United States Office of Research and Development EPA-600-R-02-011
Environmental Protection Washington, DC 20460 www.epa.gov
Agency
wEPA Procedures for the Derivation of
Equilibrium Partitioning
Sediment Benchmarks (ESBs)
for the Protection of Benthic
Organisms: Metal Mixtures
(Cadmium, Copper, Lead, Nickel,
Silver, and Zinc)
-/ I ;',. .. , . ' *W*E--
' . v . ' -'.; ' : -
-------
EPA/600/R-02/011
January 2005
Procedures for the Derivation of
Equilibrium Partitioning Sediment Benchmarks (ESBs)
for the Protection of Benthic Organisms: Metal Mixtures
(Cadmium, Copper, Lead, Nickel, Silver and Zinc)
David J. Hansen
(formerly with U.S. EPA)
Dominic M. DiToro
Univ. Delaware, Newark, DE; HydroQual, Inc..
Mahwah, NJ
Walter J. Berry
Warren S. Boothman
Robert M. Burgess
National Health and Environmental Effects Research Laboratory
Atlantic Ecology Division
Narragansett, RI
Gerald T. Ankley
David R. Mount
National Health and Environmental Effects Research Laboratory
Mid-Continent Ecology Division
Duluth, MN
Joy A. McGrath
Laurie D. DeRosa
HydroQual, Inc., Mahwah, NJ
Heidi E. Bell
Mary C. Reiley
Office of Water, Washington, DC
Christopher S. Zarba
Office of Research and Development, Washington, DC
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Atlantic Ecology Division, Narragansett, RI
Mid-Continent Ecology Division, Duluth, MN
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Notice
The Office of Research and Development (ORD) has produced this document to provide procedures for the
derivation of equilibrium partitioning sediment benchmarks (ESBs) for metal mixtures. ESBs may be useful as a
complement to existing sediment assessment tools. This document should be cited as:
U.S. EPA. 2005. Procedures for the Derivation of Equilibrium Partitioning Sediment Benchmarks
(ESBs) for the Protection of Benthic Organisms: Metal Mixtures (Cadmium, Copper, Lead,
Nickel, Silver and Zinc). EPA-600-R-02-011. Office of Research and Development. Washington,
DC 20460
This document can also be found in electronic format at the following web address:
http://www.epa.gov/nheerl/publications/
The information in this document has been funded wholly by the U. S. Environmental Protection Agency. It has
been subject to the Agency's peer and administrative review, and it has been approved for publication as an EPA
document.
Mention of trade names or commercial products does not constitute endorsement or recommendation for use.
Abstract
This equilibrium partitioning sediment benchmark (ESB) document describes procedures to derive concentrations
of metal mixtures in sediment which are protective of the presence of benthic organisms. The equilibrium
partitioning (EqP) approach was chosen because it accounts for the varying biological availability of chemicals in
different sediments and allows for the incorporation of the appropriate biological effects concentration. This
provides for the derivation of benchmarks that are causally linked to the specific chemical, applicable across
sediments, and appropriately protective of benthic organisms.
EqP can be used to calculate ESBs for any toxicity endpoint for which there are water-only toxicity data; it is not
limited to any single effect endpoint. For the purposes of this document, the ESB for mixtures of the metals
cadmium, copper, lead, nickel, silver, and zinc, the ESBAVS. , is derived based on two complementary approaches.
In the first approach, the ESBAVS. is based on the solid phase and interstitial water phase of sediments. In
sediments, these metals should not cause direct toxicity to benthic organisms if the ZSEM-AVS is < 0.0. In the
second approach, sediments containing these metals should not cause direct toxicity to benthic organisms if the sum
of the dissolved interstitial water concentrations for each of the metals (SM. d) divided by their respective Water
Quality Criteria (WQC) Final Chronic Value (FCV) is < 1.0. Uncertainty bounds on ZSEM-AVS and (2SEM-
AVS)//OC can be used to identify sediments where toxicity, because of these metals, is unlikely, uncertain, or likely.
If the ZSEM-AVS is > 0.0 or ZM.d divided by their respective FCVs is >1.0, effects may occur with increasing
severity as the degree of exceedance increases. A procedure for addressing chromium toxicity in sediments is also
included in an appendix.
The ESBs do not consider the antagonistic, additive or synergistic effects of other sediment contaminants in
combination with metal mixtures or the potential for bioaccumulation and trophic transfer of metal mixtures to
aquatic life, wildlife or humans.
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Foreword
Under the Clean Water Act (CWA), the U.S. Environmental Protection Agency (EPA) and the
States develop programs for protecting the chemical, physical, and biological integrity of the
nation's waters. To support the scientific and technical foundations of the programs, EPA's Office
of Research and Development has conducted efforts to develop and publish equilibrium partitioning
sediment benchmarks (ESBs) for some of the 65 toxic pollutants or toxic pollutant categories.
Toxic contaminants in bottom sediments of the nation's lakes, rivers, wetlands, and coastal waters
create the potential for continued environmental degradation even where water column contaminant
levels meet applicable water quality standards. In addition, contaminated sediments can lead to
water quality impacts, even when direct discharges to the receiving water have ceased.
The ESBs and associated methodology presented in this document provide a means to estimate the
concentrations of a substance that may be present in sediment while still protecting benthic
organisms from the effects of that substance. These benchmarks are applicable to a variety of
freshwater and marine sediments because they are based on the biologically available concentration
of the substance in the sediments. These ESBs are intended to provide protection to benthic
organisms from direct toxicity due to this substance. In some cases, the additive toxicity for
specific classes of toxicants (e.g., metal mixtures or polycyclic aromatic hydrocarbon mixtures) is
addressed. The ESBs do not consider the antagonistic, additive or synergistic effects of other
sediment contaminants in combination with metal mixtures or the potential for bioaccumulation and
trophic transfer of metal mixtures to aquatic life, wildlife or humans.
ESBs may be useful as a complement to existing sediment assessment tools, to help assess the
extent of sediment contamination, to help identify chemicals causing toxicity, and to serve as
targets for pollutant loading control measures.
This document provides technical information to EPA Program Offices, including Superfund,
Regions, States, the regulated community, and the public. For example, ESBs when used in the
Superfund process, would serve for screening purposes only, not as regulatory criteria, site specific
clean-up standards, or remedial goals. The ESBs do not substitute for the CWA or EPA's
regulations, nor is it a regulation itself. Thus, it cannot impose legally binding requirements on
EPA, States, or the regulated community. EPA and State decision makers retain the discretion to
adopt approaches on a case-by-case basis that differ from this technical information where
appropriate. EPA may change this technical information in the future. This document has been
reviewed by EPA's Office of Research and Development (Mid-Continent Ecology Division, Duluth,
MN; Atlantic Ecology Division, Narragansett, RI), and approved for publication.
Mention of trade names or commercial products does not constitute endorsement or
recommendation of use.
This is contribution AED-02-048 of the Office of Research and Development National Health and
Environmental Effects Research Laboratory's Atlantic Ecology Division.
Front cover image provided by Wayne R. Davis and Virginia Lee.
ill
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
IV
-------
Contents
Contents
Notice ii
Abstract a
Foreword in
Acknowledgments ix
Executive Summary xi
Glossary of Abbreviations xiii
Section 1
Introduction 1-1
1.1 General Information 1-1
1.2 Applications of Sediment Benchmarks 1-4
1.3 Overview 1-5
Section 2
Partitioning of Metals in Sediments 2-1
2.1 Metal Toxicity in Water-Only and in Interstitial Water of Sediment Exposures 2-1
2.1.1 Toxicity Correlates to Metal Activity 2-2
2.1.2 Toxicity Correlates to Interstitial Water Concentration 2-4
2.2 Solid-Phase Sulfide as the Important Binding Component 2-8
2.2.1 Metal Sorption Phases 2-8
2.2.2 Titration Experiments 2-9
2.2.2.1 Amorphous FeS 2-11
2.2.2.2 Sediments 2-11
2.2.3 Correlation to Sediment AVS 2-12
2.2.4 Solubility Relationships and Displacement Reactions 2-13
2.2.5 Application to Mixtures of Metals 2-14
Section 3
Toxicity of Metals in Sediments 3-1
3.1 General Information 3-1
3.1.1 Terminology 3-1
3.2 Predicting Metal Toxicity: Short-Term Studies 3-1
3.2.1 Spiked Sediments: Individual Experiments 3-1
3.2.2 Spiked Sediments: All Experimental Results Summarized 3-5
3.2.3 Field Sediments 3-9
3.2.4 Field Sites and Spiked Sediments Combined 3-11
3.2.5 Conclusions from Short-Term Studies 3-13
3.3 Predicting Metal Toxicity: Long-Term Studies 3-14
3.3.1 Life-Cycle Toxicity Tests 3-14
3.3.2 Colonization Tests 3-16
3.3.3 Conclusions from Chronic Studies 3-17
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
3.4 Predicting Toxicity of Metals in Sediments 3-17
3.4.1 General Information 3-17
3.4.2 EqP Theory for SEM, AVS, and Organic Carbon 3-19
3.4.3 Data Sources 3-20
3.4.4 Acute Toxicity Uncertainty 3-20
3.4.5 Chronic Toxicity Uncertainty 3-22
3.4.6 Summary 3-22
Section 4
Derivation of Metal Mixtures ESB .v_ wri_s 4-1
AV o! W {J^s
4.1 General Information 4-1
4.2 Sediment Benchmarks for Multiple Metals 4-2
4.2.1 AVS Benchmarks 4-2
4.2.2 Interstitial Water Benchmarks 4-2
4.2.3 Summary 4-3
4.3 Example Calculation of ESB s for Metals and EqP-Based Interpretation .. 4-3
4.4 ESB for Metals vs. Environmental Monitoring Databases 4-5
4.4.1 Data Analysis 4-5
4.4.1.1 Freshwater Sediments 4-5
4.4.1.2 Saltwater Sediments 4-7
4.5 Bioaccumulation 4-7
Section 5
Sampling and Analytical Chemistry 5-1
5.1 General Information 5-1
5.2 Sampling and Storage 5-1
5.2.1 Sediments 5-2
5.2.2 Interstitial Water 5-2
5.3 Analytical Measurements 5-3
5.3.1 Acid Volatile Sulfide 5-4
5.3.2 Simultaneously Extracted Metals 5-4
5.3.3 Total Organic Carbon 5-4
5.3.4 Interstitial Water Metal 5-4
Section 6
Sediment Benchmark Values:
Application and Interpretation 6-1
6.1 AVS ESB 6-1
6.2 Interstitial Water ESB 6-1
Section 7
References 7-1
Appendix A A-I
Appendix B B-I
Appendix C c-i
Appendix D D-I
VI
-------
Contents
Tables
Table 2-1. Cadmium binding capacity and AVS of sediments 2-11
Table 2-2. Metal sulfide solubility products and ratios 2-13
Table 3 -1. Toxicity of sediments from freshwater and saltwater lab-spiked sediment tests, field locations,
and combined lab-spiked and field sediment tests 3-8
Table 3-2. Summary of the results of full life-cycle and colonization toxicity tests conducted in the
laboratory and field using sediments spiked with individual metals and metal mixtures 3-15
Table 3 -3. Test-specific data for chronic toxicity of freshwater and saltwater organisms compared to
(ZSEM-AVS)//OC 3-26
Table 4-1. Water quality criteria (WQC) final chronic values (FCV) based on the
dissolved concentration of metal 4-2
Table 4-2. Example calculations of ESB .,7C,,,,& for metal mixtures: three sediments 4-4
AV a: W CjL,
Figures
Figure 2-1. Acute toxicity to grass shrimp (Palaemonetes pugio) of total cadmium and cadmium activity
with different concentrations of the complexing ligands NTA and chloride as salinity 2-2
Figure 2-2. Acute toxicity of total copper and copper activity to the dinoflagellate Gonyaulax tamarensis
with and without the complexing ligand EDTA 2-3
Figure 2-3. Specific growth rates of a diatom (Thalassiosira pseudonana) and a unicellular algae
(Monochrysis lutheri) versus total copper and copper activity for a range of concentrations
of the complexing ligands Tris and natural DOC in river water 2-4
Figure 2-4. Copper accumulation in oysters (Crassostrea virginica) versus total copper and copper
activity with different levels of the complexing ligand NTA 2-5
Figure 2-5. Mean survival of the amphipod Rhepoxynius abronius versus dissolved cadmium concentration
for 4-day toxicity tests in seawater and 0- and 4-day tests in interstitial water 2-6
Figure 2-6. Mortality versus interstitial water cadmium activity for sediments from Long Island Sound,
Ninigret Pond, and a mixture of these two sediments 2-7
Figure 2-7. Toxicity of copper to Hyalella azteca versus copper concentrations in a water-only exposure
and interstitial water copper concentrations in sediment exposures using Keweenaw
Watershed sediments 2-7
Figure 2-8. Cadmium titrations of amorphous FeS 2-9
Figure 2-9. Concentrations of ionic iron and cadmium in the supernatant from titration of FeS by Cd2+ 2-10
Figure 2-10. Cadmium titration of sediments from Black Rock Harbor, Long Island Sound, Hudson River,
and Ninigret Pond 2-12
Vll
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Figure 3-1. Percentage mortality of amphipods (Ampelisca abdita andRhepoxynius hudsoni) exposed to
sediments from Long Island Sound, Ninigret Pond, and a mixture of these two sediments as a
function of the sum of the concentrations of metals in sediments expressed as dry weight,
interstitial water cadmium activity, and the sediment cadmium/AVS ratio 3-3
Figure 3-2. Concentrations of individual metals in interstitial water of sediments from Long Island Sound
and Ninigret Pond in the mixed metals experiment as a function of SEM/AVS ratio 3-4
Figure 3-3. Percentage mortality of freshwater and saltwater benthic species in 10-day toxicity tests in
sediments spiked with individual metals (Cd, Cu, Pb, Ni, Ag, or Zn) or a metal mixture
(Cd, Cu, Ni, andZn) 3-6
Figure 3-4. Percentages of the 184 spiked sediments from Figure 3-3 that were nontoxic or toxic over
various intervals of concentrations of metal based on sediment dry weight (/-onol/g), IWTU,
and SEM/AVS 3-7
Figure 3-5. Percentage mortality of amphipods, oligochaetes, and polychaetes exposed to sediments from
four freshwater and three saltwater field locations as a function of the sum of the molar
concentrations of SEM minus the molar concentration of AVS (SEM-AVS) 3-11
Figure 3-6. Percentage mortality of freshwater and saltwater benthic species in 10-day toxicity tests in
spiked sediments and sediments from the field 3-12
Figure 3 -7. Comparison of the chronic toxicity of sediments spiked with individual metals or metal
mixtures to predicted toxicity based on SEM-AVS 3-18
Figure 3-8. Percent mortality versus SEM-AVS and (2SEM-AVS)//OC for saltwater field data without Bear
Creek and Jinzhou Bay, freshwater field data, freshwater spiked data, and saltwater spiked data . 3-21
Figure 3-9. Percent mortality versus (SEMMetal-AVS)//oc for each metal in spiked sediment tests using
Ampelisca, Capitella, Neanthes, Lumbriculus, andHelisoma 3-23
Figure 3-10. Percent mortality versus (SEMAg-AVS)//oc for silver and (2SEM-AVS)//OC for a mixture
experiment using Cd, Cu, Ni, andZn 3-24
Figure 3-11. Comparison of the chronic toxicity of sediments spiked with individual metals or metal
mixtures to predicted toxicity based on (SEM-AVS)//OC 3-25
Figure 4-1. SEM-AVS values versus AVS concentrations in EMAP-Great Lakes sediments from
Lake Michigan. Plot (A) shows all values; plot (B) has the ordinate limited to SEM-AVS
values between-10 and+10^mol/g 4-6
Figure 4-2. SEM-AVS values versus AVS concentrations in EMAP-Estuaries Virginian Province;
REMAP-NY/NJ Harbor Estuary; NOAA NST-Long Island Sound; Boston Harbor;
and Hudson-Raritan Estuaries 4-8
Figure 4-3. (ZSEM-AVS)//OC versus AVS concentrations in EMAP-Estuaries Virginian Province;
REMAP-NY/NJ Harbor Estuary; NOAA NST-Long Island Sound; Boston Harbor; and
Hudson-Raritan Estuaries 4-9
Vlll
-------
Acknowledgements
Acknowledgments
Coauthors
David J. Hansen formerly with U.S. EPA
Dominic M. Di Toro University of Delaware, Newark, DE; HydroQual, Inc., Mahwah, NJ
Walter J. Berry* U.S. EPA, NHEERL, Atlantic Ecology Division, Narragansett, RI
Warren S. Boothman U.S. EPA, NHEERL, Atlantic Ecology Division, Narragansett, RI
Robert M. Burgess** U.S. EPA, NHEERL, Atlantic Ecology Division, Narragansett, RI
Gerald T. Ankley U.S. EPA, NHEERL, Mid-Continent Ecology Division, Duluth, MN
David R. Mount U.S. EPA, NHEERL, Mid-Continent Ecology Division, Duluth, MN
Joy A. McGrath HydroQual, Inc., Mahwah, NJ
Laurie D. De Rosa HydroQual, Inc., Mahwah, NJ
Heidi E. Bell* U.S. EPA, Office of Water, Washington, DC
Mary C. Reiley U.S. EPA, Office of Water, Washington, DC
Christopher S. Zarba U.S. EPA, Office of Research and Development, Washington, DC
Significant Contributors to the Development of the Approach and Supporting Science
Herbert E. Allen University of Delaware, Newark, DE
Gerald T. Ankley U.S. EPA, NHEERL, Mid-Continent Ecology Division, Duluth, MN
Dominic M. Di Toro University of Delaware, Newark, DE; HydroQual, Inc., Mahwah, NJ
David J. Hansen formerly with U.S. EPA
Landis Hare Universite du Quebec, Sainte-Foy, Quebec, Canada
John D. Mahony Manhattan College, Riverdale, NY
Richard C. Swartz formerly with U.S. EPA
Christopher S. Zarba U.S. EPA, Office of Research and Development, Washington, DC
Technical Support and Document Review
Robert M. Burgess U.S. EPA, NHEERL, Atlantic Ecology Division, Narragansett, RI
Patricia A. DeCastro Computer Sciences Corporation, Narragansett, RI
Tyler K. Linton Great Lakes Environmental Center, Columbus, OH
David R. Mount U.S. EPA, NHEERL, Mid-Continent Ecology Division, Duluth, MN
Robert L. Spehar U.S. EPA, NHEERL, Mid-Continent Ecology Division, Duluth, MN
*Principal U.S. EPA contact
** Series Editor
IX
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Executive Summary
This equilibrium partitioning sediment benchmark (ESB) document describes procedures to
derive concentrations of metal mixtures in sediment which are protective of the presence of
benthic organisms. The equilibrium partitioning (EqP) approach was chosen because it
accounts for the varying biological availability of chemicals in different sediments and allows
for the incorporation of the appropriate biological effects concentration U.S. EPA (2003a).
This provides for the derivation of benchmarks that are causally linked to the specific chemical,
applicable across sediments, and appropriately protective of benthic organisms.
Equilibrium partitioning theory predicts that these metals partition in sediment between acid
volatile sulfide (AVS, principally iron monosulfide), interstitial (pore) water, benthic organisms,
and other sediment phases such as organic carbon. Biological responses of benthic organisms
to these metals in sediments are different across sediments when the sediment concentrations
are expressed on a dry weight basis, but similar when expressed on a ZSEM-AVS or interstitial
water basis. The difference between the sum of the molar concentrations of simultaneously
extracted metal (ZSEM, the metal extracted in the AVS extraction procedure) minus the molar
concentration of AVS accurately predicts which sediments are not toxic because of these metals.
The use of (ZSEM-AVS)//OC reduces variability associated with prediction of when sediments
will be toxic.
EqP can be used to calculate ESBs for any toxicity endpoint for which there are water-only
toxicity data; it is not limited to any single effect endpoint. For the purposes of this document,
the ESB for mixtures of the metals cadmium, copper, lead, nickel, silver, and zinc is based on
the solid phase and interstitial water phase of sediments. In sediments, these metals should not
cause direct toxicity to benthic organisms if the ZSEM-AVS is <0.0. Alternatively, sediments
containing these metals should not cause direct toxicity to benthic organisms if the sum of the
dissolved interstitial water concentrations for each of the metals (2M. d) divided by their
respective water quality criteria final chronic value (FCV) is <1.0. Uncertainty bounds on
ZSEM-AVS and (ZSEM-AVS)//OC can be used to identify sediments where toxicity, because of
these metals, is unlikely, uncertain, or likely. If an FCV is not available, a secondary chronic
value (SCV) can be substituted. Ancillary analyses conducted as part of this derivation suggest
that the sensitivity of benthic/epibenthic organisms is not significantly different from pelagic
organisms; for this reason, the FCV and the resulting ESBAVS.WQC should be fully applicable to
benthic organisms. The ESBAVS.WQCs should be interpreted as chemical concentrations below
which adverse effects are not expected. At concentrations above the ESBAVS.WQCs, effects may
occur with increasing severity as the degree of exceedance increases. In principle, above the
upper confidence limit effects are expected if the chemical is bioavailable as predicted by EqP
theory. A sediment-specific site assessment would provide further information on chemical
bioavailability and the expectation of toxicity relative to the ESBAVS.WQCs and associated
uncertainty limits. An appendix addresing chromium toxicity in sediments is also included in
this document.
As discussed, while this document uses the WQC or AVS values, the EqP methodology can be
used by environmental managers to derive a benchmark with any desired level of protection, so
long as the water-only concentration affording that level of protection is known. Therefore, the
resulting benchmark can be species or site-specific if the corresponding water-only information
is available. For example, if a certain water-only effects concentration is known to be
XI
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
economically important benthic species, the FCV or SCV for that benthic species could be used
to derive the benchmark. Such a benchmark might be considered as providing "site-specific
protection" for a species or endpoint, if the goal is to derive a benchmark for that particular
site or species. Another way to make an ESB site-specific would be to incorporate information
on unusual partitioning, if suspected, at the site (see U.S. EPA 2003b).
The ESBs do not consider the antagonistic, additive or synergistic effects of other sediment
contaminants in combination with metal mixtures or the potential for bioaccumulation and
trophic transfer of metal mixtures to aquatic life, wildlife or humans. Consistent with the
recommendations of EPA's Science Advisory Board, publication of these documents does not
imply the use of ESBs as stand-alone, pass-fail criteria for all applications; rather, ESB
exceedances could be used to trigger the collection of additional assessment data. When
using the AVS approach, the ESBAVS.WQC applies to sediments having AVS concentrations
£ 0.1 jjmol/g.
