitative  estimate of the degree to which the



 available data support these assertions.
                                        5-4

-------
    The data used in the uncertainty analysis are from the water-only and sediment toxicity tests
                          ;
                          *                                                            '
 that have been conducted to fulfill minimum database requirements for the development of SQC


 (see Section 4.3  and Technical Basis Document, U.S. EPA, 1993a).   These freshwater and


 saltwater tests span a range  of chemicals and organisms; they include both water-only and


 sediment exposures  and they are replicated within each chenucal-organism-exposure media


 treatment.  These data were analyzed using an analysis of variance (ANOVA) to estimate the


 uncertainty (i.e. the variance) associated  with  varying the exposure media and that associated


 with experimental error.  If the EqP model were perfect, then there would be only experimental


 error.  Therefore, the uncertainty associated with the use of EqP is the variance associated with


 varying exposure media.


    The data used  in the  uncertainty analysis are illustrated  in Figure 4-2.  The  data for


phenanthrene are  summarized in Appendix B.  LCSOs for sediment and water-only tests were


computed from these data.  The EqP model can be used to normalize the data in order to put


it on a common basis. The LCSOs from water-only exposures (LCSOw; /*g/L) are related to the


organic carbon-normalized LCSOs from sediment exposures QJC5QSiOC;  pg/goc)  via the


partitioning equation:
             LC50S>OC = KocLC50w                                      (5-3)





The EqP model asserts that the toxicity of sediments expressed on an organic carbon basis equals


the toxicity in water tests multiplied by the KQC.  Therefore, both LC50S(OC and Koc^LCSOw


are estimates of the true LCSOoc for each chemical-organism pair.   In this analysis, the




                                         5-5

-------
 uncertainty of KQC is not treated separately. Any error associated with KOC will be reflected in
                          i
 the uncertainty attributed to varying the exposure media.

    In order to perform an analysis of variance, a model of the random variations is required.

 As discussed above, experiments that seek to validate equation 5-3 lire subject to various sources

 of random variations. A number of chemicals and organisms have been tested. Each chemical-

 organism pair was tested in water-only exposures and in different: sediments. Let a represent

 the random variation due to this source.  Also, each experiment is replicated.  Let € represent

 the random variation due to this source. If the model were perfect, there would be no random   •

 variations other than that due to experimental error which is reflected in the replications. Hence

 or represents the uncertainty due to the approximations inherent in the model and € represents

 the experimental error.  Let (oj2 and  (k                                          (5-4)

 where ln(LC50)iJik,  are either ImXCSOw) or ln(LC50SfOC) corresponding to  a water-only or

 sediment exposure; /tj are the population of ln(LC50) for chemical-organism pair i.  The error

 structure is  assumed to be lognormal which corresponds to assuming that the errors are

proportional to the  means,  e.g. 20%, rather than absolute quantities, e.g.  1 /tg/goc-  The

 statistical problem is to  estimate /tb (aj2, and (
-------
  TABLE 5-2: ANALYSIS OF VARIANCE FOR DERIVATION OF
             SEDIMENT 'QUALITY CRITERIA CONFIDENCE LIMITS FOR
                                PHENANTHRENE.
Source of Uncertainty
Exposure media
Replication
Sediment Quality Criteria
Parameter Value
G*g/goc)
o« 0.39

-------
                TABLE 5-3. SEDIMENT QUALITY CRITERIA
                      CONFIDENCE LIMITS FOR PHENANTHRENE
Sediment Quality Criteria
95% Confidence Limits (ae/e~^
Type of
Water Body
Fresh Water
Salt Water
SQCoc
Atg/goc
180
240
Lower
85
110
Upper
390
510
5.3   COMPARISON OF PHENANTHRENE SQC AND UNCERTAINTY
      CONCENTRATIONS TO SEDIMENT CONCENTRATIONS THAT ARE TOXIC OR
      PREDICTED TO BE CHRONICALLY ACCEPTABLE.

      Insight into  the  magnitude of  protection afforded to benthic  species  by SQC
          i  . _    •        "                                   ;
concentrations and 95.% confidence intervals can be determined from effect concentrations from

toxicity tests with benthic species exposed to sediments spiked with phenanthrene and sediment

concentrations predicted to be chronically safe to organisms tested in water-only exposures

(Figure 5-1).  This is because, effect concentrations in sediments can be predicted from water-
                                                             i
only toxicity data and KQC values (See Section 4). Chronically acceptable concentrations are

extrapolated from genus mean acute values (GMAV) from water-only,  96-hour lethality tests

using acute-chronic ratios (ACR). Therefore, -it may be reasonable to combine these two
                                                                          i
predictive  procedures to  estimate  for phenanthrene,  chronically  acceptable  sediment

concentrations (Predicted Genus Mean Chronic Value, PGMCV) from GMAVs (Appendix A),

ACRs (Table 3-2) and the KQC (Table 5-1):

            PGMCV = (GMAV * ACR)Koc.                          (5-7)

      In Figures 5-1 and 5-2, each PGMCV for fishes, arthropods or other invertebrates tested
                                     5-8

-------
     ~   105
      o
      o
     0)
     1
     UJ
     5
     o
     O
     DC
     X
     O
     UJ
     5
     CO
     z
     UJ
     o
     o
     UJ
    Q
    UJ
    DC
    O.
         104
          103
102
          10
          Water-only tests: (96HR LC50 •«• ACR) KQC

          A  Arthropods
          O  Other Invertebrates
          O  Fishes
                      ACR = 2.01
                                                           t
                                                           O
          a
                                              upper:
                        20
                          40
60
80
100
                    PERCENTAGE RANK OF FRESHWATER GENERA
Figure 5-1.  Comparison between SQC concentrations and 95% confidence intervals, effect
            concentrations from benthic organisms exposed to phenanthrene-spiked sediments
            and  sediment concentrations predicted to be chronically safe in fresh water
            sediments.  Concentrations predicted to be chronically safe (Predicted Genus
            Mean Chronic Values, PGMCV) are derived from the Genus Mean Acute Values
            (GMAV) from water-only 96-hour lethality tests, Acute Chronic Ratios (ACR)
            and KOC values. PGMCV = (GMAV -5- ACR)Koc. Symbols for PGMCVs are
            A for arthropods, O for fishes and D for other invertebrates.  Solid symbols are
            benthic genera; open symbols water column genera. Arrows indicate greater than
            values.  Error bars around sediment LC50 values indicate observed range of
            LCSOs.
                                      5-9

-------
                    Water-only tests: (96HR LC50 •<- ACR) KQC
•—• 10»*
o
o
•£?
CO
^2J*
LU
^ 104
>
g
EC
X
1 103
HI
Si
CO

z
UI
CJ o
Q 102
UI
t-
o
Q
III
DC
CL
1O
f A Arthropods
" D Other Invertebrates
I - O Fishes
ACR «= 2.01
Sediment Tests: IQdlCSO
•rfr£. estuarlus * 3929 fig/Qoc
range 3 tests '3820-4050 ^
6fr L. Dlumulosus » 7585 ua/g^r. B
range 3 tests - 6490 1? 8200^ •
A A
O
7
* •
- ••
: • A .-.
• '••
•



M
* ^ ,
m
m

• -

t • 1 t 1 1 1


'

^P» H
t t jr*
• B t


















1 II I
                        20
40
60
80
100
                    PERCENTAGE RANK OF SALTWATER GENERA
Figure 5-2.  Comparison between SQC concentrations and 95% confidence intervals, effect
            concentrations from benthic organisms exposed to phenanthrene-spiked sediments
            and  sediment concentrations predicted to be chronically  safe in salt water
            sediments.  Concentrations predicted to be chronically safe (Predicted Genus
            Mean Chronic Values, PGMCV) are derived from the Genus Mean Acute Values
            (GMAV) from water-only 96-hour lethality tests, Acute Chronic Ratios (ACR)
            and KOC values. PGMCV = (GMAV -^ ACR)Koc. Symbols for PGMCVs are
            A for arthropods, O for fishes and D for other invertebrates. Solid symbols are
            benthic genera; open symbols water column genera. .Arrows indicate greater than
            values.  Error bars around sediment LC50 values indicate observed range of
            LC50s.
                                      5-10

-------
 in water is plotted against the percentage rank of its sensitivity. Results from toxicity tests with
                         .;

 benthic organisms exposed to sediments spiked with phenanthrene (Table 4-1) are placed in the


 PGMCV rank appropriate to the test-specific effect concentration.  (For example, the 10-day


 LC50 for E. estuarius (3,929 ftg/goc) is placed between the PGMCV of 3,514 ftgfSoc for the


 snail, Nassarius. and the PGMCV of 6,155 ftg/goc for the minnow, Cvprinodon.') Therefore,


 LC50 or other effect concentrations are intermingled in this figure with concentrations predicted


 to be chronically safe.  Care  should be taken by the  reader in interpreting these data with


 dissimilar endpoints.  The following discussion of SQC, organism sensitivities and PGMCVs is


 not intended to provide accurate predictions of the responses of taxa or communities of benthic


 organisms relative to specific concentrations of phenanthrene in sediments in the field.  It is,


 however, intended to guide scientists and managers through the complexity of available data


 relative to potential risks to benthic taxa posed by sediments contaminated with phenanthrene.


       The freshwater SQC for phenanthrene (180 /tg/goc) is less than any of the PGMCVs for


freshwater genera.  In fact, PGMCVs for all 19 freshwater genera are greater than the upper


95 % confidence interval of the  SQC (390 jig/goc). For phenanthrene, the PGMCVs range over


an order of magnitude from the most  sensitive to  the most tolerant genus.  Chronic effect


concentrations  may,  however,  occur  at concentrations below  saturation.   A sediment
                                          ••»

concentration 20 times the SQC would include the PGMCVs of one-half of the 12 benthic genera


tested including stoneflies, mayflies, isopods and catfish.  Tolerant benthic genera such as the


amphipod Gammarus and the  crayfish  Orconectes might be expected to  not be chronically


impacted in sediments with phenanthrene concentrations 1000X the SQC.  This large margin


of safety between all PGMCVs and the SQC results from the need to lower the FCV to. protect
                                       5-11

-------
intergravel dwelling embryos and sac fry of rainbow trout, Oncorhvnchus mykiss.  Benthic
                        /                                        ,
organisms in habitat where salmonids early life stages are absent may be over protected by this

criterion unless species with similar sensitivities are resident at the site.

       The saltwater SQC for phenanthrene (240 /ig/goc) is less than any of the 11 PGMCVs

for saltwater genera.  Only the PGMCV for the mysid shrimp Mysidopsis bahia (314 /tg/goc)

is lower than the upper 95% confidence interval for the SQC.  For phenanthrene, PGMCVs

from the most sensitive to the most tolerant saltwater genus range over an order of magnitude.