Tier 1 and Tier 2 ESB values were developed to reflect differing degrees of data availability
and uncertainty. Tier 1 ESBs have been derived for metal mixtures in this document, and for
the nonionic organic insecticides endrin and dieldrin, and polycyclic aromatic hydrocarbon
(PAH) mixtures in U.S. EPA (2003c, d, e). Tier 2 ESBs are reported in U.S. EPA (2003f).
xn
-------
Glossary
Glossary of Abbreviations
Ag
Ag2S
AVS
ccc
Cd
{CcP}
[CcP]
[Cd]B
[CdS(s)]
Cr
Cu
CWA
DOC
EDTA
EMAP
EPA
EqP
ESB(s)
ESBAVS:WQC
Joe
FCV
Fe
{Fe2+}
[Fe2+]
[FeS(s)]
Silver
Silver monosulfide
Acid volatile sulfide
Criteria continuous concentration
Cadmium
Activity of ionic cadmium (mol/L)
Concentration of ionic cadmium (mol/L)
Concentration of added cadmium (mol/L)
Concentration of bound cadmium (mol/L)
Concentration of solid-phase cadmium sulfide (mol/L)
Chromium
Concentration of contaminant in sediment
Sediment LC50 Concentration
Copper
Clean Water Act
Dissolved organic carbon
Ethlyenediaminetetra-acetic acid
Environmental Monitoring and Assessment Program
U.S. Environmental Protection Agency
Equilibrium partitioning
Equilibrium partitioning sediment benchmark(s)
Equilibrium partitioning sediment benchmark(s) for metal mixtures based
on the Water Quality Criteria Final Chronic Values or Acid Volatile Sulfide
Fraction of organic carbon in sediment
Final chronic value
Iron
Activity of ionic iron (mol/L)
Concentration of ionic iron (mol/L)
Concentration of solid-phase iron sulfide (mol/L)
Xlll
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
[FeS(s)]i Concentration of initial solid-phase iron sulfide (mol/L)
FeS Iron monosulfide
GFAA Gas Furnace Atomic Absorption Spectrophotometry
HECD U.S. EPA, Health and Ecological Criteria Division
IW Interstitial water
IWBU Interstitial water benchmarks unit
IWTU Interstitial water toxic unit
^FeS Solubility product for FeS(s) [(mol/L)2]
^MS Solubility product for MS(s) [(mol/L)2]
Koc Organic carbon-water partition coefficient
Kp Sediment-interstitial water partition coefficient
Ksp Solubility product constant
LC50 Concentration estimated to be lethal to 50% of the test organisms within
a specified time period
M2+ Divalent metalcadmium, copper, lead, nickel, silver, or zinc
MOH+ Metal hydroxide
MS Metal sulfide
Mn Manganese
{M2+ } Divalent metal activity (mol/L)
[M2+] Concentration of ionic metal (mol/L)
[M] A Concentration of added metal (mol/L)
[M] B Concentration of bound metal (mol/L)
[Md] Dissolved metal concentration in the interstitial water
[MS(s)] Concentration of solid-phase metal sulfide (mol/L)
[MT] Total cold extractable metal (mol/L)
NA Not applicable, not available
NAS National Academy of Sciences
Ni Nickel
NOAA National Oceanographic and Atmospheric Administration
NOEC No observed effect concentration
NST National Status and Trends monitoring program
NTA Nitrilotriacetic acid
xiv
-------
Glossary
NTIS
Pb
OEC
ORD
OST
POC
REMAP
s2-
{s2-}
[S2-]
SAB
SD
SEM
[SEMJ
[SEMcd]
[SEMCJ
[SEMpb]
TIE
TOC
WQC
Zn
[ZCd(aq)]
[ZFe(aq)]
[ZM(aq)]
[ZS(aq)]
National Technical Information Service
Lead
Observed effect concentration
U.S. EPA, Office of Research and Development
U.S. EPA, Office of Science and Technology
Paniculate organic carbon
Regional Environmental Monitoring and Assessment Program
Sulfide ion
Activity of sulfide (mol/L)
Concentration of sulfide (mol/L)
U.S. EPA Science Advisory Board
Standard deviation
Simultaneously extracted metals
Simultaneously extracted metals, concentration of the combined metals
Simultaneously extracted metals, Cd concentration (^mol/g)
Simultaneously extracted metals, Cu concentration (^mol/g)
Simultaneously extracted metals, Pb concentration (^mol/g)
Simultaneously extracted metals, Ni concentration (^mol/g)
Simultaneously extracted metals, Ag concentration (^mol/g)
Simultaneously extracted metals, Zn concentration (^mol/g)
Toxicity identification evaluation
Total organic carbon
Water quality criteria
Zinc
Concentration of total dissolved Cd2+ (mol/L)
Concentration of total dissolved Fe2+ (mol/L)
Concentration of total dissolved M2+ (mol/L)
Concentration of total dissolved S2" (mol/L)
XV
-------
Eq
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
xvi
-------
Section 1
Introduction
Introduction
1.1 General Information
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 (Larsson, 1985; Salomons etal., 1987;
Burgess and Scott, 1992). The absence of
defensible equilibrium partitioning sediment
benchmarks (ESBs) make it difficult to accurately
assess the extent of the ecological risks of
contaminated sediments and to identify, prioritize,
and implement appropriate cleanup activities and
source controls (U.S. EPA 1997a, b, c).
As a result of the need for a procedure to
assist regulatory agencies in making decisions
concerning contaminated sediment problems, the
U.S. Environmental Protection Agency (EPA)
Office of Science and Technology, Health and
Ecological Criteria Division (OST/HECD) and
Office of Research and Development National
Health and Environmental Effects Research
Laboratory (ORD/NHEERL) established a
research team 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 the
derivation of benchmarks in all situations (U.S.
EPA, 1989, 1992). The equilibrium partitioning
(EqP) approach was selected for nonionic organic
chemicals because it presented the greatest
promise for generating defensible, national,
numeric chemical-specific benchmarks applicable
across a broad range of sediment types. The
three principal observations that underlie the EqP
approach to establishing sediment benchmarks are
as follows:
1. The concentrations of nonionic organic
chemicals in sediments, expressed on an organic
carbon basis, and in interstitial waters correlate to
observed biological effects on sediment-dwelling
organisms across a range of sediments.
2. Partitioning models can relate sediment
concentrations for nonionic organic chemicals on
an organic carbon basis to freely-dissolved
concentrations in interstitial water.
3. The distribution of sensitivities of benthic
organisms to chemicals is similar to that of water
column organisms; thus, the currently established
water quality criteria (WQC) final chronic values
(FCV) or secondary chronic values (SCV) can be
used to define the acceptable effects concentration
of a chemical freely-dissolved in interstitial water.
Because of their widespread release and
persistent nature, metals such as cadmium, copper,
lead, nickel, silver, and zinc are commonly elevated
in aquatic sediments. These metals, in addition to
nonionic organic chemicals, are of potential
concern to aquatic environments. Thus, there
have been various proposals for deriving sediment
benchmarks for protecting benthic communities
using measurement of total sediment metals
followed by comparison with background metal
concentrations, or in some cases, an effects-based
endpoint (Sullivan etal., 1985; Persaud etal., 1989;
Long and Morgan, 1990; Ingersoll etal., 1996;
MacDonaldetal., 1996). An important limitation
to these types of approaches is that the causal
linkage between the measured concentration of
metals and the observed toxicity cannot be
established, in part because of the procedures used
to derive correlative values, and because values
derived are based on total rather than bioavailable
metal concentrations. That is, for any given total
metal concentration, adverse toxicological effects
may or may not occur, depending on the
physicochemical characteristics of the sediment of
concern (Tessier and Campbell, 1987; Luoma,
1989; DiToro etal., 1990).
1-1
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Many researchers have used elaborate
sequential extraction procedures to identify
sedimentary physicochemical fractions with which
metals are associated in an attempt to understand
the biological availability of metals in sediments
(Tessier et al., 1979; Luoma and Bryan, 1981).
Key binding phases for metals in sediments
included iron and manganese oxides and organic
carbon. Shortcomings with these approaches have
limited their application largely to aerobic
sediments instead of anaerobic sediments, where
metals are often found in the greatest
concentrations (see Section 2).
In developing ESBs for metals that causally
link metals concentrations to biological effects and
that apply across all sediments, it is essential that
bioavailability be understood. Therefore, the EqP
approach was selected as the technical basis for
deriving ESBs for metals. Different studies have
shown that although total (dry weight) metal
concentrations in anaerobic sediments are not
predictive of bioavailability, metal concentrations in
interstitial water are correlated with observed
biological effects (Swartz et al., 1985; Kemp and
Swartz, 1986). However, as opposed to the
situation for nonionic organic chemicals and
organic carbon (see Di Toro et al., 1991), sediment
partitioning phases controlling interstitial water
concentrations of metals were not readily
apparent. A key partitioning phase controlling
cationic metal activity and metal-induced toxicity in
the sediment-interstitial water system is acid
volatile sulfide (AVS) (Di Toro et al., 1990,1992).
AVS binds, on a molar basis, a number of cationic
metals of environmental concern (cadmium,
copper, lead, nickel, silver, and zinc), forming
insoluble sulfide complexes with minimal biological
availability. (Hereafter in this document, the use
of the term "metals" will apply only to these six
metals.)
The data that support the EqP approach for
deriving sediment benthmarks for nonionic organic
chemicals were reviewed by Di Toro et al. (1991)
and U.S. EPA (1997a; 2003a). The utility of the
EqP approach for deriving sediment benchmarks
for metals (U.S. EPA, 1994a) was reviewed and
endorsed by EPA's Science Advisory Board
(SAB) in 1994 and 1999 (U.S. EPA, 1995a, 1999).
The data that support the EqP approach for
deriving sediment benchmarks for metals
presented in this document were taken largely
from a series of papers published in the December
1996 issue of Environmental Toxicology and
Chemistry by Ankley et al. (1996), Berry et al.
(1996), DeWitt et al. (1996), Di Toro et al.
(1996a,b), Hansen et al. (1996a,b), Leonard et al.
(1996a), Liber et al. (1996), Mahony et al. (1996),
Peterson et al. (1996), and Sibley et al. (1996). In
addition, publications by Di Toro et al. (1990,
1992), Ankley etal. (1994), U.S. EPA(1995a), and
Berry et al. (1999) were of particular importance
in the preparation of this document.
The same three general principles observed in
applying the EqP approach to nonionic organic
chemicals listed above also apply with only minor
adjustments to deriving ESBs for mixtures of the
cationic metalscadmium, copper, lead, nickel,
silver, and zinc:
1. The concentrations of these six metals in
sediments, normalized to the concentration of
AVS and simultaneously extracted metals
(SEM) (the metals extracted with AVS) in
sediments and dissolved in interstitial waters,
correlate with observed biological effects to
sediment-dwelling organisms across a range of
sediments (Di Toro et al., 1992).
2. Partitioning models can relate sediment
concentrations for cationic divalent metals
(and monovalent silver) on an AVS basis to the
absence of freely-dissolved concentrations in
interstitial water.
3. The distributions of sensitivities of benthic and
water column organisms to organic chemicals
and metals are similar (U.S. EPA, 2003a);
thus, the currently established WQC FCVs
can be used to define the acceptable effects
concentration of the metals freely dissolved in
interstitial water.
The EqP approach, therefore, assumes that (1)
the partitioning of the metal between sediment
AVS (or any other binding factors controlling
1-2
-------
Introduction
bioavailability) and interstitial water approximates
equilibrium; (2) organisms receive equivalent
exposure from interstitial water-only exposure or
from exposure to any other equilibrated sediment
phase: either from interstitial water via respiration,
sediment via ingestion, or sediment-integument
exchange, or from a mixture of exposure routes;
(3) for the cationic metals cadmium, copper, lead,
nickel, zinc, and silver, partitioning of metal
between the solid phase and interstitial water can
be predicted based on the relative concentrations
of AVS and SEM; (4) the WQC FCV
concentration is an appropriate effects
concentration for freely-dissolved metal in
interstitial water; and (5) the toxicity of metals in
interstitial water is no more than additive.
For the first time, the Agency is publishing
ESBs that account for bioavailability in sediments
and the potential for effects of a metal mixture in
the aquatic environment, thus providing an
ecologically relevant benchmark. Two equally
applicable ESBs for metals, a solid phase and an
interstitial water phase, are described. The solid-
phase AVS ESBs is defined as the S.[SEM.] <
[AVS] (total molar concentration of simultaneously
extracted metal is less than or equal to the total
molar concentration of acid volatile sulfide). Note
that cadmium, copper, lead, nickel, and zinc are
divalent metals so that one mole of each metal can
bind only with one mole of AVS. The molar
concentrations of these metals are compared with
AVS on a one-to-one basis. Silver, however, exists
predominantly as a monovalent metal, so that silver
monosulfide (Ag2S) binds two moles of silver for
each mole of AVS. Therefore, SEM, by
Ag J
convention will be defined as the molar
concentration of silver divided by two, [Ag]/2,
which is compared with the molar AVS
concentration. The interstitial water phase ESB is
S[Mid]/[FCVid] <1 (the sum of cadmium, copper,
nickel, lead, and zinc of the concentration of each
individual metal dissolved in the interstitial water
divided by the metal-specific FCV based on
dissolved metal is less than or equal to one; note
that at present EPA does not have an FCV for
silver). This latter value is termed an interstitial
water benchmark unit (IWBU). A requirement of
the IWBUapproach is that the toxicities of
interstitial water metal concentrations be additive.
The data presented in this document support the
additivity of the toxicity of metal mixtures in water.
Importantly, both the solid-phase AVS ESB
and interstitial water ESB are no-effect
benchmarks; that is, they predict sediments that
are acceptable for the protection of benthic
organisms. These ESBs, when exceeded, do not
unequivocally predict sediments that are
unacceptable for the protection of benthic
organisms. The solid-phase AVS benchmark
avoids the methodological difficulties of interstitial
water sampling that may lead to an overestimate
of exposure and provides information on the
potential for additional metal binding. Because the
AVS benchmark does not include other metal-
binding phases of sediments, the interstitial
benchmark is also proposed. The use of both the
AVS and interstitial water benchmarks will
improve estimates of risks of sediment-associated
metals. For example, the absence of significant
concentrations of metal in interstitial water in toxic
sediments having SEMAVS demonstrates that
metals in these sediments are unavailable. The
(SSEM-AVS)//OC correction, although not an ESB,
can be used to refine the prediction of sediments
where protection of benthic organisms is
acceptable, uncertain, or unacceptable.
ESBs based on the EqP approach are
developed using the latest available scientific data
and are suitable for providing guidance to
regulatory agencies because they are
Numeric values
Chemical-specific
Applicable to most sediments
Predictive of biological effects
Protective of benthic organisms
It should be emphasized that these
benchmarks are intended to protect benthic
organisms from the direct effects of these six
metals in sediments that are permanently
inundated with water, intertidal, or inundated
periodically for durations sufficient to permit
1-3
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
development of benthic assemblages. They do not
apply to occasionally inundated soils containing
terrestrial organisms. The ESBs do not consider the
antagonistic, additive or synergistic effects of other
sediment contaminants in combination with metal
mixtures or the potential for bioaccumulation and
trophic transfer of metal mixtures to aquatic life, wildlife
or humans. The ESBs presented in this document
are the recommended concentrations of cadmium,
copper, lead, nickel, silver, and zinc in sediment
that will not adversely affect most benthic
organisms. ESB values may be adjusted to
account for future data or site-specific
considerations (U.S. EPA, 2003b).
This document includes the theoretical basis
and the supporting data relevant to the derivation
of an ESBfor cadmium, copper, lead, nickel, silver,
and zinc and their mixture. 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, 1985a); "Ambient Water Quality Criteria for
Cadmium" (U.S. EPA, 1985b); "Ambient Water
Quality Criteria for Copper" (U.S. EPA, 1985c);
"Ambient Water Quality CriteriaSaltwater
Copper Addendum" (U.S. EPA, 1995c); "Ambient
Water Quality Criteria for Lead" (U.S. EPA,
1985d); "Ambient Water Quality Criteria for
Nickel" (U.S. EPA, 1986); "Ambient Water
Quality Criteria for Silver" (U.S. EPA, 1980); and
"Ambient Water Quality Criteria for Zinc" (U.S.
EPA, 1987) is necessary in order to understand the
following text, tables, and calculations.
1.2 Applications of Sediment Benchmarks
ESBs are meant to be used with direct toxicity
testing of sediments as a method of evaluation
assuming the toxicity testing species is sensitive to
the chemical of interest. They provide a chemical -
by-chemical specification of what sediment
concentrations are protective of benthic aquatic
life. The EqP method should be applicable to
nonionic organic chemicals with aKow above 3.0.
Examples of other chemicals to which this
methodology applies include endrin, dieldrin, and
poly cyclic aromatic hydrocarbon (PAH) mixtures.
For the toxic chemicals addressed by the ESB
documents Tier 1 (U.S. EPA, 2003c, d, e, and this
document) and Tier 2 (U.S. EPA, 2003f) values
were developed to reflect the differing degrees of
data availability and uncertainty. Tier 1 ESBs are
more scientifically rigorous and data intensive than
Tier 2 ESBs. The minimum requirements to derive
a Tier 1 ESB include: (1) Each chemical's organic
carbon-water partition coefficient (Koc) is derived
from the octanol-water partition coefficient (KQW)
obtained using the SPARC (SPARC Performs
Automated Reasoning in Chemistry) model
(Karickhoff et al., 1991) and the KOW-KOC
relationship from Di Toro et al. (1991). This KQC
has been demonstrated to predict the toxic
sediment concentration from the toxic water
concentration with less uncertainty than 1C
^ UL-
values derived using other methods. (2) The FCV
is updated using the most recent toxicological
information and is based on the National WQC
Guidelines (Stephan et al., 1985). (3) EqP-
confirmation tests are conducted to demonstrate
the accuracy of the EqP prediction that the KQC
multiplied by the effect concentration from a
water-only toxicity test predicts the effect
concentration from sediment tests (Swartz, 1991;
DeWittetal, 1992). Using these specifications,
Tier 1 ESBs have been derived for metal mixtures
in this document, the nonionic organic insecticides
endrin and dieldrin (U.S. EPA, 2003c, d) and PAH
mixtures (U.S. EPA, 2003e). In comparison, the
minimum requirements for a Tier 2 ESB (U.S.
EPA, 2003f) are less rigorous: (1) The Kow for the
chemical that is used to derive the Koc can be
from slow-stir, generator column, shake flask,
SPARC or other sources. (2) FCVs can be from
published or draft WQC documents, the Great
Lakes Initiative or developed from AQUIRE.
Secondary chronic values (SCV) from Suter and
Mabrey (1994) or other effects concentrations
from water-only toxicity tests can be used. (3)
EqP cconfirmation tests are recommended, but are
not required for the development of Tier 2 ESBs.
Because of these lesser requirements, there is
greater uncertainty in the EqP prediction of the
sediment effect concentration from the water-only
effect concentration, and in the level of protection
afforded by Tier 2 ESBs. Examples of Tier 2
ESBs for nonionic organic chemicals are found in
U.S. EPA (2003f).
1-4
-------
Introduction
1.3 Overview
Section 1 provides a brief review of the EqP
methodology as it applies to the individual metals
cadmium, copper, lead, nickel, silver, and zinc and
their mixture. Section 2 reviews published
experimental results that describe the toxicity
associated with the partitioning and bioavailability
of these metals in interstitial water of freshwater
and marine sediments. Section 3 reviews the
results of acute and chronic toxicity tests
conducted with spiked and field sediments that
demonstrate that the partitioning and bioavailability
of metals in sediments can be used to accurately
predict the absence of toxicity of sediment-
associated metals. Section 4 describes the AVS
benchmark and interstitial water benchmark
approaches for the derivation of the ESB for
individual metals and mixtures of metals.
Published WQC values for five of these six
dissolved metals (the silver FCV is not available)
are summarized for use in calculating IWBUs as
required in the interstitial water ESB approach.
The ESBAVS.WQC for metals is then compared with
chemical monitoring data on environmental
occurrence of SEM, AVS, and interstitial metals in
sediments from Lake Michigan, the Virginian
Province from EPA's Environmental Monitoring
and Assessment Program (EMAP), and the
National Oceanic and Atmospheric Administration
(NOAA) National Status and Trends monitoring
program (NST). Section 5 describes
recommended procedures for sampling, handling,
and analysis of metals in sediments and
interpretation of data from the sediment samples
that is needed if the assessments of risks of
sediment-associated metals are to be appropriately
based on the EqP methodology. Section 6
concludes with the ESBAVS for a mixture of
the metals: cadmium, copper, nickel, lead, silver,
and zinc and discussion of their application and
interpretation. The references cited in this
document are listed in Section 7. Appendices A
and B provide additional monitoring data.
Appendix C reports on quality assurance for this
document and Appendix D addresses chromium
toxicity in sediments.
1-5
-------
Partitioning
Section 2
Partitioning of Metals in Sediments
2.1 Metal Toxicity in Water-Only and
in Interstitial Water of Sediment
Exposures
The EqP approach for establishing sediment
benchmarks (i.e., ESBs) requires that the
chemicals be measured in phases that relate to
chemical activity in sediment. The information
provided in this section demonstrates that
biological effects correlate to metal activity. Also,
it demonstrates that biological response in
sediment exposures is the same as in water-only
exposures when sediment exposure is assessed on
the basis of interstitial water concentrations. This
is fundamental to satisfying the EqP approach for
both metals and nonionic organic chemicals.
A direct method for establishing sediment
benchmarks for metals would be to apply the
WQC FCV to measured interstitial water
concentrations. The validity of this approach
depends both on the degree to which the
interstitial water concentration represents free
metal activity, and on whether free metal activity
can be accurately measured in surface waters and
water-only toxicity tests used to derive WQC, and
in interstitial water of field sediments and
sediments spiked with metals in the laboratory.
For most metals, free metal activity cannot be
directly measured at WQC concentrations.
Therefore, present WQC are not based on free
metal activity; rather, they are based on dissolved
metals. However, many dissolved metals readily
bind to dissolved (actually colloidal) organic
carbon (DOC) forming complexes that do not
appear to be bioavailable (Bergman and Dorward-
King, 1997). Hence, sediment guidelines or
benchmarks based on interstitial water
concentrations of metals may be overly protective
in cases where not all dissolved metal is
bioavailable.
By implication, this difficulty extends to any
complexing ligand that is present in sufficient
quantity. Decay of sediment organic matter can
cause substantial changes in interstitial water
chemistry. In particular, bicarbonate increases
because of sulfate reduction, which increases the
importance of metal-carbonate complexes and
further complicates the question of the bioavailable
metal species (Stumm and Morgan, 1996).
Sampling sediment interstitial water for metals
is not a routine procedure. The least invasive
technique employs a diffusion sampler that has
cavities covered with a filter membrane (Hesslein,
1976;Carignan, 1984; Carignanetal., 1985; Allen
etal., 1993; Bufflap and Allen, 1995). The
sampler is inserted into the sediment and the
concentrations on either side of the membrane
equilibrate. Because the sampler is removed after
equilibration, the concentrations of metals inside
the sampler should be equal to the concentrations
of freely-dissolved metals in the interstitial water.
The time required for equilibration, typically
several days, depends on the size of the filter
membrane and the geometry of the cavity.
An alternative technique for separating
interstitial water is to obtain an undisturbed
sediment sample as a whole sediment or core that
can be sliced for vertical resolution, filter or
centrifuge the sample, and then filter the resultant
interstitial water twice. For anaerobic sediments,
this must be done in a nitrogen atmosphere to
prevent precipitation of iron hydroxide, which
would scavenge the metals and yield artificially
low dissolved concentrations of metals (Troup,
1974; Allen etal., 1993).