A sediment concentration 11 times the SQC would include the PGMCVs of one-half of the 10

benthic genera tested including four arthropod genera and one polychaete genus. Other genera

'of benthic polychaetes and fishes are less sensitive and might not tie expected to be chronically

impacted in sediments with phenanthrene concentrations 20X the SQC. Data from lethality tests

with two saltwater amphipods, Eohaustorius estuarius and Leptocheirus pluinulosus. substantiate

this projection; the 10 day LCSOs from three tests with each species range from 16 to 17 times

the SQC for E. estuarius and from  27 to 34 times the SQC for L. plumulosus (see Section 4).




5.4 COMPARISON OF PHENANTHRENE SQC TO STORET AND NATIONAL STATUS
       AND TRENDS DATA  FOR SEDIMENT PHENANTHRENE:

       A STORET (U.S.  EPA, 1989b) data retrieval was  performed to obtain a preliminary
                                                                                i
assessment of the concentrations of phenanthrene in the sediments of the nation's water bodies.

Log probability plots of phenanthrene concentrations on a dry weight basis in sediments are

shown in Figure 5-3.  Phenanthrene is found at varying concentrations in sediments from rivers,

lakes and near coastal water bodies in the United  States. Median concentrations are generally

about 0.1 /ig/g in each of the three water bodies.  There is significant variability with


                                        5-12

-------
   2|
   B
   «5
   £B
  Is
Figure 5-3.
                   STREAM

                   TOTAL SAMPLES: 584
                   MEASURED  SAMPLES: 175

                               10   20
                               SO
        80  80
                                                              88    88.8
                              TOTAL'SAMPLES: 50""
                              MEASURED  SAMPLES: 29
                               «««<*
                              10  20
                              60
        80  80
                              TOTAL SAMPLES: 87
                              MEASURED SAMPLES:  28

                   10   20
60
                                                80  80
                                                  88    88.8
                                  PROBABBJTY
Probability distribution of concentrations of phenanthrene in sediments from
streams, lakes and estuaries in the United States from 1986 to 1990, from the
STORET (U.S. EPA, 1989b) database, compared to the phenanthrene SQC values
of 18 ug/g in freshwater sediments having TOC = 10%  and  1.8 jtg/g in
freshwater sediments having TOC = 156 and compared to SQC values for
saltwater sediments of 24/tg/g when TOC =10% and 2.4 ug/g when TOC=1%.
The upper dashed line on each figure represents the SQC value when TOC =
10%, the lower dashed line represents the SQC when TOC = 1 %.
                        5-13

-------
phenanthrene concentrations in sediments ranging over seven orders of magnitude within the
                         ',
country.

       The SQC for phenanthrene can be compared to existing concentrations of phenanthrene

in sediments of natural water systems in the United States as contained in the STORET database

(U.S. EPA, 1989b).  These data are generally reported on a dry weight basis, rather than an

organic carbon normalized basis. Therefore, SQC values corresponding to sediment organic

carbon levels of 1 to 10% are compared to phenanthrene's distribution in sediments as examples

'only. For fresh water sediments, SQC values are 1.8 jtg/g dry weight in sediments having 1 %

organic carbon and 18 pg/g dry weight in sediments having 10% organic carbon; for marine

sediments SQC are 2.4 jig/g dry weight and 24 /tg/g dry weight, respectively.   Figure 5-3

presents the. comparisons  of these SQC to probability distributions of observed sediment

phenanthrene levels for streams and lakes (fresh water systems, shown on the upper panels) and

estuaries (marine systems, lower panel). For streams (n = 584) the SQC of 1.8 jtg/g dry weight

for 1 % organic carbon fresh water sediments is exceeded for 4% of the data and the SQC of 18

pg/g dry weight for fresh water sediments having 10% TOC is exceeded by less than 2% of the

data. For lakes (n = 50) neither the SQC for 1 % organic carbon firesh water sediments nor the

SQC for fresh water sediments with 10 % organic^ carbon are exceeded by the post 1986 samples.

 Similarly, in estuaries, the data (n = 87) indicate that neither the criteria of 2.4 ug/g dry weight

for salt water sediments having 1 % organic carbon nor the criteria of 24 j*g/g dry weight for

salt water sediments having 10% organic carbon are exceeded by the post 1986 samples.

       The  phenanthrene distribution in Figure 5-3  includes data  from some  samples hi

which  the phenanthrene concentration was below the detection limit.  These data are indicated
                                        5-14

-------
 on the plot as "less than" symbols (<), and plotted at the reported detection limits. Because
                         3                          '
 these values represent upper bounds and not measured values the percentage of samples in which

 the SQC values are actually exceeded may be less than the percentage reported.

       A second database developed as part of the National Status and Trends Program (NOAA,

 1991) is also  available  for  assessing contaminant  levels in marine  sediments  that are

 representative of areas away from sources of contamination. The probability distribution for

 these data, which can be directly expressed on an organic carbon basis, is compared to the

 saltwater SQC for phenanthrene (240/tg/goc) on Figure 5-4. Data presented are from sediments

 with 0.20 to 31.9 percent organic carbon. The median organic carbon normalized phenanthrene

 concentration (about 5.0 jig/goc) is a factor of 32 below the SQC of 240 ns'Soc- Less than 1 %

 of these samples (n = 900) exceeded the criteria. Hence, these results are consistent with the

 preceding comparison of the marine SQC to STOEET data.

      Regional differences in phenanthrene concentrations may affect the above conclusions

 concerning expected criteria exceedences.  This analysis also does not consider other factors

 such as the type of samples collected (i.e., whether samples were from surficial grab samples

 or vertical core profiles), or the relative frequencies and intensities of sampling in different study

 areas.  It is presented as an aid in  assessing ^the range of reported  phenanthrene sediment

 concentrations and the extent to which they may exceed the SQC.