Although either technique is suitable for
research investigations, they require more than the
normally available sampling capabilities. If solid-
phase chemical measurements were available
2-1
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
from which interstitial water metal activity could
be deduced, this would obviate the need for
interstitial water sampling and analysis, circumvent
the need to deal with complexing ligands, and
provide fundamental insight into metal-binding
phases in sediments needed to predict
bioavailability. The recommended procedures for
suitable sampling, handling, and analytical
techniques for interstitial water and sediments are
provided in Section 5 of this document.
2.1.1 Toxicity Correlates to Metal Activity
A substantial number of water-only exposures
indicate that biological effects can be correlated to
divalent metal activity {M2+}. Although other
forms of metal may also be bioavailable (e.g.,
MOH+), DOC and certain other ligand-complexed
fractions of the metal render it unavailable to
organisms. Results from some of these exposures
are summarized below.
Acute toxicity of various concentrations of
cadmium to grass shrimp (Palaemonetes pugio)
has been determined in water containing the
complexing ligand nitrilotriacetic acid (NTA) or
chloride (as salinity), each of which forms
cadmium complexes (Sunda et al., 1978). The
concentration response curves as a function of
total cadmium are quite different at varying
concentrations of NTA and chloride (Figure 2-1, A
and B). However, if the organism response is
evaluated with respect to measured Cd2+ activity, a
single concentration-response relationship results
(Figure 2-1, C and D). Comparable results have
been reported by Anderson and Morel (1978) for
the dinoflagellate Gonyaulax tamarensis exposed
to copper-ethylene diamine tetra-acetic acid
(EDTA) complexes (Figure 2-2, A and C).
Likewise, Allen et al. (1980) observed that when
the concentration of zinc is held constant and the
concentration of the complexing ligand NTA is
varied, growth (cells/mL) ofMicrocystis
aeruginosa decreases as the addition of NTA is
100
s/j
5°
Iff*
5.0 6.0
Total Cadmium (-log CdT)
100
50
0
B
Salinity (%o)
AA 4.8 + 0.4
8.4 + 0.2
TT 16.3 + 03
20.0 + 0.3
4» 28.9 + 0.6
4 5
Total Cadmium (-log CdT)
100
80
60
40
20
0
- c
6.0 7.0
Cadmium Activity (p[Cd24])
100
so
60
40
20
D
Salinity (%o)
A 4.8
8.4
T 16.S
20.0
28.9
6.0 7.0
Cadmium Activity (p[Cd2+])
Figure 2-1. Acute toxicity to grass shrimp (Palaemonetes pugio) of total cadmium (top) and cadmium
activity (bottom) with different concentrations of the complexing ligands NTA (left) and
chloride as salinity (right) (figures from Sunda et al., 1978).
2-2
-------
Partitioning
increased (Figure 2-2B). The authors correlated
the effect to free zinc activity as shown in Figure
2-2D. A single concentration-response
relationship is shown for the diatom,
Thalassiosira pseudonana, and the unicellular
alga, Monochrysis lutheri, exposed to copper and
the complexing ligand Tris (Sunda and Guillard,
1976) as well as copper and DOC from natural
river water (Sunda and Lewis, 1978) when
exposure concentration is expressed as metal
activity (Figure 2-3, A, B, C, and D, respectively).
Metal bioavailability, as measured by metal
accumulation into tissues of organisms, has also
100
90
80
70
60
50
40
30
20
10
0
100
90
80
70
60
50
40
30
20
10
0
With EDTA
Without EDTA
56789
Total Copper (-log (CuT))
.
s
z
9 10 11 12 13 14
Copper Activity (pCu)
10
7.
6.
10
in_l
10 -
10 4H
10
o AAP w/l:l M Fe:EDTA
w/o T.M.
4.8x10-7
+ 1.0 x 10-7
+ 2.5 x 10-7
+ 4.0x10-7
+ 6.0x 10-7
+ 1.0x10-6
+ 5.0x10-6
+ 1.0 x 10-5
MZn
MNTA
MNTA
MNTA
MNTA
MNTA
MNTA
MNTA
12
Days
16
20
10
.6-
« 10
C«
U
(M
o
5-
10
10
D
K'atpHV.O
D -EDTA 13.46
A -NTA 7.64
X -ODS 6.31
O -CMOS 5.25
- Builder M 4.33
* - Control
1.0 2.0 3.0 4.0 5.0
Free Zinc (M/L x 107)
Figure 2-2. Acute toxicity of total copper (A) and copper activity (C) to the dinoflagellate Gonyaulax
tamarensis with and without the complexing ligand EDTA (figures from Anderson and
Morel, 1978). Toxicity of zinc to Microcystis aeruginosa showing growth of cells/mL versus
time with different levels of the complexing ligands EDTA and NTA (B) and number of cells
at 5 days as a function of free zinc concentration (D) (figures from Allen et al., 1980).
2-3
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
^
s >,
ii
o **
« a
fa .a
*gig2 ma* :
: 0.5 mM tris pH 7.7
k 1.0 nlOmMtris
fe
2.00
10% River Water
30% River Water
90% River Water
4 5 5.6
Total Copper (-log CuT)
-7.4
4.00
5.00 6.00
Total Copper (-log CuT)
7.00
2.0
II
II
9 10
Copper Activity (pCu)
0.00
D
10% River Water
30% River Water
A 90% River Water
6.00
7.00 8.00
Copper Activity (pCu)
Figure 2-3. Specific growth rates of a diatom (Thalassiosira pseudonana) (left) and a unicellular algae (Monochrysis
lutheri) (right) versus total copper (top) and copper activity (bottom) for a range of concentrations of
the complexing ligands Tris (left; from Sunda and Guillard, 1976) and natural DOC in river water
(right; from Sunda and Lewis, 1978).
been examined (Zamuda and Sunda, 1982).
Uptake of copper by oysters is correlated not to
total copper concentration (Figure 2-4A), but to
copper activity (Figure 2-4B).
The implication to be drawn from these
experiments is that the partitioning model required
for establishing a sediment benchmark should
predict dissolved metal activity in interstitial water,
and that the benchmark based on dissolved metal
would be conservative. The following subsection
examines the utility of this idea.
2.1.2 Toxicity Correlates to Interstitial
Water Concentration
This subsection presents early data that first
indicated the equivalence of interstitial water
concentrations and water-only exposures. Many
more data of this sort are presented in Section 3.
Swartz et al. (1985) tested the acute toxicity of
cadmium to the marine amphipod Rhepoxynius
abronius in sediment and water. An objective of
the study was to determine the contributions of
interstitial and particle-bound cadmium to toxicity.
A comparison of the 4-day LC50 value of
cadmium in interstitial water (1.42 mg/L) with the
4-day LC50 value of cadmium in water without
sediment (1.61 mg/L) indicated no significant
difference between the two (Figure 2-5). The
LC50 represents the chemical concentration
estimated to cause lethality to 50% of the test
organisms within a specified time period.
Experiments were performed to determine the
role of AVS in cadmium-spiked sediments using
the amphipods Ampelisca abdita and
Rhepoxynius hudsoni (Di Toro et al., 1990).
Three sediments were used: a Long Island Sound
sediment with high AVS, aNinigret Pond sediment
with low AVS concentration, and a 50/50 mixture
of the two sediments Figure 2-6 presents a
2-4
-------
Partitioning
150
a
_o
V* ^.
2 £ 100
"a >»
15
u 5
Q> fef) m
Q, ^ 50
O '""'
O
01
0.
200
^ 180
E**
T3 160
U 140
|| 12.
"S * 100
"S *
| "o 80
^^ 60
a 40
o
U
g 20
0
A
D A
0 *
0
.
D *
I o A '
01 0.1 1.0 10.0
Total Copper Concentration (//mol)
_ B Total NTA (//mol)
Winter
A l.o
010.0
-
Summer
1 J. Q T
-i. A 1 n
9 "8.1
4 A *10'°
-
4
1
t
i A ]i I
1 HH AT
^' To
8 9 10 11
Copper Activity (pCu)
Figure 2-4. Copper accumulation in oysters (Crassostrea virginica) versus total copper (A) and copper activity (B)
with different levels of the complexing ligand NTA (figures from Zamuda and Sunda, 1982).
2-5
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
20
^ 15
o
"3
GO
e
2 10
1
1
ft!
5
0
(
_ Seawater
. . \ \ Interstitial
piJL « p -l- -. w&ter
t=4 t=0
\
\ \
* \ \
ill I
J 1 2 3 4 5
Dissolved Cadmium Concentration (mg/L)
Figure 2-5. Mean survival of the amphipod Rhepoxynius abronius versus dissolved cadmium concentration for 4-
day toxicity tests in seawater (symbols) and 0- and 4-day tests (bars) in interstitial water (figure from
Swartz et al., 1985).
comparison of the observed mortality in three
sediments with the interstitial water cadmium
activity measured with a specific ion electrode.
Four-day water-only and 10-day sediment toxicity
tests were performed. The water-only response
data for A. abdita and R. hudsoni are included for
comparison although these data represent a
shorter duration exposure. These experiments
also demonstrate the equivalence of organism
response to metal concentrations in interstitial
water and in water-only exposures.
An elegant experimental design was employed
by Kemp and Swartz (1986) to examine the
relative acute toxicity of particle-bound and
dissolved interstitial cadmium. They circulated
water of the same cadmium concentration through
different sediments. This resulted in different bulk
sediment concentrations, but the same interstitial
water concentrations. They found no statistically
significant difference in organism response for the
different sediments. Because the interstitial water
concentrations were the same in each treatment,
2-6
-------
Partitioning
I
100
so
60
40
20
1
LI Sound
O Mixture
C> Ninigret Pond
Water-Only Exposure
A Rhepoxynius
& Ampelisca
_] ! i i i i
I 1 1
-5.00
-3.00
-1.00
1.00
3.00
Log10CjT Activity (mg/L)
Figure 2-6. Mortality versus interstitial water cadmium activity for sediments from Long Island Sound, Ninigret
Pond, and a mixture of these two sediments. Water-only exposure data are from separate experiments
with both Ampelisca abdita and Rhepoxynius hudsoni. The line is a joint fit to both water-only data
sets (figure from Di Toro et al., 1990).
100
80
g 60
li 40
20
0
1
O O «
O Water-Only Exposure
Sediment Exposure *
.
*
o
i i i i i i i i I i i i i i i i i I i i i i i i i i
10 100 1000
Copper C^g/L)
Figure 2-7. Toxicity of copper to Hyalella azteca versus copper concentrations in a water-only exposure (o) and
interstitial water copper concentrations in sediment exposures () using Keweenaw Watershed
sediments (figure from Ankley et al., 1993).
2-7
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
that is, the circulating water concentrations
established the interstitial water concentrations,
these experiments confirmed the hypothesis of
equal response to concentrations in water-only and
interstitial water.
A series of 10-day toxicity tests using the
amphipod Hyalella azteca was performed to
evaluate bioavailability of copper in sediments
from two sites highly contaminated with this metal:
Steilacoom Lake, WA, and Keweenaw Watershed,
MI (Ankley et al., 1993). A water-only, 10-day
copper toxicity test also was conducted with the
same organism. The mortality resulting from the
water-only test was strikingly similar to that from
the Keweenaw sediment tests when related to
interstitial water (Figure 2-7). The LC50 values
show strong agreement for the water-only (31
jWg/L) and the Keweenaw sediment test (28 ^g/L)
using the average of day 0 and day 10 interstitial
water concentrations. Steilacoom Lake 10-day
interstitial water concentrations were less than the
7 yWg/L detection limit and were consistent with the
observed lack of toxicity to H. azteca.
The data presented in this subsection, and the
data in Section 3, demonstrate that in water-only
exposures, metal activity and concentration can be
used to predict toxicity. The results of the four
experiments above demonstrate that mortality data
from water-only exposures can be used to predict
sediment toxicity using interstitial water
concentrations. Therefore, the metal activity or
dissolved concentration in interstitial water would
be an important component of a partitioning model
needed to establish sediment benchmarks. To
complete the partitioning model, one would need to
identify the solid metal-binding phase(s). The
following subsection presents data that identifies
solid-phase sulfides as the important metal-binding
phase.
2.2 Solid-Phase Sulfide as the Important
Binding Component
Modeling metal sorption to oxides in laboratory
systems is well developed, and detailed models are
available for cation and anion sorption (see Stumm
[1987] and Dzombak and Morel [1990] for
summaries). The models consider surface
complexation reactions as well as electrical
interactions by means of models of the double
layer. Models for natural soil and sediment
particles are less well developed. However,
studies suggest that the models available for cation
and anion sorption can be applied to soil systems
(Allen et al., 1980; Barrow and Ellis, 1986a,b,c;
Sposito etal., 1988). Because the ability to predict
partition coefficients is required if interstitial water
metal concentrations are to be inferred from the
total concentration, some practical model is
required. This subsection presents the state of the
science in theoretical development of metal
partitioning behavior in sediments.
2.2.1 Metal Sorption Ph ases
The initial difficulty selecting an applicable
sorption model is that available models are
complex and many of the parameter estimates
may be specific to individual soils or sediments.
However, the success of nonionic chemical
sorption models based on organic carbon suggests
that some model of intermediate complexity based
on an identification of the dominant sorption
phases may be more generally applicable.
A development in this direction has already
been presented (Jenne et al., 1986; Di Toro et al.,
1987). The basic idea was that instead of
considering only one sorption phase, as is assumed
for nonionic hydrophobic chemical sorption,
multiple sorption phases must be considered. The
conventional view of metal speciation in aerobic
soils and sediments is that metals are associated
with the exchangeable, carbonate and iron (Fe)
and manganese (Mn) oxide forms, as well as
organic matter, stable metal sulfides, and a residual
phase. In oxic soils and freshwater sediments,
sorption phases have been identified as particulate
organic carbon (POC) and the oxides of Fe and
Mn (Jenne, 1968, 1977; Oakley etal., 1980;
Luoma and Bryan, 1981). These phases are
important because they have a large sorptive
capacity. Furthermore, they appear as coatings on
the particles and occlude the other mineral
2-8
-------
Partitioning
components. It was thought that they provided the
primary sites for sorption of metals. These ideas
have been applied to metal speciation in sediments.
However, they ignore the critical importance of
metal sulfide interactions, which dominate
speciation in the anaerobic layers of the sediment.
2.2.2 Jitration Experiments
The importance of sulfide in the control of
metal concentrations in the interstitial water of
marine sediments is well documented (Boulegue et
al., 1982; Emerson etal., 1983; Davies-Colley et
al., 1985; Morse et al., 1987). Metal sulfides are
very insoluble, and the equilibrium interstitial water
metal concentrations in the presence of sulfides
are small. If the interstitial water sulfide
s
g
o
09
09
O.i
0.0
0.5
1.0
1.5
2.0
Cadmium Added (jumol Cd/^mol FeS)
Figure 2-8. Cadmium titrations of amorphous FeS. The x-axis is the amount of cadmium added normalized by
FeS initially present. The y-axis is total dissolved cadmium. The lines connecting the data points are
an aid to visualizing the data. The different symbols represent replicate experiments (figure from Di
Toro et al., 1990).
2-9
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
concentration, S2~, in sediments is large, then the
addition of metal, M2+, to the sediment would
precipitate metal sulfide (MS) following the
reaction
M2+
MS(s)
(2-1)
This appeared to be happening during a spiked
cadmium sediment toxicity test (Di Toro et al.,
1990) because a visible bright yellow cadmium
sulfide precipitate formed as cadmium was added
to the sediment. However, interstitial water sulfide
activity, {S2"}, measured with a sulfide electrode
unexpectedly indicated that there was insufficient
dissolved sulfide present in the unspiked sediment.
The lack of a significant quantity of dissolved
sulfide in the interstitial water and the evident
formation of solid-phase cadmium sulfide
suggested the following possibility. The majority of
the sulfide in sediments is in the form of solid-
phase iron sulfides. Perhaps the source of the
sulfide is from the solid-phase sulfide initially
present. As cadmium is added to the sediment,
CU
u
o
U
-25
0 0.2 0.4 0.6 0.8
1.2 1.4 1.6 1.8
Cd /FeS Molar Ratio
To
U
a ^
O h-1
700
u
a
o
U
Cd /FeS Molar Ratio
Figure 2-9. Concentrations of ionic iron (A) and cadmium (B) in the supernatant from titration of FeS by Cd2+ (Di
Toro, unpublished data). The solid line represents the result expected from theory.
2-10
-------
Partitioning
this causes the solid-phase iron sulfide to dissolve,
releasing sulfide that is available for formation of
cadmium sulfide. The reaction is
Cd2+ + FeS(s) - CdS(s) + Fe2
(2-2)
Cadmium titrations with amorphous FeS and with
sediments were performed to examine this
possibility.
2.2.2.1 Amorphous FeS
A direct test of the extent to which this
reaction takes place was performed (Di Toro et
al., 1990). A quantity of freshly precipitated iron
sulfide was titrated by adding dissolved cadmium.
The resulting aqueous cadmium activity, measured
with the cadmium electrode, versus the ratio of
cadmium added [CdA] to the amount of FeS
initially present [FeS(s)]i is shown in Figure 2-8.
The plot of dissolved cadmium versus cadmium
added illustrates the increase in dissolved cadmium
that occurs near [CdA]/[FeS(s)]i =1. It is
interesting to note that these displacement
reactions among metal sulfides have been
observed by other investigators (Phillips and
Kraus, 1965). The reaction was also postulated by
Pankow (1979) to explain an experimental result
involving copper and synthetic FeS.
These experiments plainly demonstrate that
solid-phase amorphous iron sulfide can be readily
displaced by adding cadmium. As a consequence,
the source of available sulfide must be taken into
account when evaluating the relationship between
solid-phase and aqueous-phase cadmium in
sediments.
A direct confirmation that the removal of
cadmium was through the displacement of iron
sulfide is shown in Figure 2-9. The supernatant
from a titration of FeS by Cd2+ was analyzed for
both iron and cadmium. The solid lines are the
theoretical expectations based on the stoichiometry
of the reaction.
2.2.2.2 Sediments
A similar titration procedure has been used to
evaluate the behavior of sediments taken from
four different marine environments: sediments
from Black Rock Harbor and the Hudson River,
and the sediments from Long Island Sound and
Ninigret Pond used in the toxicity tests (Di Toro et
al., 1990). The binding capacity for cadmium is
estimated by extrapolating a straight line fit to the
dissolved cadmium data. The equation is
[SCd(aq)] = max {m([CdJ - [CdJ)}
(2-3)
where [SCd(aq)] is the total dissolved cadmium,
[CdJ is the cadmium added, [CdB] is the bound
cadmium, and m is the slope of the straight line.
The different sediments exhibit quite different
Table 2-1. Cadmium binding capacity and AVS of sediments
Sediment
Black Rock Harbor
Hudson River
LI Soundd
, .-. , d,e
Mixture
Ninigret Pond
Initial AVS3
Omol/g)
175 (41)
12.6 (2.80)
15.9(3.30)
5.45 ( )
2.34 (0.73)
Final AVSb
Omol/g)
13.9 (6.43)
3.23(1.18)
0.28(0.12)
Cd Binding Capacity
Omol/g)
114(12)
8.58 (2.95)
4.57 (2.52)
1.12(0.42)
aAverage (SD) AVS of repeated measurements of the stock.
bAverage (SD) AVS after the sediment toxicity experiment.
°From Equation 2-3.
dFrom original cadmium experiment.
e50/50 mixture of LI Sound and Ninigret Pond.
Source: Di Loro et al., 1990.
2-11
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
binding capacities for cadmium, listed in Table 2-1,
ranging from approximately 1 pmol/g to more than
100 jwmol/g. The question as to whether this
binding capacity is explained by the solid-phase
sulfide present in the samples is addressed in
subsequent sections of this document.
2.2.3 Correlation to Sediment AVS
The majority of sulfide in sediments is in the
form of iron monosulfides (mackinawite and
greigite) and iron bisulfide (pyrite), of which the
former is the most reactive. These sediment
sulfides can be classified into three broad classes
that reflect the techniques used for quantification
(Berner, 1967; Goldhauber and Kaplan, 1974;
Morse et al., 1987). The most labile fraction, AVS,
is associated with the more soluble iron
monosulfides. The more resistant sulfide mineral
phase, iron pyrite, is not soluble in the cold acid
extraction used to measure AVS. Neither is the
third compartment, organic sulfide, which is
associated with the organic matter in sediments
(Landers et al., 1983).
The possibility that acid volatile sulfide is a
direct measure of the solid-phase sulfide that
reacts with cadmium is examined in Table 2-1,
which lists the sediment-binding capacity for
cadmium and the measured AVS for each
sediment, and in Figure 2-10, which indicates the
initial AVS concentration. The sediment cadmium-
binding capacity appears to be somewhat less than
the initial AVS for the sediments tested. However,
a comparison between the initial AVS of the
sediments and that remaining after the cadmium
titration is completed suggests that some AVS is
lost during the titration experiment (Table 2-1). In
any case, the covariation of sediment-binding
capacity and AVS is clear. This suggests that
AVS
AVS
2
&
I
I
u
8
BR Harbor
LI Sound
Hudson River
Ninigret Pond
0.4 -
0.2
0.0
1.0
10.0
100.0
1000.0
Cadmium Added (/zmol Cd/g dry wt)
Figure 2-10. Cadmium titration of sediments from Black Rock Harbor, Long Island Sound, Hudson River, and
Ninigret Pond. Cadmium added per unit dry weight of sediment versus dissolved cadmium. Arrows
are the measured AVS concentrations for the four sediments (figure from Di Toro et al., 1990).
2-12
-------
Table 2-2. Metal sulflde solubility products and ratios
Partitioning
Metal Sulflde Log10ATsp)2 Log10ATsp Loglo(KMS/KfeS)
FeS
NiS
ZnS
CdS
PbS
CuS
Ag2S
-3.64
-9.23
-9.64
-14.10
-14.67
-22.19
-36.14
-22.39
-27.98
-28.39
-32.85
-33.42
-40.94
-54.71
-5.59
-6.00
-10.46
-11.03
-18.55
-32.32
"Solubility products, Ksp2 for the reaction M2+ + HS' » MS(s) + H+ for FeS (mackinawite), NiS (millerite), and CdS
(greenockite) from Emerson et al. (1983). Solubility products for ZnS (wurtzite), PbS (galena), CuS (covellite), and Ag2S
(acanthite) and pK2 = 18.57 for the reaction HS' » H+ + S2' from Schoonen and Barnes (1988).
bK for the reaction M2+ + S2" « MS(s) is computed from log K 2 and pK2.
measurement of AVS is the proper quantification
of the solid-phase sulfides that can be dissolved by
the addition of ionic cadmium. The chemical basis
for this is examined below.