 5.5 LIMITATIONS TOTHE APPLICABILITY OF SEDIMENT QUALITY CRITERIA:

      Rarely, if ever,  are contaminants found alone  in  naturally occurring sediments

 Obviously, the fact that the concentration of a particular contaminant does not exceed the SQC

does not mean that other chemicals, for which there are no SQC available, are not present in
                                       5-15

-------
in 1 1 rn
           miti i i  uuiu i i  uiiiu i i  uiini i

                              o
                              A

                              8
                              <•_
                              i
                              o
mi I t  •mil in  miti» i  mum
                                    *
                                   8
       (OO 0/Dn) 1N3WIQ3S
                          5-16

-------
 concentrations sufficient to cause harmful effects.  Furthermore, even if SQC were available for
                         ; ;

 all of the contaminants in a particular sediment, there might be additive or synergistic effects


 that the criteria do not address.  In this sense the SQC represent "best case" criteria.


       The concerns about mixtures of contaminants are particularly important with the PAHs,


 which almost invariably occur as complex mixtures. Some guidance on interpretations of PAH


 concentrations is possible given the presence of SQC for phenanthrene and other  individual


 PAHs.  This is because much is known about the toxicity and structure-activity relationships of


 the so-called narcosis chemicals, a group of nonionic organic chemicals to which  the PAHs


 belong.  The toxicity of the narcosis chemicals is additive (Broderius and Kahl, 1985). The


 toxicity of these chemicals increases with  increasing  KQW (Veith et al., 1983)  and their


 bioavailability in sediments decreases as a function of its KQW Therefore, the toxicities of many


 PAHs in sediments are likely to be similar.  This explains why SQC values for fluoranthene


 (fresh: 620 pg/goc, salt: 300 ng/go^, acenaphthene (fresh:  130 jtg/goc, salt: 230 /tg/goc) and


 phenanthrene (fresh: 180  /tg/goc, salt: 240  /tg/goc) differ little and why it  is theoretically


 possible to develop an SQC for total PAHs.  EPA is currently conducting research aimed at


 development of SQC for combined PAHs.                         5      _


      It is theoretically possible that antagonistic reactions between chemicals could reduce the


 toxicity of a given chemical such that  it might not cause unacceptable effects on  benthic


 organisms  at concentrations above the SQC  when it occurs with the antagonistic  chemical:


However, antagonism has rarely been demonstrated. What should be much more common are


 instances where toxic effects occur at concentrations below the SQC because of the  additivity


 of toxicity of many common contaminants (Alabaster and Lloyd, 1982), e.g. heavy metals and
                                        5-17

-------
 PAHs, and instances where other toxic compounds for which no SQC exist occur along with
                          t

 SQC chemicals.



       'Care must be used in application of EqP-based SQC in disequilibrium conditions.  In



 gome instances site-specific SQC may be required to address this condition. EqP-based SQC



 assume that nonioriic organic chemicals are in equilibrium with tide sediment and IW and are



 associated with sediment primarily through adsorption into sediment organic carbon. In order



 for these assumptions to be valid, the chemical must be dissolved in IW and partitioned into



 sediment organic carbon. The chemical must, therefore, be associated with the sediment for a



 sufficient length of time for equilibrium to be reached. In sediments where particles like cinder,



 soot, or oil droplets contain PAHs, disequilibrium exists and criteria are over protective.  In



 liquid chemical spill situations disequilibrium  concentrations in interstitial and overlying water



 may be proportionately higher relative to sediment concentrations. In this case criteria may be



 underprotective.



       In very dynamic areas,  with highly erosional or depositional bedded sediments,



 equilibrium may not be attained with contaminants.  However, even high KQW nonionic organic



 compounds come to  equilibrium in clean  sediment in a period of days,  weeks  or months.



 Equilibrium times are shorter for mixtures of two sediments each previously at equilibrium.



 This is particularly relevant in tidal situations where large volumes of sediments are eroded and
                                                                        /


 deposited, yet near equilibrium conditions may predominate over large areas. Except for spills



and particulate  chemical, near  equilibrium is the rule and disequilibrium is uncommon!   In



instances where it is  suspected that EqP does not apply for a particular sediment because  of



disequilibrium  discussed above, site-specific methodologies may be applied (U.S. EPA, 1993b).
                                        5-18

-------
                                     SECTION 6
                              CRITERIA STATEMENT
       The procedures described in the  "Technical Basis for Deriving Numerical Sediment
 Quality Criteria for Nonionic Organic Contaminants for the Protection of Benthic Organisms by
 Using Equilibrium Partitioning" (U.S. EPA, 1993a) indicate that benthic organisms should be
 acceptably protected in freshwater sediments containing <. 180 pg phenanthrene/g organic
 carbon and saltwater sediments containing <, 240 pg phenanthrene/g organic carbon, except •
 possibly where a locally important species is very sensitive or sediment organic carbon is <
 0.2%.
             Confidence limits of 85 to  390 jig/goc for freshwater sediments and liO to 510
 Mg/goc for saltwater sediments are provided as af? estimate of the uncertainty associated with the
 degree to  which the observed concentration in sediment (ftg/goc), which may be toxic, can be
 predicted  using the organic carbon partition  coefficient (K^ and the water-only  effects
 concentration.  Confidence limits do not incorporate uncertainty associated with water quality
criteria. An understanding of the theoretical basis of the equilibrium partitioning methodology,
uncertainty, the partitioning and toxicity of phenanthrene, and sound judgement are required in
the regulatory use of SQC and their confidence limits.
These concentrations represent the U.S.  EPA's best judgement at  this  time of the levels of
                                        6-1