2.2.4 Solubility Relationships and
Displacement Reactions
Iron monosulfide, FeS(s), is in equilibrium with
aqueous-phase sulfide and iron via the reaction
FeS(s) « Fe2++ S2-
(2-4)
If cadmium is added to the aqueous phase, the
re suit is
Cd2+ + FeS(s) « Cd2++ Fe2++ S2
(2-5)
As the cadmium concentration increases, [Cd2+]
[S2"] will exceed the solubility product of cadmium
sulfide and CdS(s) will start to form. Since the
cadmium sulflde is more insoluble than iron
monosulfide, FeS(s) should start to dissolve in
response to the lowered sulfide concentration in
the interstitial water. The overall reaction is
Cd2+ + FeS(s) « CdS(s) + Fe2
(2-6)
The iron in FeS(s) is displaced by cadmium to
form soluble iron and solid cadmium sulflde,
CdS(s). The consequence of this replacement
reaction can be seen using the analysis of the
M(II)-Fe(II)-S(-II) system with both MS(s) and
FeS(s) presented in Di Toro et al. (1992). M2+
represents any divalent metal that forms a sulflde
that is more insoluble than FeS. If the added
metal, [M]A, is less than the AVS present in the
sediment then the ratio of metal activity to total
metal in the sediment-interstitial water system is
less than the ratio of the MS to FeS solubility
product constant
(2-7)
This general result is independent of the details
of the interstitial water chemistry. In particular, it
is independent of the Fe2+ activity. Of course, the
actual value of the ratio {M2+}/[M]A depends on
aqueous speciation, as indicated by Equation 2-6.
However, the ratio is still less than the ratio of the
sulfide solubility products.
This is an important finding because the data
presented in Section 2.1.1 indicate that toxicity is
related to metal activity, {M2+}. This inequality
guarantees that the metal activity, in contrast to the
total dissolved metal concentration, is regulated by
the iron sulfide-metal sulfide system.
The metal sulfide solubility products and the
ratios are listed in Table 2-2. For example, the
ratio of cadmium activity to total cadmium is less
2-13
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
than 10~1046. For nickel, the ratio is less than
10~559. By inference, this reduction in metal
activity will occur for any other metal that forms a
sulfide that is significantly more insoluble than iron
monosulfide. The ratios for the other metals in
Table 2-2 (Cu, Pb, Ag, and Zn) indicate that metal
activity for these metals will be very small in the
presence of excess AVS.
2.2.5 Application to Mixtures of Metals
A conjecture based on the sulfide solubility
products for the metals listed in Table 2-2 is that
the sum of the molar concentrations of metals
should be compared with AVS. Because all these
metals have lower sulfide solubility parameters
than FeS, they would all exist as metal sulfides if
their molar sum (and using [Ag]/2 because it is
monovalent) is less than the AVS. For this case
Si[MT]i < AVS
(2-8)
no metal toxicity would be expected, where [MT]i
is the total cold acid extractable i metal molar
concentration in the sediment (divided by 2 for
silver). On the other hand, if their molar sum is
greater than the AVS concentration, then a portion
of the metals with the largest sulfide solubility
parameters would exist as free metal and
potentially cause toxicity. For this case the
following would be true
Si[MT]i > AVS
(2-9)
These two equations are precisely the formulas
that could be employed to determine the extent of
metal toxicity in sediments assuming additive
behavior and neglecting the effect of partitioning to
other sediment phases. Whether the normalized
sum is less than or greater than 1.0 discriminates
between nontoxic and potentially toxic sediments.
The additivity does not come from the nature of
the mechanism that causes toxicity. Rather, it
results from the equal ability of the metals to form
metal sulfides with the same stoichiometric ratio of
M and S (except silver).
The appropriate quantity of metals to use in
the metals and AVS comparison is referred to as
SEM, that is, the metal extracted with the cold
acid used in the AVS procedure. This is the
appropriate quantity to use because some metals
form sulfides that are not labile in the AVS
extraction (e.g., nickel, copper). If a more
rigorous extraction were used to increase the
fraction of metal extracted that did not also
capture the additional sulfide extracted, then the
sulfide associated with the additional metal release
would not be quantified. This would result in an
erroneously high metal value relative to AVS (Di
Toroetal, 1992).
The above discussion is predicated on the
assumption that all the metal sulfides behave
similarly to cadmium sulfide. Furthermore, it has
been assumed that only acid-soluble metals are
reactive enough to affect the free metal activity.
That is, the proper metal concentration to be used
is the SEM. Both of these hypotheses were tested
directly with benthic organisms using sediment
toxicity tests. Results of these sediment-spiking
experiments with cadmium, copper, lead, nickel,
silver, zinc, and a mixture of these metals are
presented in Section 3.
2-14
-------
Toxicity of Metals in Sediments
Section 3
Toxicity of Metals in Sediments
3.1 General Information
This section summarizes data from acute and
chronic toxicity tests that demonstrate that absence of
sediment toxicity caused by metals can be predicted by
(a) the use of interstitial water concentrations of metals
or (b) comparison of molar concentrations of AVS and
SEM. Furthermore, they demonstrate that use of
(ZSEM-AVS)//OC reduces the variability associated
with prediction of when sediments will be toxic. The
ability to predict toxicity of metals in sediments,
through a fundamental understanding of chemical
bioavailability, is demonstrated using results of toxicity
tests with benthic organisms in spiked or field
sediments. A wide variety of individual benthic
species having different habitat requirements have
been tested in 10-day experiments in spiked and field
sediments, including the following: an oligochaete
(Lumbriculus variegatus), polychaetes (Capitella
capitata and Neanthes arenaceodentata), amphipods
(A. abdita,R. hudsoni, Leptocheims plumulosus, and
Hyalella azteca), a harpacticoid copepod (Amphiascus
tenuiremis), a midge (Chironomus tentans), and a
gastropod (Helisoma sp.). In addition, the approach
was tested in life-cycle tests with L. plumulosus and C.
tentans. Many other benthic species were tested in
freshwater and saltwater benthic colonization studies.
3.1.1 Terminology
Early studies on use of AVS in prediction of
biological effects (e.g., Di Toro et al., 1990) involved
the ratio of SEM to AVS, expressed as SEM/AVS. The
ratio appeared more useful in the early laboratory tests
because it caused concentration-response data from
spiking experiments with different sediments to fall on
the same line (DiToro etal., 1990,1992; Casas and
Crecelius, 1994; Peschetal., 1995; Berry etal., 1996).
Later studies, however, showed several advantages to
the use of the difference, expressed as SEM-AVS
(Hansenetal., 1996a). The two expressions
SEM/AVS < 1 and SEM-AVS < 0are functionally
equivalent. Both indicate an excess of AVS over SEM.
The advantages to using SEM-AVS are that it does not
get very large when AVS is very low (as the ratio does),
and that it can be used to develop partitioning
relationships that include other phases, such as total
organic carbon (TOC) (see Section 3.4; see also the
discussion in Section 3.2.5). For these reasons, the use
of the SEM-AVS difference is the recommended method,
and it will be used throughout the rest of this document
except in the discussion of the historical development
of AVS theory that follows. In the ensuing discussion,
SEM/AVS ratios are presented because they were
originally presented in this form.
3.2 Predicting Metal Toxicity:
Short-Term Studies
3.2.1 Spiked Sediments: Individual
Experiments
A key to understanding the bioavailability of
sediment-associated contaminants was provided by
Adams et al. (1985), who observed that the effects of
kepone, a nonionic organic pesticide, were similar
across sediments when toxicity was related to
interstitial water concentrations. Swartzetal. (1985)
and Kemp and Swartz (1986) first observed that metal
concentrations in interstitial waters of different
sediments were correlated with observed biological
effects. However, as opposed to the situation for
nonionic organic chemicals and organic carbon (see Di
Toro et al., 1991), the sediment-partitioning phases that
controlled interstitial water concentrations of metals
and metal-induced sediment toxicity were initially not
apparent.
Di Toro et al. (1990) first investigated the
significance of sulfide partitioning in controlling metal
bioavailability and metal-induced toxicity in marine
sediments spiked with cadmium. In these experiments,
the operational definition of Cornwell and Morse (1987)
was used to identify that fraction of amorphous sulfide,
or AVS, available to interact with cadmium in the
sediments. Specifically, AVS was defined as the sulfide
liberated from wet sediment when treated with cold IN
HC1 acid. Di Toro et al. (1990) found that, when
expressed on a dry weight basis, the toxicity of
cadmium in sediments in 10-day tests with the
amphipods R. hudsoni or A. abdita was sediment
specific (Figure 3-1A; fromDiToro etal., 1990).
3-1
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Toxicity increased with increasing cadmium
concentration, but the concentration-response
relationships were different for each sediment. Thus, it
would not be possible to predict whether a particular
sediment would be toxic or not. If the cadmium
concentration is expressed on an interstitial water basis
(Figure 3-1B), however, concentration response is not
sediment specific. Similar results are observed when
cadmium concentration is expressed as SEM/AVS
(Figure 3 -1C). Note that when the ratio of ^mol
Cd/^mol AVS was less than 1.0, the sediments were not
toxic, and when the ratio was greater than 1.0, the
sediments became increasingly toxic. Studies by
Carlson etal.(1991) with cadmium-spiked freshwater
sediments yielded similar results; when there was more
AVS than total cadmium, significant toxicity was not
observed in 10-day tests with an oligochaete
(L. vahegatus) or snail (Helisoma sp). Di Toro et al.
(1992), in their studies with nickel-spiked sediments
using A. abdita and field sediments contaminated with
cadmium and nickel using the freshwater amphipod
H. azteca, provided further support to the importance
of AVS in controlling metal bioavailability in sediments.
These studies suggested that it may be feasible to
derive an ESB for mixtures of metals by direct
comparison of molar AVS concentrations to the molar
sum of the concentrations of cationic metals
(specifically, cadmium, copper, lead, nickel, and zinc)
extracted with the AVS (i.e., ZSEM). They observed
that expression of metals concentrations based on the
sum of SEM concentrations is required because a
significant amount of nickel sulfide is not completely
soluble in the AVS extraction. Hence, AVS must be
used as the measure of reactive sulfide and the sum of
SEM as the measure of total reactive metal.
Casas and Crecelius (1994) further explored the
relationship of SEM and AVS, interstitial water
concentrations, and toxicity by conducting 10-day
toxicity tests with the marine polychaete C. capitata
exposed to sediments spiked with zinc, lead, and
copper. As was true in earlier studies, elevated
interstitial water metal concentrations were observed
only when SEM concentrations exceeded those of AVS.
Sediments were not toxic when SEM concentrations
were less than AVS and when the concentrations in
interstitial water were less than the water-only LC50
values. Green et al. (1993) reported results of another
spiking experiment supporting this general EqP
approach to deriving an ESB for metals. In their study,
metal-sulfide partitioning was not directly quantified,
but it was found that toxicity of cadmium-spiked marine
sediments to the meiobenthic copepod A tenuiremis
was predictable based on interstitial water, but not
sediment dry weight cadmium concentrations. Further
spiking experiments by Pesch et al. (1995) demonstrated
that 10-day survival of the marine polychaete
N. arenceodentata was comparable to controls in
cadmium- or nickel-spiked sediments with more AVS
than SEM.
Berry etal. (1996) described experiments in which
A. abdita were exposed for 10 days to two or three
sediments spiked either singly, or in combination, with
cadmium, copper, lead, nickel, and zinc. As in previous
studies, significant toxicity to the amphipod did not
occur when AVS concentrations exceeded those of
SEM. They compared observed mortality with
interstitial water metal concentrations expressed as
interstitial water toxic units (IWTUs)
IWTU = [M ,]/LC50
(3-1)
where [Md] is the dissolved metal concentration in the
interstitial water, and the LC50 is the concentration of
the metal causing 50% mortality of the test species in a
water-only test. If interstitial water exposure in a
sediment test is indeed equivalent to that in a water-
only test, then 1.0 IWTU should result in 50% mortality
of the test animals. Berry etal. (1996) reported that
significant (>24%) mortality of the saltwater amphipod
occurred in only 3.0% of sediments with less than 0.5
IWTU, whereas samples with greater than 0.5 IWTUs
were toxic 94.4% of the time. Berry et al. (1996) also
made an important observation relative to interstitial
water metal chemistry in their mixed-metals test;
chemical equilibrium calculations suggest that the
relative affinity of metals for AVS should be silver>
copper>lead>cadmium>zinc>nickel (Emerson et al.,
1983; Di Toro et al., 1992); hence, the appearance of the
metals in interstitial water as AVS is exhausted should
occur in an inverse order. For example, zinc would
replace nickel in a monosulfide complex and nickel
would be liberated to the interstitial water, and so on.
Berry et al. (1996) observed this trend in sediments
spiked with cadmium, copper, nickel, and zinc (Figure
3-2). Furthermore, an increase in the concentration of a
metal in a sediment with a low sulfide solubility product
constant (K ) theoretically would displace a
previously unavailable and nontoxic metal with a higher
K making that metal available to bind to other
sediment phases or enter interstitial water to become
toxic. Berry et al. (1999) exposed the saltwater
amphipod A. abdita to sediments spiked with silver.
When AVS was detected in the sediments, they were
not toxic and interstitial water contained no detectable
silver. For sediments that contain no detectable AVS,
3-2
-------
Toxicity of Metals in Sediments
1
I
o
o
100
80
60
40
20
0
A
LI Sound
Mixture
O NigretPond
-o
10 100 1000 10000
Sediment Cadmium (//g Cd/g dry wt)
100000
B
Water Only
Exposure
A Ampelisca
& Rhepoxy.iius
100
80
60
40
20
0
0.00001 0.0001 0.001 0.01 0.1 1 10
Cadmium Activity (mg Cd2+/L)
100 1000
100
80
60
40
20
c
LI Sound
Mixture
O Nigret Pond
0.01 0.10 1 10
Sediment Cadmium (umol Cd/V/m AYS)
100
Figure 3-1. Percentage mortality of amphipods (Ampelisca abdita and Rhepoxynius hudsoni) exposed to sediments
from Long Island Sound, Ninigret Pond, and a mixture of these two sediments as a function of the
sum of the concentrations of metals in sediments expressed as: (A) dry weight, (B) interstitial water
cadmium activity, and (C) the sediment cadmium/AVS ratio (figures from Di Toro et al., 1990).
3-3
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
B
3
£
a
i
I
1
I
S
I
f
I
10000-
1000-
100-
10-
1-
0.1-
0.01-
Ni *
Zn
Cd
0.01
0.1
10
100
1000
SEM/AVS
10000.
1000-
100-
10-
1-
0.1-
0.01.
B
Cd»
0.01
0.1
10
100
1000
SEM/AVS
Figure 3-2. Concentrations of individual metals in interstitial water of sediments from Long Island Sound (A) and
Ninigret Pond (B) in the mixed metals experiment as a function of SEM/AVS ratio. Concentrations
below the interstitial water detection limits, indicated by arrows, are plotted at one-half the detection
limit. K is the sulfide solubility product constant (figures from Berry et al., 1996).
3-4
-------
Toxicity of Metals in Sediments
any SEM silver that is detected is dissolved interstitial
silver, because silver sulfide and silver chloride
precipitate are not extracted using the standard AVS
procedure.
3.2.2 Spiked Sediments: All Experimental
Results Summarized
This summary includes data from amphipods
exposed in 10-day toxicity tests to saltwater sediments
spiked with cadmium, copper, lead, nickel, silver, or zinc
and their mixtures (DiToroetal., 1990; Berry etal., 1996,
1999); polychaetes exposed to sediments spiked with
cadmium, copper, lead, nickel, or zinc (Casas and
Crecelius, 1994; Pesch etal., 1995) ;copepods exposed
to sediments spiked with cadmium (Green et al., 1993;
measured interstitial cadmium but not AVS); and
freshwater tests using oligochaetes and snails exposed
to sediments spiked with cadmium (Carlson etal., 1991).
Seven species (freshwater and saltwater) and sediments
from seven different locations were described. AVS
concentrations ranged from 1.9 to 65.7 ^mol/g dry
weight, andTOCrangedfrom0.15%to 10.6%in these
sediments.
Overall, the results of these experiments
demonstrate that predictions of the toxicity of
sediments spiked with metals using the total metal
concentration on a dry weight basis are not based on
scientific theories of bioavailability and will have
considerable error (Figures 3-3 A and 3-4A). Sediments
having <24% mortality are considered nontoxic as
defined by Berry et al. (1996), which is indicated by the
horizontal line in Figure 3-3. Furthermore, the
concentration range where it is 90% certain that the
sediment may be either toxic or nontoxic, shown as
dashed lines in Figure 3-3, is almost two orders of
magnitude for dry weight metals, a little over an order
of magnitude for IWTUs, and only a half order of
magnitude for SEM/AVS (see Section 3.4 for a
description of the derivation of the uncertainty limits).
The uncertain range for dry weight metals is
approximately equal to the sum of the uncertainty range
for SEM/AVS plus the range in the AVS concentrations
of the spiked sediments in the database. If sediments
with a lower AVS concentration had been tested,
effects would have occurred at a lower dry weight
concentration, and if sediments with lower or higher
AVS concentrations had been tested, the uncertainty
range would increase. Importantly, the uncertainty
range for IWTUs or SEM/AVS would likely not be
altered.
Even given the above, it is visually tempting to
select a cutoff at a dry weight concentration of 1.0
,wmol/g to indicate the separation of sediments that are
toxic or nontoxic. This would be inappropriate because
toxicity of metals in sediments when concentrations are
expressed as dry weights have been shown to be
sediment specific (Figure 3-1A). Also, had sediments
with lower or higher AVS concentrations been tested,
the cutoff would have been at lower or higher dry
weight concentrations. However, to further
demonstrate the risks of establishing a dry weight
cutoff, the data from the 184 spiked sediments in Figure
3 -3 were re-analyzed. A visually based cutoff of 1.0
^tmol/g dry weight, and theoretically based cutoffs of
0.5 IWTU and 1.0 SEM/AVS were selected. Sediment
concentrations were numerically ordered. Those with
concentrations less than the cutoffs were divided into
three groups containing approximately the same
number of sediments (15,22, or 25 sediments per group
for dry weight metal concentrations, IWTUs, and SEM/
AVS, respectively). Similarly, sediments containing
greater concentrations were divided into six groups (21,
16, or 14 sediments per group for dry weight metal
concentrations, IWTUs, and SEM/AVS, respectively).
The percentages of nontoxic (<24% mortality) and toxic
(>24% mortality) sediments in each group are plotted in
a stacked bar plot (Figure 3-4). Not surprisingly,
because the distribution was visually selected, most
sediments having less than 1.0 ^mol/g dry weight metal
were not toxic. The same was true for the
lexicologically selected cutoffs of 0.5 IWTUs and SEM/
AVS ratios of 1.0. The advantage of using IWTUs and
SEM/AVS becomes more clear when the sediments
above the cutoffs are considered. For dry weight metal
concentrations, more of the sediments in the first four
sediment groups (up to 26.8 ^tmol/g dry weight) were
nontoxic than were toxic. It was only in the two
sediment groups that contained the highest
concentrations, >27.6 ^mol/g dry weight, that toxic
sediments predominated after the first two sediment
groups. In contrast, toxic sediments predominated in
only the first two sediment groups above the IWTU
cutoff and after the first sediment group above the
SEM/AVS ratio cutoff.
In some cases, the dry weight metal concentrations
required to cause acute mortality in these experiments
were very high relative to those often suspected to be
of lexicological significance infield sediments (e.g.,
Figures 3-1A and 3-3 A). This has sometimes been
interpreted as a limitation of Ihe use of SEM and AVS lo
predicl melal-induced loxicily. However, Ihe range of
AVS in Ihese sedimenls spiked wilh melals is similar lo
3-5
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Mortality (%) Mortality (%) Mortality (%)
100
80
60
40
20
0'
0
100
80
60
40
20
0-j
0
100
80
60
40
20,
34% 56% 10%
A 1. _1
- " b O.ODCDOBOO OD ooboD mo coo -
1 0° a ' o d>o ii#"H9lb gij 44,14111 i IMIIII
.01 0.1 1 10 100 1000
Total Metal or SEM (/umol/g dry wt)
51% 25% 24%
d OO OO OO O .OOD CO OtXDO.fi
1 . . t,
'
1
, . : .
O O_ 0 (
«-->.£--! -n
8 £ oft o o 0*0 o
^8 °ooi, a i ° 8
8 So'oi'SiJi'oS ' lulllorf>t<% "$' 'd>
.01 0.1 1 10 100 1000
Interstitial Water Toxic Units
64% 27% 9%
i i 1 1 inn i i 1 1 HIM i i i Mini n i 1 1 HIM i i 1 1 HIM 1 1 1 1 INI
C ' '
ob>aQpBoaD ^oo a o a> o o e
i.wj: \
f 1 0
f °
1,0
lo
ir °
° °^o-hl"
' o ^ Is*
ootfo,*.?*0.. f1><= o
o "j^lxLsX01
0.001 0.01 0.1 1 10 100 1000
SEM/AVS
Figure 3-3.
Percentage mortality of freshwater and saltwater benthic species in 10-day toxicity tests in sediments
spiked with individual metals (Cd, Cu, Pb, Ni, Ag, or Zn) or a metal mixture (Cd, Cu, Ni, and
Zn). Mortality is plotted as a function of: (A) the sum of the concentrations of the respective metal
or metal mixture in /^mol metal per gram dry weight of sediment; (B) IWTU; and (C) SEM/AVS
ratio. Data below the detection limits are plotted at IWTU=0.01 and SEM/AVS = 0.001. Heavy
dashed lines are the theoretically based cutoffs of 0.5 IWTU and a SEM/AVS ratio of 1.0. Light
vertical dashed lines are the 90% uncertainty bound limits derived as in Section 3.4. The percentage
of the total number of sediments (n = 184) within the bounded limits is provided above each of the
three panels for the purpose of comparison (silver data from Berry et al., 1999; all other data
modified after Berry et al., 1996).
3-6
-------
Toxicity of Metals in Sediments
s«
a
o
'.a
03
I
u
e
OJ
PH
I
(B
s
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
that of sediments commonly occurring in the field. The
important point here is that even a sediment with only a
moderate concentration of AVS has a considerable
capacity for sequestering metals as a metal sulfide, a
form that is notbioavailable (Di Toro et al., 1990).
In contrast, the combined data from all available
freshwater and saltwater spiked-sediment experiments
support the use of IWTUs to predict mortality of
benthic species in spiked-sediment toxicity tests
(Figure 3-3B). Mortality in these experiments was
sediment independent when plotted against IWTUs.
Sediments with IWTUs of <0.5 were generally not toxic.
Of the 96 sediments with IWTUs <0.5,96.9% were not
toxic, whereas 76.4% of the 89 sediments with IWTUs
>0.5 were toxic (Table 3-1). This close relationship
between IWTUs and sediment toxicity in sediments
spiked with metals was also observed in studies with
field sediments contaminated with metals (see Section
3.2.3 below), as well as sediments spiked with nonionic
organic chemicals (Adams et al., 1985; Swartz et al.,
1990; Di Toro etal., 1991), and field sediments
contaminated with nonionic organic chemicals (Hoke et
al., 1994; Swartz etal., 1994).
Table 3-1. Toxicity of sediments from freshwater and saltwater lab-spiked sediment tests, field locations, and
combined lab-spiked and field sediment tests as a function of the molar concentrations of SEM and
AVS (SEM/AVS or the SEM-AVS), interstitial water toxic units (IWTUs), and both SEM/AVS or SEM-
AVS and IWTUs
Percent of Sediments
Study Type/Parameter
Laboratory Spike:
SEM/AVS or SEM-AVSC
IWTUd
SEM/AVS or SEM-AVSC; IWTUd
Field:
SEM/AVS or SEM-AVSC
IWTUd
SEM/AVS or SEM-AVSC; IWTUd
Lab-Spike and Field:
SEM/AVS or SEM-AVS0
IWTUd
SEM/AVS or SEM-AVS0; IWTUd
Value
<1.0or<0.0
>1.0or>0.0
<0.5
>0.5
<1.0or <0.0;<0.5
>1.0or>0.0; >0.5
<1.0or <0.0
>1.0or>0.0
<0.5
>0.5
<1.0or<0.0;<0.5
>1.0or>0.0; >0.5
<1.0or<0.0
>1.0or>0.0
<0.5
>0.5
<1.0or<0.0;<0.5
>1.0or>0.0; >0.5
n
101
95
96
89
83
78
57
79
79
53
49
45
158
174
175
142
132
123
Nontoxic
98.0
26.3
96.9
23.6
97.6
14.1
98.2
59.5
98.7
45.3
100.0
33.3
98.1
42.0
97.7
31.7
98.5
21.1
rj, D
Toxic
2.0
73.7
3.1
76.4
2.4
85.9
1.8
40.5
1.3
54.7
0.0
66.7
1.9
58.0
2.3
68.3
1.5
78.9
aNontoxic sediments <24% mortality.
Toxic sediments >24% mortality.
cAn SEM/AVS ratio of <1.0 or an SEM-AVS difference of <0.0 indicates an excess of sulfide and probable nontoxic
sediments. An SEM/AVS ratio of >1.0 or an SEM-AVS difference of >0.0 indicates an excess of metal and potentially
toxic sediments.
An IWTU of <0.5 indicates a probable nontoxic interstitial water concentration of less than one-half of the water-only
LC50 of the same duration. An IWTU of >0.5 indicates a possibly toxic interstitial water concentration of greater than
one-half of the water-only LC50 of the same duration.
Source: Modified from Hansen et al., 1996a.
3-8
-------
Toxicity of Metals in Sediments
The interstitial water metal concentrations in
spiked-sediment studies were most often below the
limit of analytical detection in sediments with SEM/AVS
ratios below 1.0 (Berry etal., 1996). Above an SEM/
AVS ratio of 1.0, the interstitial metals concentrations
increased up to five orders of magnitude with
increasing SEM/AVS ratio. This increase of several
orders of magnitude in interstitial water metals
concentration with an increase of only a factor of two
or three in sediment concentration is the reason why
mortality is most often complete in these sediments,
and why the chemistry of anaerobic sediments controls
the toxicity of metals to organisms living in aerobic
microhabitats. It also explains why toxicities of
different metals in the same sediment to different
species when expressed on the basis of sediment
metals concentration are so similar. Interstitial water
metals were often below or near detection limits when
SEM/AVS ratios were only slightly above 1.0,
indicating the presence of other metal-binding phases
in sediments.
The combined data from all available freshwater
and saltwater spiked-sediment experiments also support
the use of SEM/AVS ratios to predict sediment toxicity
to benthic species in spiked-sediment toxicity tests. All
tests yield similar results when mortality is plotted
against SEM/AVS ratios (Figure 3-3C). Mortality in
these experiments was sediment independent when
plotted on an SEM/AVS basis. With the combined data,
98.0% of the 101 metals-spiked sediments with SEM/
AVS ratios < 1.0 were nottoxic, whereas 73.7% of the 95
sediments with SEM/AVS ratios >1.0 were toxic (Table
3-1).
The overall data show that when both SEM/AVS
ratios and IWTUs are used, predictions of sediments
that would be toxic were improved. Of the 83 sediments
with SEM/AVS ratios < 1.0 and IWTUs <0.5,97.6% were
not toxic, whereas 85.9% of the 78 sediments with SEM/
AVS ratios >1.0 and IWTUs >0.5 were toxic (Table 3-1).
These results show that SEM/AVS and IWTUs are
accurate predictors of the absence of mortality in
sediment toxicity tests; however, predictions of
sediments that might be toxic are less accurate. The
fact that a significant number of sediments (26.3%)
tested had SEM/AVS ratios of > 1.0 but were nottoxic
indicates that other binding phases, such as organic
carbon (Mahony et al., 1996), may also control
bioavailability in anaerobic sediments.
Organism behavior may also explain why some
sediments with SEM/AVS ratios of >1.0 were nottoxic.
Many of the sediments that had the highest SEM/AVS
ratios in excess of 1.0 that produced little or no
mortality were from experiments using the polychaete
N. arenaceodentata (see Pesch etal., 1995). In these
experiments, this polychaete did not burrow into some
of the test sediments with the highest concentrations,
thereby limiting its exposure to the elevated
concentrations of metals in the interstitial water and
sediments. This same phenomenon may also explain
the low mortality of snails, Heliosoma sp., in freshwater
sediments with high SEM/AVS ratios. These snails are
epibenthic and crawl onto the sides of test beakers to
avoid contaminated sediments (GL. Phipps, U.S. EPA,
Duluth, MN, personal communication). Increased
mortality was always observed in sediments with SEM/
AVS ratios >5.9 in tests with the other five species.
Similarly, a significant number of sediments (23.6%)
with >0.5 IWTUs were nottoxic. This is likely the
result of interstitial water ligands, which reduces the
bioavailability and toxicity of dissolved metals;
sediment avoidance by polychaetes or snails; or
methodological problems in contamination-free
sampling of interstitial water. Ankley etal. (1991)
suggested that a toxicity correction for the hardness of
the interstitial water for freshwater sediments is needed
to compare toxicity in interstitial water with that in
water-only tests. Absence of a correction for hardness
would affect the accuracy of predictions of metal-
induced sediment toxicity using IWTUs. Furthermore,
a significant improvement in the accuracy of metal-
induced toxicity predictions using IWTUs might be
achieved if DOC binding in the interstitial water is taken
into account. Green et al. (1993) and Ankley et al.
(1991) hypothesized that increased DOC in the
interstitial water reduced the bioavailability of cadmium
in sediment exposures, relative to the water-only
exposures. Green et al. (1993) found that the LC50
value for cadmium in an interstitial water exposure
without sediment was more than twice that in a water-
only exposure, and that the LC50 value for cadmium in
interstitial water associated with sediments was more
than three times that in a water-only exposure.
3.2.3 Field Sediments
In addition to short-term laboratory experiments
with spiked sediments, there have been several
published studies of laboratory toxicity tests with
metal-contaminated sediments from the field. Ankley et
al. (1991) exposed!, var/'egatoandtheamphipod/f.
azteca to 17 sediment samples along a gradient of
cadmium and nickel contamination from a freshwater/
3-9
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
estuarine site in Foundry Cove, NY. In 10-day toxicity
tests, H. azteca mortality was not significantly different
from controls in all sediments where SEM (cadmium
plus nickel) was less than AVS. Mortality was greater
than controls only in sediments with more SEM than
AVS. L. variegatus was far less sensitive to the
sediments than H. azteca, which correlates with the
differential sensitivity of the two species in water-only
tests with cadmium and nickel.
In 10-day toxicity tests with the saltwater
amphipodA abdita in these same sediments, Di Toro
et al. (1992) observed that metals concentrations
ranging from 0.1 to 28 ^mol SEM/g sediment were not
toxic in some sediments, whereas metals concentrations
ranging from 0.2 to 1,000 ^mol SEM/g sediment were
lethal in other sediments. These results indicate that
the bioavailable fraction of metals in sediments varies
from sediment to sediment. In contrast, the authors
also observed a clearly discernible mortality-
concentration relationship when mortality was related
to the SEM/AVS molar ratio (i.e., there was no
significant mortality where SEM/AVS ratios were <1.0,
mortality increased in sediments having SEM/AVS
ratios of 1.0 to 3.0, and there was 100% mortality in
sediments with ratios >10). The sum of the IWTUs for
cadmium and nickel ranged from 0.08 to 43.5.
Sediments with <0.5 IWTUs were always nontoxic,
those with >2.2 IWTUs were always toxic, and two of
seven sediments with intermediate IWTUs (0.5 to 2.2)
were toxic. Molar concentrations of cadmium and
nickel in the interstitial water were similar. However,
cadmium contributed over 95% to the sum of the toxic
units because cadmium is 67 times more toxic to A
abdita than nickel. The latter illustrates the utility of
interstitial water concentrations of individual metals in
assigning the probable cause of mortality in benthic
species (Hansenetal., 1996a).
In tests with the same sediments from Foundry
Cove,Peschetal. (1995) observed that 6 of the 17
sediments tested had SEM/AVS ratios <1.0 and IWTUs
<0.5, and none of the 6 were toxic to the polychaete N.
arenaceodentata. Interestingly, the other 11 sediments
containing SEM/AVS ratios >1.0 were also not toxic.
The results are not surprising given that in these
particular tests only one sediment had >0.5 IWTUs, N.
arenaceodentata is not sensitive to cadmium and
nickel, and the polychaetes did not burrow into
sediments containing toxic concentrations of these
metals.
Ankley etal. (1993) examined the significance of
AVS as a binding phase for copper in freshwater
sediments from two copper-impacted sites. Based on
interstitial water copper concentrations in the test
sediments, the 10-day LC50 for/f. azteca was 31 /-ig/L;
this compared favorably with a measured LC50 of 28
/-ig/L in a 10-day water-only test. Sediments having
SEM/AVS ratios <1.0 were not toxic. They also
observed no toxicity in several sediments with
markedly more SEM than AVS, suggesting that copper
was not biologically available in these sediments.
Absence of copper in interstitial water from these
sediments corroborated this lack of bioavailability.
This observation suggested the presence of binding
phases in addition to AVS for copper in the test
sediments. Two studies suggest that an important
source of the extra binding capacity in these sediments
was organic carbon (U.S. EPA, 1994a; Mahony et al.,
1996).
Hansenetal. (1996a) investigated the biological
availability of sediment-associated divalent metals to A.
abdita and H. azteca in sediments from five saltwater
locations and one freshwater location in the United
States, Canada, and China using 10-day lethality tests.
Sediment toxicity was not related to dry weight metals
concentrations. In the locations where metals might be
likely to cause toxicity, 49 sediments had less SEM than
AVS and <0.5 IWTUs, and no toxicity was observed. In
contrast, one-third of the 45 sediments with more SEM
than AVS and >0.5 IWTUs were toxic (Table 3-1).
Hansen et al. (1996a) made an observation that is
important to interpretation of toxicity of sediments from
field locations, particularly those from industrial
harbors. They observed that if sediments with SEM/
AVS ratios <1.0 are toxic, even if metals concentrations
on a dry weight basis are very high, the toxicity is not
likely to be caused by metals. Furthermore, it is
incorrect to use such data to reach the conclusion that
the EqP approach is not valid. This is because when
SEM/AVS ratios were <1.0, there was an almost
complete absence of toxicity in both spiked sediments
and field sediments where metals were the only known
source of contamination and IWTUs for metals were
O.5. When metals concentrations expressed as the sum
of the IWTUs are used in conjunction with SEM/AVS
ratios, they together provide insight that can explain
apparent anomalies between SEM/AVS ratios <1.0 and
sediment toxicity in field sediments. Joint use of both
SEM/AVS ratios and interstitial water concentrations is
also a powerful tool for explaining absence of toxicity
when SEM/AVS ratios are >1.0. Overall, when
freshwater and saltwater field sediments were tested in
the laboratory, 100% were not toxic when SEM/AVS
was < 1.0 and IWTUs were <0.5, and 66.7% were toxic
3-10
-------
Toxicity of Metals in Sediments
when SEM/AVS was >1.0 and IWTUs were ;
(Table 3-1).
:0.5
Therefore, because AVS can bind divalent metals in
proportion to their molar concentrations, Hansen et al.
(1996a) proposed the use of the difference between the
molar concentrations of SEM and AVS (SEM-AVS)
rather than SEM/AVS ratios used previously. The molar
difference provides important insight into the extent of
additional available binding capacity and the
magnitude by which AVS binding has been exceeded
(Figure 3-5). Further, absence of organism response
when AVS binding is exceeded can indicate the
potential magnitude of other important binding phases
in controlling bioavailability. Figure 3-5 shows that for
most nontoxic freshwater and saltwater field sediments,
1 to 100 ^mol of additional metal would be required to
exceed the sulfide-binding capacity (i.e., SEM-AVS =
-100 to -1 ,wmol/g). In contrast, most toxic field
sediments contained 1 to 1,000 ^mol of metal beyond
the binding capacity of sulfide alone. Data on nontoxic
field sediments whose sulfide-binding capacity is
exceeded (SEM-AVS is > 1.0 ^mol/g) indicate that other
sediment phases, in addition to AVS, have significance
in controlling metal bioavailability. In comparison to
SEM/AVS ratios, use of SEM-AVS differences is
particularly informative where AVS concentrations are
low, such as those from Steilacoom Lake and the
Keweenaw Watershed, where the SEM-AVS difference
is numerically low and SEM/AVS ratios are high
(Ankley et al., 1993). Forthese reasons, SEM-AVS is
used instead of the SEM/AVS ratio almost exclusively
for the remainder of this document.
3.2.4 Field Sites and Spiked Sediments
Combined
Figure 3-6 and Table 3-1 summarize available data
from freshwater and saltwater sediments spiked with
individual metals or metal mixtures, freshwater field
sites, and saltwater field sites on the utility of metals
concentrations in sediments normalized by dry weight,
IWTUs, and SEM-AVS. These data explain the
Ctf
1
100
80
60
40
20
D
o o
o
o
D
AD
A PA * * AV
^A JII A u^X> A A *
AA .A. o 4 /O.^A A* + +
-100
-10
-1 0 1 10
SEM-AVS (^rnol/g dry wt)
100
1000
Figure 3-5. Percentage mortality of amphipods, oligochaetes, and polychaetes exposed to sediments from four
freshwater and three saltwater field locations as a function of the sum of the molar concentrations of
SEM minus the molar concentration of AVS (SEM-AVS). Sediments having s24% mortality are
considered nontoxic as defined by Berry et al. (1996), which is indicated by the horizontal dotted line
in the figure. The vertical dotted line at SEM-AVS = 0.0 /^mol/g dry wt indicates the boundary
between sulfide-bound unavailable metal and potentially available metal. The different symbols
represent field sediments from different locations (figure from Hansen et al., 1996a).
3-11
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
/^
£
S
a
et
c
o
£
g
s
M
"C
1
/^
£
J
ti
i
100
80
60
40
20
0 '
0
100
80
60
40
20
i
0
100
80
60
40
20,
<
O.i
'
j
1 1 1 1 HIM 1 III Mill 1 1 1 1 1 MM 1 1 1 1 HIM 1 1 1 1 MM
A
** o <><» GOOD o^ q> o*&&> 4fco ODO0-
0 o o 0 V %> J*""^ ~ ^
o Spiked Sediment ° o «, ° * o
- O Field Sediment ° <> o* o
° 0
O
O o s>
o v
o 4J,
o 00 °
O ° O^s
g o o o v
" .o-^f^J^iS^^^C''''-
.01 0.1 1 10 100 1
Total Metal or SEM (//mol/g dry wt)
1 1 1 1 HIM 1 III Mill 1 1 1 1 1 MM 1 1 1 1 HIM 1 1 1 1 MM
n
- " o oo4X30 OO9o+ o g oo oocoo.a
O 0 o o o o ,
* ^ o ^ o ° ° !
o O O 0 v O
0 °
0
° 0
1 ^0 o 0
<> O 0 °
O O Q N/ O (
^8 cj&^^^'o"*!
ioog $. iiiiloS^^Mmiu-j^Ai i uj
.01 0.1 1 10 100 1
Interstitial Water Toxic Units
i i i i i i
*-' 'goDoooO g ODO o 0 O$D ^>O o o o -<
I ^O^ o ^ . j
<> 0 OQ O J
1 ° o
o1
* o
It ° ° °° 0
<%OO Dt^MBft ° * o
tV °WD*^^^^^»V W^ M V V ^
50 -25 0 25 50 75 100 1
SEM-AVS
DOO
000
>
\
>
>
*
25
Figure 3-6. Percentage mortality of freshwater and saltwater benthic species in 10-day toxicity tests in spiked
sediments and sediments from the field. Mortality is plotted as a function of: (A) the sum of the
concentrations of the respective metal (Cd, Cu, Pb, Ni, Ag, or Zn) or metal mixture in ;unol metal
per gram dry weight of sediment; (B) IWTU; and (C) SEM-AVS difference. Data below the detection
limits are plotted at IWTU = 0.01 and SEM-AVS = -50 /^mol/g dry wt (silver data from Berry et al.,
1999; all other data modified after Hansen et al., 1996a).
3-12
-------
Toxicity of Metals in Sediments
bioavailability and acute toxicity of metals in sediments
(Hansenetal., 1996a; Berry etal., 1999). This analysis
contains all available data from 10-day lethality tests
where mortality, IWTUs, SEM, and AVS are known from
experiments with sediments toxic only because of
metals. The relationship between benthic organism
mortality and total dry weight metals concentrations in
spiked and field sediments is not useful to causally
relate metal concentrations to organism response
(Figures 3-4A and 3-6 A). The overlap is almost four
orders of magnitude in the bulk metals concentrations
that cause no toxicity and those that are 100% lethal for
these sediments where metals are the only source of
toxicity (see discussion in Section 3.2.2).
Data in Figure 3-6B show that over all tests, the
toxicity of sediments whose concentrations are
normalized on an IWTU basis are typically consistent
with the IWTU concept; that is, if IWTUs are < 1.0,
then sediments should be lethal to <50% of the
organisms exposed, and significant mortality probably
should be absent at <0.5 IWTUs. Of the spiked and
field sediments evaluated that had IWTUs <0.5,97.7%
of 175 sediments were nontoxic (Table 3-1). For the 142
sediments having IWTUs >0.5,68.3% were toxic.
However, and as stated above, given the effect on
toxicity or bioavailability of the presence of other
binding phases (e.g., DOC) in interstitial water, water
quality (hardness, salinity, etc.), and organism behavior,
it is not surprising that many sediments having IWTUs
> 0.5 are not toxic.
Data in Figure 3-6C show that over all tests,
organism response in sediments whose concentrations
are normalized on an SEM-AVS basis is consistent with
metal-sulfide binding on a mole to mole basis as first
described by DiToro etal. (1990), and later
recommended for assessing the bioavailability of
metals in sediments by Ankley et al. (1994). Saltwater
and freshwater sediments either spiked with metals or
from field locations with SEM-AVS differences <0.0
were uniformly nontoxic (98.1% of 158 sediments)
(Table 3-1). The majority (58.0%) of 174 sediments
having SEM-AVS >0.0 were toxic. It is not surprising
that many sediments having SEM-AVS >0.0 are not
toxic given the effect on toxicity or bioavailability of the
presence of other sediment phases that also affect
bioavailability (see Section 3-4; Di Toro et al., 1987,
2000; Mahony etal., 1996).
Over all tests, the data in Figure 3-6 indicate that
use of both IWTUs and SEM-AVS together did not
improve the accuracy of predictions of sediments that
were nontoxic (98.5% of 132 sediments; Table 3-1).
However, it is noteworthy that 78.9% of the 123
sediments with both SEM-AVS >0.0 and IWTUs >0.5
were toxic. Therefore, the approach of using SEM-AVS,
IWTUs, and especially both indicators to identify
sediments of concern is very useful.
The results of all available data demonstrate that
using SEM, AVS, and interstitial water metals
concentrations to predict the lack of toxicity of
cadmium, copper, lead, nickel, silver, and zinc in
sediments is certain. This is very useful, because the
vast majority of sediments found in the environment in
the United States have AVS concentrations that exceed
the SEM concentration (SEM-AVS <0.0) (see Section
4.4). This may incorrectly suggest that there should be
little concern about metals in sediments on a national
basis, even though localized areas of biologically
significant metal contamination do exist (Wolfe et al.,
1994; Hansenetal., 1996a; Leonard etal., 1996a). It is
potentially important that most of these data are from
field sites where sediment samples were collected in the
summer. At this time of year, the seasonal cycles of
AVS produce the maximum metal-binding potentials
(BoothmanandHelmstetter, 1992; Leonard etal., 1993).
Hence, sampling at seasons and conditions when AVS
concentrations are at a minimum is a must in
establishing the true overall level of concern about
metals in the nation's sediments and in evaluations of
specific sediments of local concern.
Predicting which sediments with SEM-AVS >0.0
will be toxic is presently less certain. Importantly, the
correct classification rate seen in these experiments is
high; that is, the accuracy of predicting which
sediments were toxic was 58.0% using the SEM and
AVS alone, 68.3% using IWTUs, and 78.9% using both
indicators. An SEM-AVS >0.0, particularly at multiple
adjacent sites, should trigger additional tiered
assessments. These might include characterization of
the spatial (both vertical and horizontal) and temporal
distribution of chemical concentration (AVS and SEM)
and toxicity, measurements of interstitial water metal,
and toxicity identification evaluations (TIEs). In this
context, the combined SEM-AVS and IWTU approach
should be viewed as only one of the many sediment
evaluation methodologies.
3.2.5 Conclusions from Short-Term Studies
Results from tests using sediments spiked with
metals and sediments from the field in locations where
toxicity is associated with metals demonstrate the value
of explaining the biological availability of metals
3-13
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
concentrations normalized by SEM-AVS and IWTUs
instead of dry weight metal concentrations.
Importantly, data from spiked-sediment tests strongly
indicate that metals are not the cause of most of the
toxicity observed infield sediments when both SEM-
AVS is <0.0 and IWTUs are <0.5 (Table 3-1). Expressing
concentrations of metals in sediments on an SEM-AVS
basis provides important insight into the available
additional binding capacity of sediments and the extent
to which sulfide binding has been exceeded.
SEM-AVS and interstitial water concentrations of
metals can aid in identifying the specific metal causing
toxicity. For example, the metal(s) in excess of AVS can
be identified by subtracting from the molar
concentration of AVS the molar concentrations of
specific metals in the SEM in order of their sulfide
solubility product constants (K 2) in the SEM.
Alternatively, interstitial water concentrations of metals
can be used to identify a specific metal causing
sediment toxicity using the toxic unit concept, if
appropriate water-only toxicity data for the tested
species are available (Hansenetal., 1996a).
Predictions of sediments not likely to be toxic,
based on use of SEM-AVS and IWTUs for all data from
freshwater or saltwater field sediment and spiked-
sediment tests, are extremely accurate (98.5%) using
both parameters. Predictions of sediments likely to be
toxic are less accurate. Nevertheless, SEM-AVS is
extremely useful in identifying sediments of potential
concern. Data were summarized from amphipod tests
using freshwater and saltwater laboratory metals-
spiked sediments and field sediments where metals
were a known problem by comparing the percentage of
sediments that were toxic with the SEM-AVS
concentration (tests with polychaetes and gastropods
were excluded because these organisms avoid
exposure) (Hansen, 1995). Seventy percent of the
sediments in these amphipod studies with an SEM-AVS
concentration of > 0.76 ^mol of excess SEM/g were
toxic. The corresponding values for 80%, 90%, and
100% of the sediments being toxic were 2.7,16, and 115
^mol of excess SEM/g, respectively.
Of course, SEM, AVS, and IWTUs can only predict
toxicity or the lack of toxicity caused by metals in
sediments. They cannot be used alone to predict
toxicity of sediments contaminated with toxic
concentrations of other contaminants. However, SEM
and AVS have been used in sediment assessments to
rule out metals as probable causative agents of toxicity
(Wolfe etal., 1994). Also, the use of SEM and AVS to
predict biological availability and toxicity of cadmium,
copper, lead, nickel, silver, and zinc is applicable only to
anaerobic sediments that contain AVS; binding factors
other than AVS control bioavailability in aerobic
sediments (DiToro etal., 1987; Tessier etal., 1993).