-------
phenanthrene in sediments that would be protective of benthic species. It is the philosophy of
                         ».

the Agency and the EPA Science Advisory Board that the use of sediment quality criteria (SQCs)


as stand-alone, pass-fail criteria is not recommended for all applications and should frequently


.trigger additional studies at sites under investigation. The upper confidence limit should be


interpreted as a .concentration above which impacts on benthic species should be expected.
                                                                    •

Conversely, the lower confidence limit should be interpreted as a concentration below which


impacts on benthic species should be unlikely.
                                          6-2

-------
                                    SECTION 7
                                  REFERENCES

Abemathy, S., A.M. Bobra, W.Y. Shiu, P.O. Wells and D. Mackay.   1986.  Acute lethal
   .    toxicity of hydrocartrons and cMorinated hydrocarbons to
       key role of organism-water partitioning. Aquat. ioxicoL  8:163-174.

Adams, W.J., R.A. Kimerle and R.C. Mosher. 1985. Aquatic safety assessment of chemicals
       sorbed to sediments.  In: Aquatic Toxicology and Hazard Assessment:   Seventh
       Symposium. Eds: R.D. Cardwell, R. Purdy and R.C. Banner. Amer. Soc. Testing and
       Materials, Philadelphia, PA. STP 854. pp. 429-453.

Alabaster, J.S. and R, Lloyd. 1982.  Water Quality Criteria for freshwater fish.  Chapter 11.
       Mixtures of Toxicants. London, Butterworth Scientific.

Banerjee,  S.; S.C. Valvani,  S.H.  Yalkowsky. 1980.  Water solubility and octanol/water
       partition  coefficients of organics:  Limitations of the solubility-partition coefficient
       correlation.  Environ. Sci. Technol. 14(10): 1227-1229.

Battelle Ocean Sciences.  1987.  Acute toxicity of phenanthrene to saltwater animals.  Report
       to U.S. EPA Criteria and Standards Division. Battelle Ocean Sciences, Duxbury,.MA.

Bidleman, T.F.  1984.  Estimation of vapor pressures for nonpolar organic compounds by
       capillary  gas  chromatography.   Anal. Chem. 56:2490-2496.   Quoted  by U.S.
       Environmental Protection Agency. 1987. Health and environmental effects profile for
       phenanthrene.  U.S. EPA, Environmental  Criteria and Assessment Office, Office of
       Health and Environmental Assessment, Office of Research and Development, Cincinnati,
       OH.  77pp.

Black,  J.A., W.J. Birge, A.G. Westerman and P.C. Francis.  1983.   Comparative aquatic
       toxicology of aromatic hydrocarbons.  Fundam. Appl. Toxicol. 3:353-358.

Broderius, S. and M.  Kahl. 1985. Acute toxicity of organic chemical mixtures to the fathead
       minnow.  Aquatic Toxicol. 6:307-322.

Brookes,  P. 1977. Mutagenicity of polycyclic aromatic hydrocarbons. Mutation Res. 39:257-
       284.

                                        7-1

-------
 Bruggeman,  W.A., J.  Van der Steen, and O. Hutzinger.   1982.,  Reversed-phase thin-layer
       chromatography of'polynuclear aromatic hydrocarbons and chlorinated biphenyls:
       Relationship with hydrophobicity as measured by aqueous solubility and octanol-water
       partition coefficient. J. Chromatogr. 238:335-346.

 Call, D.J., L.T. Brooke, S.L. Halting, S.H. Pokier and D.J. McCauley.  1986.  Toxicity of
       phenanthrene to several freshwater species. Final report to Battelle Memorial Research
       Institute, Columbus, OH.  Center for Lake Superior Environmental Studies, University
       of Wisconsin-Superior,  Superior, Wis.  18 pp.

, Chapman, G. A. 1987. • Establishing sediment criteria for chemicals-regulatory perspective. In:
       Fate and Effects of Sediment-Bound Chemicals in Aquatic Systems.  Editors: K.L.
       Dickson, A.W.  MaM and W.A. Biungs. Pergamon Press, New York. pp. 355-376.

 Chou, J.T. and P.C. Jurs.  1979.  Computer-assisted computation of partition coefficients from
       molecular structures using fragment constants.  J.Chem. Laf. Comput. Sci. 19(3): 172-
       178.

 De Bruijn, J., F. Busser, W. Seinen, and J. Hermens.  1989. Determination  of octanol/water
       partition coefficients for hydrophobic organic chemicals with the "slow-stirring" method.
       Environ.' ToxicoL Chem. 8:499-512.

De Witt, T.H., R.J. Ozretich,  R.C. Swartz, J.O.  Lamberson, D.W.  Shults, G.R. Ditsworth,
    •   J.K.P. Jones, L. Hoselton, and L.M. Smith. 1992. The effect of organic matter quality
       on the toxicity and partitioning of sediment-associated fluoranthene to the infaunal marine
       amphipod, Rhepoxvnius abronius. Environmental Toxicology and Chemistry 11:197-208.

Di Toro, D.M., 1985.  A particle interaction model of reversible organic chemical sorption.
       Chemosphere. 14(10): 1503-1538.      '

Di Toro,  D.M., C. Zarba, D.J. Hansen, R.C.  Swartz,  C.E.  Cowan, H.E. Allen, N.A.
       Thomas, P.R. Paquin, and W.J. Berry. 1991.  Technical basis for establishing sediment
       quality  criteria  for non-ionic organic chemicals  using  equilibrium partitioning.
       Environmental Toxicology and Chemistry 10:(12)1541-1583.
                                                                                 r
Eadie, B.J., P.P. Landrum, W. Faust. 1982. .Polycyclic aromatic hydrocarbons in sediments,
       pore water and  the amphipod Pontoporeia hoyi from Lake Michigan. Chemosphere
       ll(9):847-858.
                                                                                   /
Eastmond, D.A., G.M.  Booth and M.L. Lee.  1984. Polycyclic accumulation, and elimination
       of polycyclic aromatic sulfur heterocycles in Daphnia magma.  Arch.  Environ. Contain
       Toxicol. 13:105-111.
                                        7-2

-------
 Geiger, J.G. and A.L. Buikema, Jr.  1981.  Oxygen consumption and filtering rate of Daphnia
       pulex after exposure
-------
       DC, Final Report, EPA-44-/4-81-041.
                         ',                                              ,  '  '
Mackay, D., A. Bobra, and W.Y. Shui. 1980.  Relationships between aqueous solubility an
-------
 Rossi, S.S.  and J.M.  Neff.  1978.  Toxicity of polynuciear aromatic hydrocarbons to the
       polychaete Neanthes arenaceodentata. Mar. Poll. Bull. 9:220-223.

 Schuytema, G.A., A.V. Nebeker, W.L. Griffis, and C.E. Miller. 1989.  Effects of freezing
       on toxicity of sediments contaminated with DDT and endrin.  Environ. Toxicol. and
       Chem. 8(10):883-891.

 Stephan, C.E., D.I. Mount, D.J. Hansen, J.H. Gentile, G.A. Chapman, and W.A. Brangs.
       1985. Guidelines for deriving numerical national water quality criteria for the protection
       of aquatic organisms and their uses.  PB85-227049.  National Technical Information
       Service, Springfield, VA. 98 pp.

 Swartz, R.C. 1991. Acenaphthene and penanthrene files.  Memorandum to David J. Hansen
       June 26,  1991. 160 pp.

 Swartz, R.C., D.W. Schults, T.H. DeWitt, G.R. Ditsworth, and J.O. Lamberson.  1990.
       Toxicity of fluoranthene in sediment to  marine amphipods: A test of the equilibrium
       partitioning approach to sediment quality criteria. Environ. Toxicol. Chem. 9(8): 1071-
       1080.
 Trucco, R.G., F.R. Engelhardt and B. Stacey. 1983.  Toxicity, accumulation and clearance of
       aromatic hydrocarbons in Daphnia pulex. Environ. PoUut. (Series A) 31:191-202.

 U.S. Environmental Protection Agency. 1978.  In-depth studies on health and environmental
       impacts of selected water pollutants (Table of data available from Charles E.Stephan U.S.
       EPA, Duluth, MN).

 U.S. EPA. 1980. Ambient water quality criteria for polynuciear aromatic hydrocarbons. EPA-
       440/5-80-069orPB81-117806. National Technical Information Service, Springfield, VA.

 U.S. Environmental Protection Agency. 1985. Appendix B - Response to public comments on
       "Guidelines for deriving numerical national water quality criteria for the protection of
       aquatic organisms and their uses." July 19, 1985. Fed. Regist. 50:30793-30796.

 U.S. Environmental Protection Agency. 1987. Quality Criteria for Water, 1986. EPA 440/5-
       86-001. May 1,1987. U.S. Government Printing Office No. 955-002-00000-8. 406pp.

U.S. Environmental Protection Agency. 1989a.  Sediment classification methods compendium.
       Watershed Protection Division, U.S. EPA.  280 pp.

U.S. Environmental Protection Agency. 1989b.  Handbook: Water Quality Control Information
       System, STORET.  Washington, D.C., 20406.
                                       7-5

-------
 U.S. Environmental Protection Agency.  1989c.  Briefing Report to the EPA Science Advisory
       Board on the Equilibrium Partitioning Approach to Generating Sediment Quality Criteria.
       Office of Water Regulations and Standards, Criteria and Standards Division, 132 pp.

 U.S. Environmental Protection Agency. 1993a. Technical Basis for Deriving Sediment Quality
       Criteria for Nonionic Organic Contaminants for the Protection of Benthic Organisms by
       Using Equilibrium Partitioning.  (In Review).

 U.S. Environmental  Protection Agency.   1993b.   Guidelines for Deriving Site-Specific
       Sediment "Quality Criteria for the Protection of Benthic Organisms.  (In Review).
               »

 U.S. Environmental Protection Agency. 1993c. Guide for the Use and Application of Sediment
       Quality Criteria for Non-Ionic Organic Chemicals.  (In Review).

Veith, G.D., D.J.  Call, and L.T. Brooke. 1983.  Stnicture-toxicity relationships for the fathead
       minnow, Pimephales promelas: narcotic industrial chemicals.  Can. J. Fish. Aquat. Sci.
       40:743-748.

Verschueren,  K.  1983.  Handbook of Environmental Data on Organic Chemicals,  Second
       Edition. Van Nostrand Reinhold Company, New York.  1310pp.