Measurement of interstitial water metal may be useful
for evaluations of these and other metals in aerobic and
anaerobic sediments (Ankley etal., 1994). Even with
these caveats, the combined use of SEM, AVS, and
interstitial measurements is preferable to all other
currently available sediment evaluation procedures to
causally assess the implications to benthic organisms
of these six metals associated with sediments (see
discussion in Section 5, Sampling and Analytical
Chemistry, for further guidance).
3.3 Predicting Metal Toxicity:
Long-Term Studies
Taken as a whole, the short-term laboratory
experiments with metal-spiked and field-collected
sediments present a strong argument for the ability to
predict the absence of metal toxicity based on sediment
SEM and AVS relationships and/or interstitial water
metal concentrations. However, if this approach is to
serve as a valid basis for ESB derivation, comparable
predictive success must be demonstrated in long-term
laboratory and field experiments where chronic effects
could be manifested (Luoma and Carter, 1993; Meyer et
al., 1994). This demonstration was the goal of
experiments described by Hare et al. (1994), DeWitt et
al. (1996), Hansen etal. (1996b), Liber etal. (1996), and
Sibley etal. (1996). An important experimental
modification to these long-term studies, as opposed to
the short-term tests described in Section 3.2, was the
collection of horizon-specific chemistry data. This is
required because AVS concentrations often increase,
and SEM-AVS differences decrease, with an increase in
sediment depth (Howard and Evans, 1993; Leonard et
al., 1996a); hence, chemistry performed on homogenized
samples might not reflect the true exposure of benthic
organisms dwelling in surficial sediments (Luoma and
Carter, 1993; Hare etal., 1994; Peterson etal., 1996).
3.3.1 Life-Cycle Toxicity Tests
DeWitt et al. (1996) conducted an entire life-cycle
toxicity test with the marine amphipod L. plumulosus
exposed for 28 days to cadmium-spiked estuarine
sediments (Table 3-2). The test measured effects on
survival, growth, and reproduction of newborn
amphipods relative to interstitial water and SEM/AVS
3-14
-------
Toxicity of Metals in Sediments
Table 3-2. Summary of the results of full life-cycle and colonization toxicity tests conducted in the laboratory and
field using sediments spiked with individual metals and metal mixtures
Measured SEM-AVSa
Toxicity Test
Life Cycle:
Leptocheirus
plumulosus
Chironomus
tentans
Dura-
tion
Metal(s) (days)
Cadmium 28
Zinc 56
Omol/g)
NOEC(s)b
-3.5, -2.0,
0.78,2.0
-2.6, -1.4,
6.4
OEC(s)C Effect
8.9, 15.6 Mortality 100%
21 .9, 32.4 Larval mortality 85%-
100%
Reference
DeWittetal., 1996
Sibley etal., 1996
Colonization:
Laboratory-
saltwater
Weight, emergence, and
reproduction reduced
Cadmium 118 -13.4 8.0,27.4 Fewer polychaetes, shifts
in community
composition, fewer
species, bivalves absent,
tunicates increased
Hansen et al.,
1996b
Field-saltwater
Field-freshwater
Field-freshwater
Cadmium,
copper,
lead,
nickel, zinc
Cadmium
Zinc
120 -0.31,
-0.06,
0.02
-365 -0.07,0.08,
0.34
368 -3.6, -3.5,
-2.9,
-2.0, 1.0d
No effects observed
2.2 Reduced Chironomus
salinarius numbers
Bioaccumulation
No effects observed
Boothman et al.,
2001
Hare etal., 1994
Liber etal., 1996
aSEM-AVS differences are used instead of SEM/AVS ratios to standardize across the studies referenced. An SEM-AVS difference of
<0.0 is the same as an SEM/AVS ratio of <1.0. An SEM-AVS difference of >0.0 is the same as an SEM/AVS ratio of >1.0.
NOECs = no observed effect concentration(s); all concentrations where response was not significantly different from the control.
cOECs = observed effect concentration(s); all concentrations where response was significantly different from the control.
Occasional minor reductions in oligochaetes (Naididae).
normalization. Seven treatments of Cd were tested: 0
(control), -3.5, -2.0,0.78,2.0,8.9, and 15.6 SEMcd-AVS
differences (measured concentrations). Gradients in
AVS concentration as a function of sediment depth
were greatest in the control treatment, decreased as the
SEMcd ratio increased, and became more pronounced
over time. Depth gradients in SEMcd-AVS differences
were primarily caused by the spatial and temporal
changes in AVS concentration, because SEMcd
concentrations changed very little with time or depth.
Thus in most treatments SEMcd-AVS differences were
smaller at the top of sediment cores than at the bottom.
This is expected because the oxidation rate of iron
sulfide in laboratory experiments is very rapid (100% in
60 to 90 minutes) but for cadmium sulfide it is slow
(10% in 300 hours) (Mahony et al., 1993; Di Toro et al.,
1996a). Interstitial cadmium concentrations increased
in a dramatic stepwise fashion in treatments having a
SEM-AVS difference of > 8.9 ,wmol of excess SEM, but
were below the 96-hour LC50 value for this amphipod in
lesser treatments. There were no significant effects on
survival, growth, or reproduction in sediments
containing more AVS than cadmium (-3.5 and -2.0
,wmol/g) and those with a slight excess of SEMcd (0.78
and 2.0 ^mol/g), in spite of the fact that these samples
contained from 183 to 1,370 /j,g cadmium/g sediment.
All amphipods died in sediments having SEM-AVS
differences > 8.9 ,wmol excess SEM/g. These results are
3-15
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
consistent with predictions of metal bioavailability from
10-day acute tests with metal-spiked sediments (i.e.,
that sediments with SEMcd-AVS differences <0.0 are
not toxic, interstitial water metal concentrations are
related to organism response, and sediments with
SEMcd-AVS differences >0.0 may be toxic).
Sibley et al. (1996) reported similar results from a
56-day life-cycle test conducted with the freshwater
midge C. tentans exposed to zinc-spiked sediments
(Table 3 -2). The test was initiated with newly hatched
larvae and lasted one complete generation, during
which survival, growth, emergence, and reproduction
were monitored. In sediments where the molar
difference between SEM and AVS (SEM-AVS) was <0.0
(dry weight zinc concentrations were as high as 270
mg/kg), concentrations of zinc in the sediment
interstitial water were low and no adverse effects were
observed for any of the biological endpoints measured.
Conversely, when SEM-AVS was 21.9 and 32.4 ^trnol of
excess SEM/g, interstitial water concentrations of zinc
increased (being highest in surficial sediments), and
reductions in survival, growth, emergence, and
reproduction were observed. Over the course of the
study, the absolute concentration of zinc in the
interstitial water in these treatments decreased because
of the increase in sediment AVS and loss of zinc from
twice-daily renewals of the overlying water.
3.3.2 Colonization Tests
Hansen et al. (1996b) conducted a 118-day benthic
colonization experiment in which sediments were spiked
to achieve nominal cadmium/AVS molar ratios of 0.0
(control), 0.1,0.8, and 3.0 and then held in the
laboratory in a constant flow of unfiltered seawater
(Table 3-2). Oxidation of AVS in the surficial 2.4 cm of
the control treatment occurred within 2 to 4 weeks and
resulted in sulfide profiles similar to those occurring in
sediments in nearby Narragansett Bay, RI (Boothman
andHelmstetter, 1992). In the nominal 0.1 cadmium/
AVS treatment, measured SEMcd was always less than
AVS (SEM-AVS = -13.4 ,wmol AVS/g in the surficial 2.0
cm), interstitial cadmium concentrations (<3 to 10 ^g/L)
were less than those likely to cause biological effects,
and no significant biological effects were detected. In
the nominal 0.8 cadmium/AVS treatment (SEM-AVS =
8.0 ^mol SEM/g), measured SEMQd commonly exceeded
AVS in the surficial 2.4 cm of sediment, and interstitial
cadmium concentrations (24 to 157 ,wg/L) were
sufficient to be of toxicological significance to highly
sensitive species. In this treatment, shifts in the
presence or absence of organisms were observed over
all taxa, and there were fewer macrobenthic polychaetes
(Mediomastus ambiseta, Streblospio benedicti, and
Podarke obscura) and meiofaunal nematodes. In the
nominal 3.0 cadmium/AVS treatment (SEM-AVS of 27.4
^mol SEM/g), concentrations of SEMcd were always
greater than AVS throughout the sediment column.
Interstitial cadmium ranged from 28,000 to 174,000 ^g/L.
In addition to the effects observed in the nominal 0.8
cadmium/AVS treatment, the following effects were
observed: (a) sediments were colonized by fewer
macrobenthic and polychaete species and
harpacticoids, (b) the sediments had lower densities of
diatoms, and (c) bivalve molluscs were absent. Over all
treatments, the observed biological responses were
consistent with predicted possible adverse effects
resulting from elevated SEMcd-AVS differences in
surficial sediments and interstitial water cadmium
concentrations.
Boothman et al. (2001) conducted a field
colonization experiment in which sediments from
Narragansett Bay, RI, were spiked with an equimolar
mixture of cadmium, copper, lead, nickel, and zinc at
nominal SEM/AVS ratios of 0.1,0.8, and 3.0; placed in
boxes; and replaced in Narragansett Bay (Table 3-2).
The AVS concentrations decreased with time in surface
sediments (0 to 3 cm) in all treatments where the
nominal SEM/AVS ratio was <1.0 (SEM-AVS decreased
from - 0.31 to - 0.06 ^mol SEM/g in the surficial 2.0 cm)
but did not change in subsurface (6 to 10 cm)
sediments or in the entire sediment column where
nominal SEM/AVS ratios exceeded 1.0 (SEM-AVS = 0.02
^mol AVS/g). SEM decreased with time only where
SEM exceeded AVS. The concentration of metals in
interstitial water was below detection limits when there
was more AVS than SEM. When SEM exceeded AVS,
significant concentrations of metals were present in
interstitial water, and appeared in the order of their
sulfide solubility product constants. Interstitial water
concentrations in these sediments decreased with time,
although they exceeded the WQC in interstitial water
for 60 days for all metals, 85 days for cadmium and zinc,
and 120 days for the entire experiment for zinc. Benthic
faunal assemblages in the spiked-sediment treatments
were not different from those of the control treatment.
Lack of biological response was consistent with the
vertical profiles of SEM and AVS. AVS was greater
than SEM in all surface sediments, including the top 2
cm of the 3.0 nominal SEM/AVS treatment, because of
oxidation of AVS and loss of SEM. The authors
speculated that interstitial metal was likely absent in the
surficial sediments in spite of data demonstrating the
presence of significant measured concentrations.
3-16
-------
Toxicity of Metals in Sediments
Interstitial water in the 3.0 nominal SEM/AVS treatment
was sampled from sediment depths where SEM was in
excess, rather than in the surficial sediments. Important
to the biological data are the surficial sediments, where
settlement by saltwater benthic organisms first occurs.
Also, there was a storm event that allowed a thin layer
of clean sediment to be deposited on top of the spiked
sediment (W.S. Boothman, U.S. EPA, Narragansett, RI,
personal communication). These data demonstrate the
importance of sampling sediments and interstitial water
in sediment horizons where benthic organisms are
active.
Hare et al. (1994) conducted an approximately
1 -year field colonization experiment in which
uncontaminated freshwater sediments were spiked with
cadmium and replaced in the oligotrophic lake from
which they originally had been collected (Table 3-2).
Cadmium concentrations in interstitial waters were very
low at cadmium-AVS molar differences <0.0, but
increased markedly at differences >0.0. The authors
reported reductions in the abundance of only the
chironomid Chironomus salinarius in the 2.2 ,wmol
excess SEM/g treatment. Cadmium was accumulated by
organisms from sediments with surficial SEM
concentrations that exceeded those of AVS. These
sediments also contained elevated concentrations of
cadmium in interstitial water.
Liber et al. (1996) performed a field colonization
experiment using sediments having 4.46 ,wmol of sulfide
from a freshwater mesotrophic pond (Table 3-2).
Sediments were spiked with 0.8,1.5,3.0,6.0, and 12.0
^mol of zinc, replaced in the field, and chemically and
biologically sampled over 12 months. There was a
pronounced increase in AVS concentrations with
increasing zinc concentration; AVS was lowest in the
surficial 0 to 2 cm of sediment with minor seasonal
variations. With the exception of the highest spiking
concentration (approximately 700 mg/kg, dry weight),
AVS concentrations remained larger than those of SEM.
Interstitial water zinc concentrations were rarely
detected in any treatment, and were never at
concentrations that might pose a hazard to benthic
macroinvertebrates. The only observed difference in
benthic community structure across the treatments was
a slight decrease in the abundance of Naididae
oligochaetes at the highest spiking concentration. The
absence of any noteworthy biological response was
consistent with the absence of interstitial water
concentrations of biological concern. The lack of
biological response was attributed to an increase in
concentrations of iron and manganese sulfides
produced during periods of diagenesis, which were
replaced by the more stable zinc sulfide, which is less
readily oxidized during winter months. In this
experiment, and theoretically in nature, excesses of
sediment metal might be overcome over time because of
the diagenesis of organic material. In periods of
minimal diagenesis, oxidation rates of metal sulfides, if
sufficiently great, could release biologically significant
concentrations of the metal into interstitial waters. The
phenomenon should occur metal by metal in order of
their sulfide solubility product constants.
3.3.3 Conclusions from Chronic Studies
Over all full life-cycle and colonization toxicity
tests conducted in the laboratory and field using
sediments spiked with individual metals and metal
mixtures (Table 3-2), no sediments with an excess of
AVS (SEM-AVS < 0.0) were toxic (Figure 3-7).
Conversely, all sediments where chronic effects were
observed, and 7 of 19 sediments where no effects were
observed, had an excess of SEM (SEM-AVS >0.0)
(Table 3-2; Figure 3-7). Therefore, the results from all
available acute and chronic toxicity tests support the
use of SEM-AVS < 0.0 as an ESB that can be used to
predict sediments that are unlikely to be toxic.
3.4 Predicting Toxicity of Metals in
Sediments
3.4.1 General Information
The SEM-AVS method for evaluating toxicity of
metals in sediments (Di Toro et al., 1990,1992) has
proven to be successful at predicting the lack of metal
toxicity in spiked and field-contaminated sediments
(Berry etal., 1996;Hansenetal., 1996a). However,
because SEM-AVS does not explicitly consider the
other sediment phases that influence interstitial water-
sediment partitioning, and in spite of its utility in
identifying sediments of possible concern, it was never
intended to be used to predict the occurrence of
toxicity. The proposed sediment quality criteria for
metals using SEM, AVS, and IWTUs in Ankley et al.
(1996)now referred to as ESBs or equilibrium
partitioning sediment benchmarkswere constructed
as "one-tailed" guidelines. They should be used to
predict the lack of toxicity but not its presence. Thus
the problem of predicting the onset of toxicity in metal-
contaminated sediments remained unsolved.
This section introduces a modification of the SEM-
AVS procedure in which the SEM-AVS difference is
3-17
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Predicted
Toxicity Uncertain
D
Experimental OEC
Experimental NOEC
_L
-20
-10
o 10
SEM-AVS (
20
30
40
Figure 3-7. Comparison of the chronic toxicity of sediments spiked with individual metals or metal mixtures to
predicted toxicity based on SEM-AVS (data from Table 3-2). Horizontal dashed line separates
experimental observed effect concentrations (solid columns) from no observed effect concentrations
(shaded columns). Values at SEM-AVS < 0.0 ^mol/goc are predicted to be nontoxic. Values at SEM-AVS
>0.0 ^mol/goc are indicative of sediments that are likely to be toxic or toxicity is uncertain.
3-18
-------
Toxicity of Metals in Sediments
normalized by the fraction of organic carbon, foc, in a
sediment. This section is largely taken from Di Toro et
al. (2000). Their publication should be consulted for
additional information about the utility of the/oc
procedure and comparison of this procedure with the
sediment guidelines of Long et al. (1995a) and
MacDonaldetal. (1996). The (ZSEM-AVS)//OC
procedure significantly improves prediction of mortality
by accounting for partitioning of metals to sediment
organic carbon, as well as the effect of AVS. In
addition, the approach used by Di Toro et al. (2000) to
derive (ZSEM-AVS)//OC uncertainty bounds for
identifying sediments that are likely to be toxic, are of
uncertain toxicity, or are nontoxic has applicability to
SEM/AVS ratios, SEM-AVS differences, andlWTUs.
Although not used as an ESB, the uncertainty bounds
should be useful in prioritizing sediments of concern
for further evaluations.
3.4.2 EqP Theory for SEM, AVS, and
Organic Carbon
The EqP model provides for the development of
causal sediment concentrations that predict toxicity or
lack of toxicity in sediments (Di Toro etal., 1991). The
sediment concentration Cs that corresponds to a
measured LC50 in a water-only exposure of the test
organism is
C* = K LC50
s p
(3-2)
where Cs* is the sediment LC50 concentration (jug/kg
dry wt), K (L/kg) is the partition coefficient between
interstitial water and sediment solids, and LC50 is the
concentration causing 50% mortality Cwg/L). For
application to metals that react with AVS to form
insoluble metal sulfides, Equation 3-2 becomes
Cs*=AVS+£pLC50
(3-3)
where AVS is the sediment concentration of acid
volatile sulfides. Equation 3-3 simply states that
because AVS can bind the metal as highly insoluble
sulfides, the concentration of metal in a sediment that
will cause toxicity is at least as great as the AVS that is
present. The sediment metal concentration that should
be employed is the SEM concentration, because any
metal that is bound so strongly that IN of hydrochloric
acid cannot dissolve it is not likely to be bioavailable
(Di Toro et al., 1992). Of course, this argument is
theoretical, which is why so much effort has been
expended to demonstrate experimentally that this is
actually the case (Di Toro et al., 1992; Hare et al., 1994;
Berry etal., 1996; Hansen etal., 1996a; Sibley etal.,
1996). Therefore, the relevant sediment metal
concentration is SEM, and Equation 3-3 becomes
(34)
The basis for the AVS method is to observe that if
the second term in Equation 3-4 is neglected, then the
critical concentration is SEM = AVS, and the criterion
for toxicity or lack of toxicity is SEM-AVS < 0.0 (amol/g
dry wt).
The failure of the difference to predict toxicity
when there is an excess of SEM is due to neglect of the
partitioning term KLC50. Note that ignoring the term
does not affect the prediction of lack of toxicity in that
it makes the condition conservative (i.e., smaller
concentrations of SEM are at the boundary of toxicity
and no toxicity).
The key to improving prediction of toxicity is to
approximate the partitioning term rather than ignore it
(Di Toro et al, in prep.). In sediments, the organic
carbon fraction is an important partitioning phase, and
partition coefficients for certain metals at certain pHs
have been measured (Mahony etal., 1996). This
suggests that the partition coefficient K in Equation 3-
4 can be expressed using the organic carbon-water
partition coefficient, KQC, together with the fraction
organic carbon in the sediment, foc
Tf = f Tf C\ *\\
P Joe Jvoc vj"~v
Using this expression in Equation 3-4 yields
SEM = AVS +/oc Koc LC50 (3-6)
Moving the known terms to the left side of this
equation yields
SEM-AVS = K LC5Q (3-7)
foc
If both KQC and LC50 are known, then Equation 3 -7 can
be used to predict toxicity.
The method evaluated below uses (2SEM-AVS)/
foc as the predictor of toxicity and evaluates the critical
concentrations (the right side of Equation 3-7) based
on observed SEM, AVS, /oc, and toxicity data (Di Toro
et al, in prep.). If multiple metals are present, it is
necessary to use the total SEM
2SEM=2 [SEM,.]
(3-8)
3-19
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
to account for all the metals present. Note that (ZSEM-
AVS)//OC is the organic carbon-normalized excess SEM
for which we use the notation
(£SEM-AVS)//OC =
, SEM-AVS
/c
'oc
(3-9)
3.4.3 Data Sources
Data from toxicity tests using both laboratory-
spiked and field-collected sediments were compiled
from the literature. Four sources of laboratory-spiked
tests using marine sediments (Casas and Crecelius,
1994;Peschetal., 1995; Berry etal., 1996,1999)andone
using freshwater sediments (Carlson etal., 1991) were
included. Two sources for metal-contaminated field
sediments were included (Hansen et al., 1996a; Kemble
etal., 1994). The field data from the sediments where
metals were not the probable cause of toxicity (Bear
Creek and Jinzhou Bay) (Hansen et al., 1996a) were
excluded. Data reported included total metals, SEM,
AVS, foc, and 10- or 14-day mortality. In Hansen etal.
(1996a), data were reported for five saltwater and four
freshwater locations, but organic carbon
concentrations were not available for freshwater field
sediments from three locations. Organic carbon data
for the Keweenaw Watershed were obtained separately
(E.N. Leonard, U.S. EPA, Duluth, MN, personal
communication).
Laboratory-spiked and field sediment data were
grouped for analysis. Mortality data were compared
against the SEM-AVS difference and the SEM-AVS
difference divided by thefoc. For each comparison,
two uncertainty bounds were computed: a lower-bound
concentration equivalent to a 95% chance that the
mortality observed would be less than 24% (the
percentage mortality considered to be toxic) (see Berry
et al., 1996) and an upper-bound concentration
equivalent to a 95% chance that the observed mortality
would be greater than 24%. The lower-bound
uncertainty limit was computed by evaluating the
fraction of correct classification starting from the
lowest x-axis value. When the fraction correct dropped
to below 95%, the 95thpercentile was interpolated.
The same procedure was applied to obtain the upper-
bound uncertainty limit. These uncertainty bounds are
the concentration range where it is 90% certain that the
sediment may be either toxic or not toxic.
3.4.4 Acute Toxicity Uncertainty
Mortality in the laboratory-spiked and field-
contaminated sediment tests were both organism and
metal independent when plotted against the SEM-AVS
difference (Figure 3-8A). The horizontal dashed line
indicating 24% mortality is shown for reference. The
90% lower and upper uncertainty bound limits for the
SEM-AVS difference are from 1.7 and 120 ^mol/g, a
factor of 70. Thus, it appears that for both laboratory
and spiked-sediment data, toxicity is likely when the
SEM-AVS difference is >120, uncertain when the
difference is from 1.7 to 120 ^mol/g, and not likely when
the difference is < 1.7 ^mol/g.
Although use of SEM-AVS differences to predict
toxicity is not based on any theoretical foundation, use
of SEM-AVS < 0.0 to predict lack of toxicity is based on
the equilibrium partitioning model (DiToro etal., 1991)
and the chemistry of metal-sulfide interactions. The
stoichiometry of the uptake of divalent metals by AVS
is such that 1 mol of AVS will stabilize 1 mol of SEM,
except for silver, where the ratio is 2:1, hence the use of
the difference of 0.0 ^mol/g dry weight to predict lack
of toxicity. In fact it is the very low solubility of the
resulting metal sulfides that limits the interstitial water
concentrations to below toxic levels regardless of the
details of the sediment chemistry (e.g., pH, iron
concentration) as has been demonstrated in this
document and detailed in the Appendix in Di Toro et al.