Yalkowsky, S.H., S.C. Valvani, and D.  Mackay. 1983.  Estimation of the aqueous solubility
       of some aromatic compounds. Residue Rev. 85:43-55.
                                        7-6

-------
              lo
                                 VD
                                 CO
                                 en
                                 Q)
                                 0
          U)
          CO
          en
                                          0)
         r
         a
                                                   00

                                                   Cn
                   a


                   0


                   c
         m
         a
         H
                            oo
                            cn
                                                                     00
                                                                     cn
                                                                     4J
                                                                     0
                                                                     g
vo
eo
en
         co
         cn
         tH
                                                                                       -H
                                                                                       E
                                                                                       "O

                                                                                       §
CO

I
          CO
          en
                  s

                                                                                                t»H
                           CO
                           en
                           rH
                           ed

                           4J
                           0

                           O
                           U
                           u
                  S
en    fc
rH    frl
 vo
 00
 4J

 0
«	1

3
                   oo
                   cn
                   tH
                                                                                                    4J

                                                                                                    0)
                 .3
                                en
cn
iH

A
                                                                                                        cn
                                                                        eo
                                                                        0
                           to
                           CN
                                            0
                                            en
                                                                                                                                   in
                                                                                                                                   r>
                                                                                                            o
                                                                                                            in
                           tH


                            A
vo
cn
                                         cn
                                         H
                          co
                          o
u>

                                         cn
                                         rH
                                                                                                        0
                                                                                                        o
                                                                                                                 CN
                                                                                          o
                                                                                          cn
                                                     in
                                                     r-
                                                     m
                                                                                                                                            o
                                                                                                                                            in
                                                                                                                                                     ro
                                en
         cn
         tH
         *i
         A
                                                                                               o
                                                                                               in
                                                  00
                                                                    0
                                                                    cs
                          O
                          O
                          v>
                          oa
         o
         cn
         «*
                                                                                         in
                                                                                         r>
                                                                                                                                            in
                                                                                                                                            H
                                                                                                                                            iH
                                                                                                                    ro
                                                                                                                    CN
                                                                                                                                  S
                                                                                                                                                    S
                               E      E
                                                 CO
                                                          CO
                                                                   CO
                                                                                     CO
                                                                                              CO
                                                                                                       co
                                                                                                                         co
                                                                                                  e
                                                                                                                                           CO
             M
          H EH

          H> CO
                                        x       x
                             Rj    VO CD


                             |    4|   «
                                                                                                                H
                                                                                                                                                    ft
                                                                                     VJ3    X
                                                                                                       X
O

X.
a

S!
                                  S1   -o 3
                                «      -H
                               «J nJ    rH
                    BJ

                -  .g

                 S

                 u  at
                 O-H
                 o  c
                '*
Cladoceran,
Daphnia pulex
Cladoceran,
Daphnia pulex
Midge

Chi
                                                                                                                                     0
                                                                                                                                     CO
                                                                                                                                    -rl
                                                                                                                                    M
Rainbow trout

Oncorhchus

-------




















.








"0
o
£t
-S
U
e
o
TJ
i
8)

Jc
4-1
Q
K3
j






"Bo
g
S
'I
i
p£ j> -
PQ «q£
& .E
s
.,
2*
6
•M
{5°
'*B
i
&
CO
o
§8
M
*|
Pi
op
0
98
2 w


to §
3 P-




M
S
S
52
*
H






§f
d
0) 0
o1"
Q) 0
rH 0
rH t
(D C
4J C
*J-r
at t
me/
„,
••I •
***•• in
cn co
3. tH
^
•3 •
S*^ l/l
Cn oo
% H
••? *
tn co
=1. CO) H-
H • ' '
O «gt
31 CO CO
^ OJ *"*
g
1 o
co)
--

H



10

.
CO
'3
-w
at
•H
rH
Archiannelid,
Dinophilus qvroci


CO
r-
o\
*"*

* *| I
I Ut
i 01
- -C
I §
! -H
i ca
i -m
> o
i «


o
o
VD


O
O
u>

o
o
•
o
o

D

CO

H



"*

•Rt
4-1
at
4J
B
01
•D
Polychaete worm,
Neanthea arenaceo



*


§a

tn cr in *4* C^ A in ^< CN A tn A in A . g CJ Ss M" "* Blue mussel, Mvtilus eduiis Sr a \ Q) 0 1 V r O ' (0 D ) rH C i rH I ! 0) C > 4J g 4Jf at c ) oa cr in ^ CN A in *$* CN A in A in A S3 DJ H ^ Soft -shell clam, Mva arenaria • c t- ) S o \ CU O 1 U r* ' CO 0 1 rH ( 1 rH t ! CD C > 4J 0 1 4J-r- > Rt I > n cr l l tH g E M 13 Mysid, Mvsidopsis bahia t- co cn rH »7 0) •H J CD \ CD t 3 3 i -0 1 i i i « H cn tH H cn f^ CN tH cn t-i N o' t*- N a 'E M 13 Mysid, Mvaidopsis bahia tH O\ a\ iH N" 4J 1 ^, • CO cn tH «• • CO cn tH ' *. ' CO cn tH •* CO cn . tH S E M ^ a a O rH 3 Amphipod, Leptocheirus plum SS S ff o CD 0 rH C rH C. CD C 4J a Mi- Hi C n cs t 1 CO O O D OS, ^ M" ^ OI Grass shrimp, Palaemonetes puqi §t- cc > CD cr O i 0) Q I rig SS tit I a cr « • in tH ^ • in tH «, in ^ H •* in tH s E ^ U < OI Grass shrimp, Palaemonetes puqi i 3 a t 0) O 0 ' o a H < 'co c ,u a 4J-i- rt t 1 ffl cr • CO u> H ,> rt r> \D H jg :E' M «t m 3 Hermit crab, Paqurus lonqicarp §f a t 0) o 0 " a> a 1 rH ( l rH t 1 CD C 4J d 4J-r at l > . nc/ l l t in CN' A D - M" ••3 a £j Sheepshead minnow Cvprinodon varieq • Cr 3 at ci ) ft0 O " CD Q 1 rH ( 1 rH C ! 01 C i 4J a 1 4J-r i at c > n a W • cn IN . G\ d « cn CN ^ cn CN S3 £ H* *3 CO 4J - irJ Sheepshead minnow Cvprinodon variew c t~ > 0) CO \ CD cn 0 ' CD 0 1 rH 01 1 rH C ! CD C 1 4J 0 < 4J-r ) (d U > n ca ' .CO o tH co o tH CO. O H CO o • tH S e £E t' < •' ^ CD •S Atlantic ailversii Menidia menidia IN


-------
•d
smbryo, U ..lifestage and habitat unknown, x -« lifestage
nn.