(1992).
The (2SEM-AVS)/foc approach provides an
equivalent theoretical basis that is needed to derive an
appropriately normalized sediment concentration that
predicts occurrence of toxicity that is causally linked to
bioavailable metal. When percent mortality is plotted
against the organic carbon-normalized excess SEM
(ZSEM-AVS)/foc) for the same data as contained in
Figure 3-8 A, toxicity is likely when the (2SEM-AVS)/foc
is >3,000 ,wmol/goc, uncertain when the concentration is
between 130 and 3,000 ,wmol/goc, and not likely when
the concentration is <130 ,wmol/goc (Figure 3-8B). Thus,
the width of the uncertainty bound is a factor of 70 for
SEM-AVS differences and 23 for (2SEM-AVS)/foc.
If the (2SEM-AVS)/foc approach improves
predictions of sediment toxicity caused by metals, the
uncertainty bounds should narrow and the percentages
of sediments where toxicity predictions are uncertain
should decrease. If the uncertainty bound analysis is
not conducted, and SEM-AVS>0.0 is used as proposed
in Sections 3.2 and 3.3, predictions of sediment toxicity
3-20
-------
Toxicity of Metals in Sediments
100
80
0^
& 60
1 40
20
0
100
80
Mortality (%
S § §
0
-100
IIIMI i i IIIMI 1 1 i IIIMI i i i mini i i mull 1 1 1
A
-
-
v
n D "^ gpvr2v
&fejg^W
MwUE*
IIIMI i i HIM ii i i mini i i mini i i mini | | |
Ml II IMIII 1 1 MINIM 1 II IMIII 1 II IIIMI 1 II IIIMI 1 1 IIIMI
1
1
"" vv.;-- 0
v JD °
. 1
1
* 1
. n 1
£ v. ^ *1D v
^7 £n* '
AlB
1 1
ill ii IMIII i ii mini i ii IMIII i iiiiini i iiiinii i iiinii
SEM-AVS Cumol/g)
MM 1 1 1 1 HIM 1 1 1 1 Mill 1 1 1 1 IMIII 1 1 1 IIIMI 1 1 1 1
B
v
.B.5?i fl *
*W"J^B
HIM i i i HUM i i i MINI i i i mill i i i nun i i i i
1 1 1 1 IIIMI 1 1 1 1 Hill 1 1 1 1 1 Mill 1 1 1 1 Mill 1 1 1 1 Mill 1 1 1 HIM
1 1
1 1
1 v "* , v
i v i ;..
'. v o- a
l v I
1 1.
1 1
1 D 'V «
1 1
i n i *
v i v v a.di .
D a ^ ' c.'^
i
I lllllllll I lllliiiil i ilium i iiiini i iiiinii i iiinii
,000 -10,000 -1,000 -100 -10 -101 10 100 1,000 10,000 100,000 1,000,000
(SSEM-AVS)//OC Cwmol/goc)
Figure 3-8. Percent mortality versus SEM-AVS (A) and (SSEM-AVS)//"^ (B) for saltwater field data without Bear
Creek and Jinzhou Bay (D), freshwater field data (v), freshwater spiked data (- ), and saltwater spiked
data (); silver data excluded. Vertical dashed lines are the 90% uncertainty bound limits (figure from
Di Toro et al., 2000).
3-21
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
for the 267 spiked sediments are classified as uncertain
for 47.2% of the sediments. Using the uncertainty
bounds on SEM-AVS of 1.7 to 120 ^mol/g as described
in this section results in reduction in the percentage of
sediments where toxicity predictions are uncertain, to
34. 1%. Use of (ZSEM-AVS)/foc with uncertainty
bounds of 130 to 3,000 ,wmol/goc results in further
reduction in the percentage of sediments where toxicity
predictions are uncertain, to 25. 5%. Therefore, use of
the uncertainty limits of the (ZSEM-AVS)/foc approach
classifies 3 3 .7% more sediments as toxic or nontoxic
than using the uncertainty limits of SEM-AVS, and 85%
more than use of SEM-AVS without uncertainty limits.
This improvement highlights the advantages of using
(Z SEM- AVS)/foc in assessing toxicity of metal-
contaminated sediments.
c uncertainty limits applies
to all the metals regardless of their identity. Figure 3-9
presents the spiked-sediment data categorized by
identity of the metal. The field-contaminated data
cannot be included because the identity of the metal
causing toxicity cannot be unambiguously determined.
There is no apparent difference for any of the metals in
the region of overlapping survival and mortality data
between 130 and 3,000 ^mol/goc.
It is interesting to note that organic carbon
normalization appears not to work for silver. The
spiked-sediment test data are presented in Figure 3-10A
(Berry etal., 1999). Note that there is almost a complete
overlap of mortality and no mortality data. This
suggests that organic carbon is not a useful
normalization for silver partitioning in sediments.
Perhaps this is not surprising because the role of sulfur
groups is so prominent in the complexation chemistry
of silver (Bell and Kramer, 1 999).
To not depend on the identity of the metal is an
advantage in analyzing naturally contaminated
sediments in that it is difficult to decide which metal is
potentially causing the toxicity. Of course it can be
done using the sequence of solubilities of the metal
sulfides or interstitial metal concentrations (Di Toro et
al., 1992; Ankley etal., 1996). The metal-independent
method can be tested using the results of an experiment
with an equimolar mixture of cadmium, copper, nickel,
and zinc (see Figure 3 - 1 OB) . The area of uncertainty
falls within the carbon-normalized excess SEM
boundaries above.
3.4.5 Chronic Toxicity Uncertainty
The results of chronic toxicity tests with metals-
spiked sediments can also be compared to (2SEM-
AVS)//OC (Figure 3-11;Table 3-3). NotethatFigure 3-11
indicates a category for "predicted toxic." Significant
chronic effects were observed in only 1 of the 19
sediments, where the uncertainty analysis of acute
toxicity tests indicated that effects were not expected
at (2SEM-AVS)//OC <130 ^mol/goc. The concentration
in the sediment where chronic effects were observed
but not expected, i.e., (2SEM-AVS)//OC = 28 ,wmol
excess SEM/goc. The previous analysis of the results
of chronic toxicity tests using SEM-AVS indicated that
concentrations of SEM exceeded AVS in 7 of 19
nontoxic sediments. Sediment concentrations based
on (ZSEM-AVS)//OC placed these sediments in the
uncertain toxicity category. Importantly, use of (2 SEM-
AVS )//oc to classify sediments resulted in six of these
same seven sediments being correctly classified as
probably nontoxic. Chronic effects were observed in
six of the seven sediments where predictions of effects
are uncertain (130 to 3,000 ^mol/goc). This suggests
that chronic toxicity tests with sensitive benthic
species will be a necessary part of the evaluations of
sediments predicted to have uncertain effects.
3.4.6 Summary
The uncertainty bounds on SEM-AVS differences
and organic carbon-normalized excess SEM ((2SEM-
AVS)//OC) can be used to identify sediments that are
likely to be toxic, are of uncertain toxicity, or are
nontoxic. Use of (2SEM-AVS)/foc as a correction
factor for excess SEM is attractive because it is based
on the theoretical foundation of equilibrium
partitioning. Likewise, it reduces the uncertainty of the
prediction of toxicity over that of SEM-AVS
differences.
3-22
-------
Toxicity of Metals in Sediments
120
100
^ 80
£
& 60
1 40
M
* 20
0
-20
n\
'.
-
~
-
~
_
..
H i HUH i inn i HIIII 1 1
Copper
o
"O"
2°
%v
HI bum b bnml
1 1 IIIIH 1 HUM I 1 IIIIH 1 HIM 1 IIIIH 1 MM
1
Wl| ODD OO
1
1
1
1
1
1
|
1 "
1
1
1 d d d lllllld d Mil.
120
100
80
60
40
20
0
INI 1 HIIII 1 HUM 1 HIM II 1
Lead
-
~
-
_ _
O TO
- vv
.
MI i bun i bmi i bum 1
1 1 HIM 1 IIIHI H MUM 1 IIIIH 1 IIIIH 1 Illlll
1
V| 0
o
-
o
O
. .
V (S j
i
^
i
1 d dl d i md d
-10,000 -100 0 100 10,000 1,000,000 -10,000 -100 0 100 10,000 1,000,000
120
100
^ 80
^,
f60
40
o
* 20
0
-20
i
-
-
-
_
(SEMCn-AVS)//oc (Mmol/goc)
II 1 HIIII 1 HIIlD Hill 1 1
Cadmium
;^'
120
100
-3> 80
^
£> 60
"3
"C 40
|
20
0
-20
1 1 HIM 1 IIIIH I 1 MUM 1 IIIHI 1 IIIIH 1 MM
1
J CD D
' cfP
1
T D
1
O 1
0
1
t> D|
V
1 1
(SEMcd-AVS)//oc Gumol/goc)
I I
Zinc | v i o o
i
Y o
i
i
i
10
i
"d9 10 "
Oo i
\o(A/ ^7 ~
1
HIM bill M bill ii bin M IIIIIH! i nnidl i mid i mid i IIIIH] i MM
W^KcO-W
(SEMPb-AVS)//oc Gumol/goc)
120
100
80
60
40
20
0
-20
HIM HIM HIM Illlll
- Nickel
-
-
_
-
-
-«,°o
III 1 HUH 1 HUM 1 Lllllll I
IIIIIM 1 IIIHI II IIIIIH IIIIH IMIIH 1 Illlll
1
d.0°°o°n -
1 ^
i
i
i
1 ^
_
1
i
i
0 '
cPim
i
1 J Ji J iinJ J
(SEMNi-AVS)//oc G"mol/goc)
Figure 3-9. Percent mortality versus (SEMj^^-AVS)//^ for each metal in spiked sediment tests using Ampelisca
(o), Capitella (v), Neanthes (D), Lumbriculus (), and Helisoma (»). Vertical dashed lines are the 90%
uncertainty bound limits (figure from Di Toro et al., 2000).
3-23
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
120
100
SO
60
40
20
-20
. A: Silver <
-
0 0
00 <
:°8 o°
mi i i i i Inn i i i i Inn i i i i Inn 1 1 i i 1
i i i 1 1 inn i i i 1 1 mi
> od
i
1 1 1
1 1 1 1 1 INI 1
DO
0
3
0
O
1 ,
1 1 1 INI 1 1 1 1 1 INI 1 1 1 1 1 III
O
"
_
"
_
(SEMA-AVS)//OC Cuniol/goc)
C^-
5
"eS
1ZU
100
SO
60
40
20
0
-^0
- B: Mixture
-
-
-
* V **
mi i i i i Inn MI i Inn i i i i Inn 1 1 i i 1
i i i 1 1 inn i i i i inn
~l
1 1 1
1 1 1 1 1 INI 1
0
*
1 ,
*
-
-
-
-
1 1 Mill 1 III Illll 1 1 1 1 III!
-10,000 -1,000 -100 -10 0 10 100 1,000 10,000 100,000 1,000,000
(SSEM-AVS)//ocCumol/goc)
Figure 3-10. Percent mortality versus (SEMAg-AVS)//oc for silver (A) and (3JSEM-AVS)//"OC for a mixture experiment
using Cd, Cu, Ni, and Zn (B; see Berry et al., 1996). Vertical dashed lines are the 90% uncertainty
bound limits determined from Figure 3-8B (figures from Di Toro et al., 2000).
3-24
-------
Toxicity of Metals in Sediments
Predicted
Toxic
Predicted Nontoxic
> [Uncertain
Experimental OEC
D Experimental NOEC
-10000 -1000 -100 -10
10 100 1000 10000
(SEM-AVS)//OC Cumol/goc)
Figure 3-11. Comparison of the chronic toxicity of sediments spiked with individual metals or metal mixtures
to predicted toxicity based on (SEM-AVS)//^ (data from Table 3-3). Horizontal dashed line
separates experimental observed effect concentrations (solid columns) from no observed effect
concentrations (shaded columns). Values at (SEM-AVS)//^ s 130 /^mol/goc are predicted to be
nontoxic. Values between 130 and 3,000 /^mol/goc lie where the prediction of toxicity is uncertain,
and values greater than 3,000 ,umol/goc are predicted to be toxic.
3-25
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Table 3-3. Test-specific data for chronic toxicity of freshwater and saltwater organisms compared to
(SSEM-AVS)//OC
(SSEM-AVS)//oca
Omol/goc)
foe ,
Toxicity Test Metal(s) (goc/g) NOEC(s)
Life Cycle:
Leptocheirus Cadmium 0.030 -117, -66.7, 26, 63.3
plumulosus
Chironomus tentans Zinc 0.038 -68, -36.8, 168
OEC(s)c Reference
297,520 DeWittetal., 1996
576,847 Sibleyetal., 1996
Colonization:
Laboratory- saltwater
Cadmium
800, 2740
Hansenetal., 1996b
Field-saltwater
Field-freshwater
Field-freshwater
Cadmium,
copper,
lead, nickel,
zinc
Cadmium
Zinc
0.002
0.079
0.111
-155, -30, 10
-0.92,1.08,4.30
-32.7, -31.8, -26.4,
-18.2,9.1
Boothman et al.,
2001
28 Hare etal., 1994
Liber etal., 1996
a(SSEM-AVS)//oc concentrations in bold type are those between 130 and 3,000 //mol/gQQ for which the expectation of effects is
uncertain. Italics indicates concentrations where effects were observed but not expected.
NOECs = no observed effect concentration(s); all concentrations where response was not significantly different from the control.
cOECs = observed effect concentration(s); all concentrations where response was significantly different from the control.
3-26
-------
Derivation of Metal Mixtures ESBAVS:WQCs
Section 4
Derivation of Metal Mixtures
rLSBAVS:WQCs
4.1 General Information
Section 4 of this document presents the technical
basis for establishing the ESB for cadmium, copper,
lead, nickel, silver, and zinc. The basis of the overall
approach is the use of EqP theory linked to the concept
of maintaining metal activity for the sediment
interstitial water system below concentrations that
cause adverse effects. Extensive toxicological
concentration-response data from short-term and
chronic laboratory and field experiments, with both
marine and freshwater sediments and a variety of
species, indicate that it is possible to reliably predict
absence of metal toxicity based on EqP theory and
derive ESBs for metals in sediments using either of two
approaches referred to as ESBAVS:WQCs . The
ESBAVS:WQCs for the six metals that collectively predicts
absence of their toxicity in sediments can be derived by
(a) comparing the sum of their molar concentrations,
measured as SEM, with the molar concentration of AVS
in sediments (solid-phase AVS benchmark); or (b)
summing the measured interstitial water concentrations
of the metals divided by their respective WQC FCVs
(interstitial water benchmark). Lack of exceedence of
the ESBAVS:WQC based on either of these two procedures
indicates that metal toxicity should not occur.
At present, the technical basis for implementing
these two approaches is supportable. The approaches
have been presented to and reviewed by the SAB (U.S.
EPA, 1994a, 1995a, 1999).
Additional research required to fully implement
other approaches for deriving an ESBAVS:WQC for these
metals and to derive an ESB for other metals such as
mercury, arsenic, and chromium includes the
development of uncertainty estimates; part of this
would include their application to a variety of field
settings and sediment types. Finally, the ESB
approaches are intended to protect benthic organisms
from direct toxicity associated with exposure to metal-
contaminated sediments. The ESBs do not consider the
antagonistic, additive or synergistic effects of other
sediment contaminants in combination with metal
mixtures or the potential for bioaccumulation and
trophic transfer of metal mixtures to aquatic life,
wildlife or humans. They are not designed to protect
aquatic systems from metal release associated with, for
example, sediment suspension, or the transport of
metals into aquatic food webs. In particular, studies
are needed to understand the toxicological significance
of the biomagnification of metals that occurs when
predators consume benthic organisms that have
accumulated metals from sediments with more AVS
than SEM (Ankley, 1996).
The following nomenclature is used in subsequent
discussions of the ESBAVS:WQCs derivation for metal
mixtures. The ESBAVS:WQC for the metals, based on
AVS, is expressed in molar units because of the molar
stoichiometry of metal binding to AVS. Thus, solid-
phase constituents (AVS, SEM) are in ^mol/g dry
weight. The interstitial water metal concentrations are
expressed in ^mol/L or Mg/L, either as dissolved
concentrations [MJ or activities {M2+} (Stumm and
Morgan, 1981). The subscripted notation, Md, is used
to distinguish dissolved aqueous-phase molar
concentrations from solid-phase molar concentrations
with no subscript. For the combined concentration,
[SEMT], the units are ^mol of total metal per gram of
dry weight sediment. Note also that when [SEMAg] is
summed and/or compared with AVS, one-half the molar
silver concentration is applied.
One final point should be made with respect to
nomenclature. The terms nontoxic and having no effect
are used only with respect to the six metals considered
in this document. Toxicity of field- collected sediments
can be caused by other chemicals. Therefore, avoiding
exceedences of the ESBAVS:WQC for metal mixtures does
not mean that the sediments are nontoxic. It only
ensures that the six metals being considered should not
cause direct toxicity to benthic organisms. Moreover,
as discussed in detail below, exceedence of the
benchmarks for the six metals does not necessarily
indicate that metals will cause toxicity. For these
reasons,it is strongly recommended that the combined
use of both AVS and interstitial water measurements;
toxicity tests; TIEs; chemical monitoring in vertical,
horizontal, and temporal scales; and other assessment
methodologies as integral parts of any evaluation of the
effects of sediment-associated contaminants (Ankley et
4-1
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
al., 1994; Lee et al., 2000).
4.2 Sediment Benchmarks for Multiple
Metals
It is neither sufficient nor appropriate to derive an
ESB that considers each metal separately, because
metals almost always occur as mixtures in field
sediments and metal-sulfide binding is interactive.
4.2.1 AVS Benchmarks
Results of calculations using chemical equilibrium
models indicate that metals act in a competitive manner
when binding to AVS. That is, the six metalssilver,
copper, lead, cadmium, zinc, and nickel will bind to
AVS and be converted to their respective sulfides in
this sequence (i.e., in the order of increasing
solubility). Therefore, they must be considered
together. There cannot be a benchmark for just nickel,
for example, because all the other metals may be
present as metal sulfides, and therefore, to some extent,
as AVS. If these other metals are not measured as a
mixture, then the ZSEM will be misleadingly small,
and it might appear that 2[SEM]<[AVS] when in fact
this would not be true if all the metals are considered
together. It should be noted that this document
currently restricts this discussion to the six metals listed
above; however, in situations where other sulfide-
forming metals (e.g., mercury) are present at high
concentrations, they also must be considered.
The equilibrium model used to derive the
ESBAVS:WQC for a mixture of the metals is presented
below (see Ankley et al., 1996, for details). If the
molar sum of SEM for the six metals is less than or
equal to the AVS, that is, if
S. [SEM.] < [AVS]
where
(4-1)
Z [SEM,] = [SEMcd] + [SEMCJ + [SEMpb] + [SEMM]
+ [SEMZn] + l/2[SEMAg]
then the concentrations of the mixtures of metals in the
sediment are acceptable for protection of benthic
organisms from acute or chronic metal toxicity.
4.2.2 Interstitial Water Benchmarks
The application of the interstitial water benchmark
to multiple metals is complicated, not by the chemical
interactions of the metals in the sediment-interstitial
water system (as in the case with the AVS benchmark),
but rather because of possible toxic interactions. Even
if the individual concentrations do not exceed the water
quality final chronic value (FCV) of each metal
Table 4-1. Water quality criteria (WQC) final chronic value (FCV) based on the dissolved concentration of metal3
Saltwater FCV
Freshwater FCV
Metal
Cadmium
Copper
Lead
Nickel
Silver
Zinc
9.3
3.1
8.1
8.2
NAe
81
p-pC r (0.7852[ln(hardness)]-3.490-,
<^r \c J
0 96Qre(°-8545[ln(hardness)]-1.465)-|
0 7Q1 rp(1-273Pn(hardness)]-4.705)-i
n QQ7r (0.8460[ln(hardness)]+1.1645)-i
NAe
0 OQf;re(a8473[ln(hardness)]+0.7614)-|
a These WQC FCV values are for use in the interstitial water benchmarks approach for deriving ESBsAVS:WQC based on the dissolved metal
concentrations in interstitial water (U.S. EPA, 1995b).
b
For example, the freshwater FCV at a hardness of 50, 100, and 200 mg CaCO3/L are 0.62, 1.0, and 1.7 ,ug cadmium/L; 6.3, 10, and 20 //g
copper/L; 1.0, 2.5, and 6.1 ,ug lead/L; 87, 160, and 280 ^g nickel/L; and 58, 100, and 190 ^g zinc/L.
°CF = conversion factor to calculate the dissolved FCV for cadmium from the total FCV for cadmium: CF=1.101672-[(ln
hardness)(0.041838)].
dThe saltwater FCV for copper is from U.S. EPA (1995c).
eThe silver criteria are currently under revision to reflect water quality factors that influence the criteria such as hardness, DOC, chloride,
and pH, among other factors.
4-2
-------
Derivation of Metal Mixtures ESBAVS:WQcs
presented in Table 4-1, the metals could exert additive
effects that might result in toxicity (Biesinger et al.,
1986; Spehar and Fiandt, 1986; Enserink et al., 1991;
Kraak et al., 1994). Therefore, in order to address this
potential additivity, the interstitial water metal
concentrations are converted to interstitial water
benchmark units (IWBUs). This conversion is done by
dividing the individual metal interstitial water
concentrations by their respective WQC FCV and
summing these values for all the metals. IWBUs are
conceptually similar to toxic units; however, the term
IWBU was adopted because it is derived using the
FCV, which is intended to be a "no effect"
concentration (i.e., toxicity would not usually be
expected at 1.0 IWBUs).
For freshwater sediments, the FCVs are hardness
dependent for all of the divalent metals under
consideration, and thus, need to be adjusted to the
hardness of the interstitial water of the sediment being
considered. Because there are no FCVs for silver in
freshwater or saltwater, this approach is not applicable
to sediments containing significant concentrations of
silver (i.e., ZSEM>AVS). Because silver has the
smallest solubility product (see Table 2-2) and the
greatest affinity for AVS, it would be the last metal to
be released from the AVS or the first metal to bind with
AVS. Therefore, it is unlikely that silver would occur
in the interstitial water of any sediment with
measurable AVS (Berry et al., 1996).
For the ith metal with a total dissolved
concentration, [M.J, the IWBU is
S.[SEM;] < [AVS]
- [Mj.dJ
"' FCV; H
(4-2)
where
r [M,d] = [MCd.dl , [MCu.d] , [Mpb,d]
<[FCViid] [FCVcd,d] [FCVCu,d] [FCVPb>d]
, [MNj.d] ,
[FCVNi,d] [FCVZlljd]
4.2.3 Summary
In summary, the sediment benchmarks for these six
metals are not exceeded, and benthic organisms are
sufficiently protected, if the sediment meets either one
of the following benchmarks.
or
.
' FCVW
(4-1)
(4-2)
If the AVS or interstitial water ESBAVS:WQCs are
exceeded, there is reason to believe that the sediment
might be unacceptably contaminated by these metals.