•3
o
3
•H-
JJ
g
^
• •
0
- Phenanthren

<
X
*r4

£
0)
Pi
w
H
H

f
(S
rH

U
Hi
0
. rH
-ri
§
•n
n
1
u

0
S1
JJ
m
0
IH
-H
s
3
rH
8

I
U

9
O
-H .
B
-H
. &
n
| «
! 1
*rt ._j

jj H
3
a)
r< -U
C -H
1 «
J2 JG
.>

.
•§,
3
|
§
rH
IH
n
fe
, »
renewal
n
o:
0
-H
JJ
m
n
CO
•O
o
a
jj
1
. measured. .
r*i
S
•H
W
•s
H
53
+•
*-^»
g
-H
£
•O
S
3
m
oi
0
n
&
g
-H
JJ
g
JJ
B
0
|
in
00
on
rH
.
rH •
a
0
a
0
jj
CQ
E
i.
a
g
•H
jj
a
0

^
o
in
u •
n
o
o
§~
W
o\
0
3

0

he geometric mean of acute values by species for benthic
jj
m
•H
CQ
0
-H
U
60-
i

3
gj
0
JJ
§
B: Habitat He
i lifestages.
0 S
«!
BQ ^

£$ ®
O
1
RJ
B
-H
JJ
•g

m
0
•H
U
0
&

M *
o
Ej
g
IH
O
§
0
: Geometric m
CO
1

.»
|
pecies, habitats and lifestages within the genus.
m
m
m
• j
a
m
rH
a

0
8
BJ
•s
I
V: Geometric
|rf
8
tH
(ti
S
1

-------
i CO
 O
                    CQ
                    8
                o
I,
i
      Q


      1

      B


M DO 63

H « CJ
O t? W
w o a.
CO CQ CO
            H
            cn
            en
                            N
                            XI

                           5
                             cn
                             cn
                             iH
                                                        XI
                                                        H
                                                        CO
                        cn
                        cn
                                ro cn w co (T> TJ< r-
                                »* ro
            cooooooooco

            OHHrlHHHHO
                                                1-1 ro cn c« t* in
                                                ^c ro N en * *
                                                          OCOOOCnCOiHH
                                                            r4 co in cn iH o cn
OOOOOOOOO
     O CM VD VD f CO !••
     <* VD o t^ in co o
          H i-l CM ^« O
        co co   co co in


          H H C4 << CO
                                                       ooooooooo
                                                            VD t> O U) CO t» (N
                                                                        < CO CN
                                                                         in co
                                                            «   N ^ in in

                                                       ooi-fotr-cotHin«j<
                                                            H 1-1 o << IB ^i cn
                                     in in   in in
                          in in in" in cj r^ o cq t
                                     tH H cn in cn
                                                         in    in in in in in
                                                                      CO
                                                                           O
                                                                           H
                           •»    m
                           «f
                           0)
                          ffl    m
                          j:    3

                          i«ti
                          coo u
                               «^
                               O  w

                            0)  -4J

                            §•&§
                              i t-l  O
                               cow
   (0

*n5
   k
   (S
   3
   X)
   m
   (D
                                                               in w vo o CN o] IH
                                                               »* <«» m tt rt  o o o o o o o
                                                                         «D 03 H ^ tN H CO
                                                                         in attn u> co vo in
                                                                                         Cn
                                                                                         vo
                                                                        ' CO Cl O *> O t» «N
                                                                         H m in in vo co co
                                                                                        rt
                                                              in   in in •     in
                                                               •    •  •       •

                                                         CM              iH 01 O\
                                                                                   O)    m
r-H O  V4
CQ    O
  . .-«
O-H  O
H CO

-------































^^
I)
g

*"*
0
c
1
AJ
0

Ol
0)
M
APPEND

CO

Fj
g
" -a
M
S
s


^

gS



23
5 .
IJII~
S S ^iit
E ^"
S? J3 <
S o*"*'
g£5
8

o 3
a
SC9
K
H O
EH
Sis
S S SM
8|g

g
M
la]*
g"~
§


Q
13
co
. £
S5-V.CO
HUH
SDH
MOj 0
Q 5 (M
U O O<
CO CO CO
j
^ H rH
o\ a\ a\
d O\ O\
«H ' H H
« .
J? • " N
•W - -W 4J
g « «
co eo co

M ca •* t» r> ro'i" o\<«)ocfto\meo cnmotninr^fro

(N VD VO VO U> VD U> U> CS U> VO f. («.
«o en en en o\ en en o> co ro in in in in in in in ro vo u> vo u> u> \o u> en  rH t^ co o o «r u>o
minert<*j<4i\OrH t- N c\ o eo en «*< t-cNVCi-icncocovo
rt N CO CO t-l H CO « W> t-l H H N CS * CO H



«* rH COU>r- CO if rH N in rH U> fft*f rf
* * «•• ••• • ,, . •••
oovoHcnr-ooot^ oomo
-------

-------