Further evaluation and testing would, therefore, be
necessary to assess actual toxicity and its causal
relationship to the metals of concern. If data on the
sediment-specific SEM, AVS, and organic carbon
concentrations are available, the uncertainty bounds for
(SSEM-AVS)//OC described in Section 3.4 could be
used to further classify sediments as those in which
metals are not likely to cause toxicity, metal toxicity
predictions are uncertain, or metal toxicity is likely.
For sediments in which toxicity is likely or uncertain,
acute and chronic tests with species that are sensitive to
the metals suspected to be of concern, acute and
chronic sediment TIEs, in situ community assessments,
and seasonal and spatial characterizations of the SEM,
AVS, and interstitial water concentrations would be
appropriate (Ankley et al., 1994).
4.3 Example Calculation of ESBAVS:WQCs
for Metals and EqP-Based
Interpretation
To assist users of these ESBAVS WQCs for mixtures of
metals, example calculations for deriving solid-phase
and interstitial water ESB
s are provided in Table
AVS:WQC'
4-2. For each of the three sediments, the calculations
began with measured concentrations (in bold) of AVS
(//g/g), SEM; (/wg/g), and interstitial water metal (//g/
L). All other values were calculated. The specific
concentrations in each of the these sediments were
selected to provide examples of how the chemical
measurements are used with the ESBAVS WQC to
determine the acceptability of a specific sediment and
how the risks of sediment-associated metals can be
evaluated within the technical framework of the EqP
approach. Sediments are arranged in the table in
decreasing order of their sulfide solubility product
constants (see Section 2.2.5).
Sediment A contains relatively high concentrations
of metals in the SEM, between 14.2 and 16.5 //g/g for
copper, lead, and zinc. However, because there is
sufficient AVS (0.96 /^mol/g) in the sediment, the solid-
phase ESBAVS:WQC is -0.343 (/^mol/g), and there is no
4-3
-------
*
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Table 4-2. Example calculations of ESBAVS.WQCs for metal mixtures: three sediments.
Sediment Concentration
Sediment Analyte
A AVS
SEMNi
SEMZn
SEMcd
SEMpb
SEMCu
SEMAg
SSEM = 0.617 Mmol/g;
B AVS
SEMNi
SEMZn
SEMcd
SEMpb
SEMCu
SEMAg
SSEM = 46.5 ,umol/g;
C AVS
SEMNi
SEMZn
SEMcd
SEMpb
SEMcu
SEMAg
SSEM = 10.28 Mmol/g;
Mg/g*
30.8
2.85
16.5
0.05
14.2
16.0
SEM-AVS =
1310
34.0
2630
82.9
282
227
NDC
SEM-AVS = 5
146
269
12.4
573
66.2
4.44
NDC
SEM-AVS =
,umol/g
0.96
0.048
0.25
0.001
0.068
0.25
Interstitial Water Concentration
Metal (Mid) ,ug/L
Nickel NDC (<0.8)
Zinc NDC (<5.0)
Cadmium NDC (<0.2)
Lead NDC (<0.7)
Copper NDC (<0.6)
Silver
-0.343 ,umol/g
40.8
0.58
40.2
0.74
1.36
3.58
NDc
.71 ,umol/g
4.57
4.58
0.19
5.12
0.32
0.07
NDc
5.71 jumol/j
Nickel 4.8
Zinc 43.2
Cadmium NDC (<0.01)
Lead NDC (<0.10)
Copper NDC (<0.05)
Silver NDC (<0.01)
FCVb
8.2
81
9.3
8.1
3.1
LSBAVS:WQC
SEM-AVS IW
IWBU Omol/g)
<0.10
<0.06
<0.02
<0.09
<0.19
SIWBU <0.46 -0.34 <0.46
160
100
1.0
2.5
11
0.03
0.43
<0.01
<0.04
<0.005
SIWBU -0.46 5.71 -0.46
Nickel 26.3
Zinc 4.3
Cadmium 24.9
Lead NDC (<0.10)
Copper NDC (<0.05)
Silver NDC (<0.01)
87
58
0.62
1.0
6.3
0.30
0.07
40.1
<0.10
<0.008
1 SIWBU -40.47 5.71 -40.5
a Molecular weights: sulfur, 32.06; nickel, 58.7; zinc, 65.4; cadmium, 112; lead, 207; copper, 63.5; silver, 108.
b Saltwater sediment: sediment A. Freshwater sediments: sediment B, interstitial hardness 100 mg/L; sediment C, 50 mg/L.
°ND = not detected.
metal detected in the interstitial water. This sediment is
acceptable for protection of benthic organisms from
direct toxicity of the metals in the sediment. Silver was
not measured in this sediment. However, because AVS
is present, any silver in the sediment is not of
lexicological concern and none should occur in
interstitial water. One final consideration is the need
for detection limits for metals in the sediment that are
significantly below their respective WQC FCVs. For
this sediment there were no detectable metals in the
interstitial water and SIWBU was <0.46.
Sediment B is from a Superfund site heavily
contaminated with all of the metals (SSEM = 46.5
4-4
-------
Derivation of Metal Mixtures ESB
AVS:WQCa
, but most severely with zinc (2,630 ,wg/g).
There is an excess of SEM in this sediment (SEM-AVS
= 5.71 jwmol/g). Importantly for sediment B, the
interstitial concentrations of the metals were all less
than the WQC FCVs and the ZIWBU was <1.0 (~
0.46). Therefore, this sediment is acceptable for
protection of benthic organisms from direct toxicity of
this mixture of metals in the sediment. It should be
noted that, if interstitial metal concentrations had not
been quantified, the sediment would have exceeded the
ESBAVS:WQC and additional testing would be advisable.
A possible explanation for the absence of significant
metals in the interstitial water of this sediment is its
higher organic carbon concentration (foc = 0.05). The
(ZSEM-AVS)//OC of 114 ,wmol excess SEM/goc for this
sediment is, therefore, predicted to be nontoxic because
it is <130 ^mol excess SEM/goc (see Section 3.4.4).
Sediment C is heavily contaminated with
approximately equimolar concentrations of cadmium
and nickel. It exceeds the ESBAVS:WQC for metals for
both solid and interstitial water phases. The ZSEM
(10.28 ,umol/g) exceeds the AVS (4.57 /-onol/g);
therefore, SEM-AVS = 5.71 ,wmol excess SEM/g, a
concentration identical to that of sediment B. Although
lead and copper are found in the sediment, they are not
found in detectable concentrations in the interstitial
water. This is because they have the lowest sulfide
solubility product constants and the sum of their SEM
concentrations (0.39 ^mol/g) is less than AVS. If the
dry weight concentrations of metals had been analyzed,
silver and additional copper and nickel might have
been detected. Silver will not be detected in the SEM
or interstitial water when AVS is present (see Section
3.2.1). Nickel, cadmium, and zinc occur in interstitial
water because in the sequential summation of the SEM.
concentrations in order of increasing sulfide
solubilities, the concentrations of these metals exceed
the AVS. Therefore, these three metals are found in the
SEM that is not a metal sulfide and in the interstitial
water, and contribute to the ZIWBU (-40.47) as well
as to the overall exposure of benthic organisms.
Because only cadmium concentrations exceed the
WQC FCV, any effects observed in toxicity tests or in
faunal analyses with this sediment should principally be
a result of cadmium. This sediment is low in organic
carbon concentration (TOC = 0.2%;/oc = 0.002). The
organic carbon-normalized concentration (S SEM-
AVS//OC) of 2,855 ,wmol excess SEM/goc was within the
uncertainty bounds of 130 to 3,000 ^mol excess SEM/
goc, suggesting that additional evaluations should be
conducted (see Section 3.4.4).
4.4 ESBAVS:WQC for Metals vs.
Environmental Monitoring Databases
This section compares the ESB
based on
AVS or IWBUs with chemical monitoring data from
freshwater and saltwater sediments in the United States.
This comparison of AVS-SEM and interstitial water
concentrations is used to indicate the frequency of
sediments in the United States where metals toxicity is
unlikely. When data were available in the monitoring
programs, (2SEM-AVS)//OC is used to indicate
sediments where toxicity is unlikely, likely, or
uncertain. When toxicity or benthic organism
community health data are available in conjunction
with these concentrations it is possible to speculate as
to potential causes of the observed effects. These data,
however, cannot be used to validate the usefulness of
the AVS approach because sediments that exceed the
benchmarks are not always toxic, and because observed
sediment toxicity may be the result of unknown
substances.
4.4.1 Data Analysis
Three monitoring databases were identified that
contain AVS, SEM, and/oc information; one also had
data on concentrations of metals in interstitial water.
Toxicity tests were conducted on all sediments from
these sources. The sources are the Environmental
Monitoring and Assessment Program (EMAP)
(Leonard et al., 1996a), the National Oceanographic
and Atmospheric Administration National Status and
Trends monitoring program (NOAA NST) (Wolfe et
al., 1994; Long et al., 1995b, 1996), and the Regional
Environmental Monitoring and Assessment Program
(REMAP) (Adams et al., 1996).
4.4.1.1 Freshwater Sediments
The AVS and SEM concentrations in the 1994
EMAP database from the Great Lakes were analyzed
by Leonard et al. (1996a). A total of 46 sediment grab
samples and 9 core samples were collected in the
summer from 42 locations in Lake Michigan. SEM,
AVS, TOC, interstitial water metals (when sufficient
volumes were present), and 10-day sediment toxicity to
the midge C. tentans and the amphipod H. azteca were
measured in the grab samples (the concentrations are
listed in Appendix A).
The AVS concentrations versus SEM-AVS
differences from Appendix A are plotted in Figure 4-1.
Grab sediment samples containing AVS concentrations
4-5
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
10
f
3
I
-10
-20
-30
-40
0.01
V
o.i
10
100
10
1
3
C/5
-5
B
-10
0.01
0.1 1 10
Acid Volatile Sulfide (/. mol/g)
100
Figure 4-1. SEM-AVS values versus AVS concentrations in EMAP-Great Lakes sediments from Lake Michigan. Data
are from surficial grab samples only. Plot (A) shows all values; plot (B) has the ordinate limited to SEM-
AVS values between -10 and +10 /^mol/g.
4-6
-------
Derivation of Metal Mixtures ESB
AVS:WQCa
below the detection limit of 0.05 ^tmol/g AVS are
plotted at that concentration. Forty-two of the 46
samples (91%) had SEM-AVS differences greater than
0.0. Thirty-six of these had less than 1.0 ^mol of
excess SEM/g sediment; and none had over 5.8 ^mol
excess SEM/g sediment. Sediments with SEM
concentrations in excess of that for AVS have the
potential to be toxic because of metals. However, the
majority of sediments with an excess of SEM had low
concentrations of both AVS and SEM. For 20 of these
Lake Michigan sediments, interstitial water metals
concentrations were measured. The sum of the IWBUs
for cadmium, copper, lead, nickel, and zinc was always
less than 0.4 (Leonard et al., 1996a). In 10-day toxicity
tests using C. tentans and H. azteca, no toxicity was
observed in 81% of the 21 sediments not exceeding the
ESBAVS:WQC, Leonard et al. (1996a) concluded that
when toxicity was observed it was not likely from
metals, because of the low interstitial water metals
concentrations. These data demonstrate the value of
using both SEM-AVS and IWBUs to evaluate the risks
of metals in sediments.
4.4.1.2 Saltwater Sediments
Saltwater data from a total of 398 sediment
samples from 5 monitoring programs representing the
eastern coast of the United States are included in
Appendix B. The EMAP Virginian Province database
(U.S. EPA, 1996) consists, in part, of 127 sediment
samples collected from August to mid-September 1993
from randomly selected locations in tidal rivers and
small and large estuaries from the Chesapeake Bay to
Massachusetts (Strobel etal., 1995). The NOAA data
are from Long Island Sound, Boston Harbor, and the
Hudson River Estuary. Sediments were collected from
63 locations in the coastal bays and harbors of Long
Island Sound in August 1991 (Wolfe et al., 1994).
Sediment samples from 30 locations in Boston Harbor
were collected in June and July 1993 (Long et al.,
1996). Sediment samples from 38 locations in the
Hudson River Estuary were collected from March to
May 1991 (Long etal., 1995b). Sediment samples
were collected in the REMAP program from 140
locations from the New York/New Jersey Harbor
Estuary System (Adams et al., 1996). All of the above
sediment grab samples were from approximately the
top 2 cm of undisturbed sediment.
For saltwater sediments, the molar concentration of
AVS typically exceeds that for SEM (SEM-AVS <0.0
,wmol/g) for most of the samples across the entire range
of AVS concentrations (Figure 4-2). A total of 68 of
the 398 saltwater sediments (17%) had an excess of
metal, and only 4 of the 68 (6%) had over 2 ^trnol
excess SEM/g. As AVS levels increase, fewer and
fewer sediments have SEM-AVS differences that are
positive; none occurred when AVS was >8.1 ^mol/g.
Interstitial water metal was not measured in these
saltwater sediments. Only 5 of the 68 sediments (7%)
having excess of up to 0.9 ,wmol SEM/g were toxic in
10-day sediment toxicity tests with the amphipodA
abdita, whereas 79 of 330 sediments (24%) having an
excess of AVS were toxic. Toxicity was not believed to
be metals related in the 79 toxic sediments where AVS
was in excess over SEM. Metals were unlikely the
cause of toxicity in those sediments having an excess of
SEM because there was only <0.9 ^tmol excess SEM/g.
Finally, the absence of toxicity in sediments having an
excess of SEM of up to 4.4 ^mol/g indicates significant
metal-binding potential over that of AVS in some
sediments. Organic carbon concentrations from 0.05%
to 15.2% (average 1.9%) provide for some of this
additional metal binding.
Organic carbon, along with SEM and AVS, was
measured in these 398 saltwater sediments. Therefore,
the (2SEM-AVS)//OC concentrations of concern can be
compared with the organic carbon-normalized
concentrations of SEM-AVS differences (Figure 4-3).
No sediments containing an AVS concentration in
excess of 10 ^tmol/g had an excess of SEM; that is, all
(ZSEM-AVS)//OC values were negative. Excess of
SEM relative to AVS became more common as
sediment AVS decreased. None of the sediments
contained greater than 130 ^mol excess SEM/goc, the
lower uncertainty bound from Section 3.4. This
indicates that metals concentrations in all of the
sediments monitored in the summer by EPA EMAP and
REMAP and by NOAA are below concentrations of
concern for benthic organisms.
4.5 Bioaccumulation
The data appear to suggest that, for these
sediments collected from freshwater and marine
locations in the United States, direct toxicity caused by
metals in sediments is expected to be extremely rare.
Although this might be true, these data by themselves
are inconclusive. Importantly, it would be
inappropriate to use the data from the above studies to
conclude that metals in sediments are not a problem.
In all of the above studies, the sediments were
conducted in the summer when the seasonal
biogeochemical cycling of sulfur should produce the
highest concentrations of iron monosulfide, which
4-7
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
50
o
3
VI
in
-50
-100
-150
-200'
0.01
10.0
O EMAP
D REMAP
O NOAA
0.1
10
-10.0
0.01
0.1 1 10
Acid Volatile Sulfide (^mol/g)
D
o
100
1000
100
1000
Figure 4-2. SEM-AVS values versus AVS concentrations in EMAP-Estuaries Virginian Province (U.S. EPA, 1996);
REMAP-NY/NJ Harbor Estuary (Adams et al., 1996); NOAA NST-Long Island Sound (Wolfe et al.,
1994); Boston Harbor (Long et al., 1996); and Hudson-Raritan Estuaries (Long et al., 1995b). Plot A
shows all values; plot B has the ordinate limited to SEM-AVS values between -10 and +10 //imol/g (see
data in Appendix B).
4-8
-------
Derivation of Metal Mixtures ESB
AVS:WQCa
W)
"o
3
o
w
o
s
^
w
5000
-5000
-10000
-15000
-20000
O EMAP
D REMAP
ONOAA
0.01
0.1
10
100
1000
1000
500
-500
-1000
0.01
100
1000
Acid Volatile Sulfide (//mol/g)
Figure 4-3. (ZSEM-AVS)//OC versus AVS concentrations in EMAP-Estuaries Virginian Province (U.S. EPA, 1996);
REMAP-NY/NJ Harbor Estuary (Adams et al., 1996); NOAANST-Long Island Sound (Wolfe et al.,
1994); Boston Harbor (Long et al., 1996); and Hudson-Raritan Estuaries (Long et al., 1995b). Plot A
shows all values; plot B has the ordinate limited to (ZSEM-AVS)//^ values between -10 and +10 /^mol/g
(see data in Appendix B).
4-9
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
might make direct metal-associated toxicity less likely
than in the winter/spring months. Accurate assessment
of the extent of the direct ecological risks of metals in
sediments requires that sediment monitoring occur in
the months of minimum AVS concentration; typically,
but not always, in November to early May. These yet-
to-be-conducted studies must monitor, at a minimum,
SEM, AVS,/OC, interstitial water metal, and toxicity.
Bioaccumulation of metals from sediments when
SEM is less than AVS was not expected based on EqP
theory. However, there is a significant database that
demonstrates that metals concentrations inbenthic
organisms increase when metals concentrations in
sediments on a dry weight basis increase (Ankley,
1996). This has caused considerable debate (Lee et al.,
2000a,b) because it suggests that metal bioavailability
may be related to dry weight metals concentrations, and
if the increase in bioaccumulated metal is related to
effects, then effects may be related to dry weight metals
concentrations. Most importantly, these studies, and all
other AVS-related testing, has overwhelming
demonstrated that toxic effects of metals are absent in
sediments when SEM is less than AVS, even when
bioaccumulation is observed, and that toxicity is not
related to dry weight metals concentrations. For
example, careful evaluation of Lee et al. (2000b)
results, demonstrates that in order to understand and
predict metal toxicity AVS normalization is critical.
Although Lee et al. (2000b) note the accumulation of
metal by the test organisms, no adverse effects were
reported. This suggests that the bioaccumulated metals
may not be lexicologically available or of sufficient
concentration in the organism to cause effects. In
addition, these metals do not biomagnify to higher
trophic levels in aquatic ecosystems (Suedel et al.,
1994). Therefore, an ESBAVS:WQC based on the
difference between the concentrations of SEM and AVS
is appropriate for protecting benthic organisms from
the direct effects of sediment-associated metals, and
not for protecting against metal bioaccumulation.
4-10
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Section 5
Sampling and Analytical Chemistry
5.1 General Information
This section provides guidance on procedures for
sampling, handling, and analysis of metals in
sediments, and on the interpretation of data from the
sediment samples that are needed if the assessments of
the risks of sediment-associated metals are to be
appropriately based on the EqP methodology. The
design of any assessment should match the goal of the
specific assessment and how evaluation tools such as
ESB
s are to be applied.
AVS:WQO
Results of the short- and long-term laboratory and
field experiments conducted to date using sediments
spiked with individual metals and mixtures of metals
represent convincing support for the conclusion that
absence (but not necessarily presence) of metal toxicity
can be reliably predicted based on metal-sulfide
relationships or interstitial water metal concentrations.
In contrast, much confusion exists on how to use this
convincing evidence to interpret the significance of
metals concentrations in sediments from the field.
Using these observations as a basis for predicting
metal bioavailability, or deriving anESBAVS:WQC, raises a
number of conceptual and practical issues related to
sampling, analytical measurements, and effects of
additional binding phases. Many of these were
addressed by Ankley et al. (1994). Those most salient
to the proposed derivation of the ESBAVS:WQCs are
described below.
5.2 Sampling and Storage
Accurate prediction of exposure of benthic
organisms to metals is critically dependent on sampling
appropriate sediment horizons at appropriate times.
This is because of the relatively high rates of AVS
oxidation caused by natural processes in sediments and
the requirement that oxidation must be avoided during
sampling of sediments and interstitial water. In fact, the
labile nature of iron monosulfides has led some to
question the practical utility of using AVS as a basis for
an EqP-derived ESB for metals (Luoma and Carter, 1993;
Meyer etal., 1994). For example, there have been many
observations of spatial (depth) variations in AVS
concentrations, most of which indicate that surficial
AVS concentrations are less than those in deeper
sediments (Boothman and Helmstetter, 1992; Howard
andEvans, 1993; Brumbaugh etal., 1994; Hare etal.,
1994; Besser etal., 1996; Hansen etal., 1996b; Leonard
etal., 1996a; Liber etal., 1996; Boothman etal., 2001).
This is likely because of oxidation of AVS (principally
FeS) at the sediment surface, a process enhanced by
bioturbation (Peterson et al., 1996).
In addition to varying with depth, AVS can vary
seasonally. For example, in systems where overlying
water contains appreciable oxygen during cold-weather
months, AVS tends to decrease, presumably because of
a constant rate of oxidation of the AVS linked to a
decrease in its generation by sulfate-reducing bacteria
(Herlihy and Mills, 1985; Howard and Evans, 1993;
Leonard etal., 1993). Because of potential temporal and
spatial variability of AVS, it appears that the way to
avoid possible underestimation of metal bioavailability
is to sample the biologically "active" zone of sediments
at times when AVS might be expected to be present at
low concentrations. It is recommended that, at a
minimum, AVS and SEM measurements be made using
samples of the surficial (0 to 2.0 cm) sediments during
the period from November to early May. Minimum AVS
concentrations may not always occur during cool-
weather seasons; for example, systems that become
anaerobic during the winter can maintain relatively
large sediment AVS concentrations (Liber et al., 1996).
Therefore, AVS, SEM, and interstitial metal
concentrations may need to be determined seasonally.
Importantly, the biologically active zones of some
benthic communities may be within only the surficial
first few millimeters of the sediment, whereas other
communities may be biologically active at depths up to
a meter. In order to determine the potential for exposure
to metals, sediment and interstitial water samples from
multiple sediment horizons may be required.
The somewhat subjective aspects of these
sampling recommendations have been of concern.
Multiple sediment samples are necessary because of
the dynamic nature of the metal-binding phases in
sediments. Depending on the depth of bioturbation,
the possible oxidation rates of specific metal sulfides,
and the extent of possible metal concentrations, the
horizontal and vertical resolution of the needed
monitoring is likely to be site specific. Even if neither
5-1
-------
Sampling and Analytical Chemistry
of the sediment benchmarks is violated in extensive
monitoring programs, metals concentrations on a dry
weight basis may be high and widely distributed. This
may be a good reason to conduct monitoring studies to
determine the extent of metal bioaccumulation in
benthic food chains. Furthermore, if the ultimate fate of
the sediments is unknown, risk assessments to evaluate
future risks caused by dynamic processes may be
desirable.
Research suggests that the transient nature of AVS
may be overstated relative to predicting the fate of all
metal-sulfide complexes in aquatic sediments.
Observations from the Duluth EPA laboratory made in
the early 1990s indicate that AVS concentrations in
sediments contaminated by metals such as cadmium
and zinc tended to be elevated over concentrations
typically expected in freshwater systems (G.T. Ankley,
U.S. EPA, Duluth, MN, personal communication). The
probable underlying basis for these observations did
not become apparent, however, until a recent series of
spiking and metal-sulfide stability experiments. The
field colonization study of Liber et al. (1996)
demonstrated a strong positive correlation between the
amount of zinc added to test sediments and the
resultant concentration of AVS in the samples. In fact,
the initial design of their study attempted to produce
test sediments with as much as five times more SEMZn
(nominal) than AVS; however, the highest measured
SEMZn/AVS ratio achieved was only slightly larger than
1. Moreover, the expected surficial depletion and
seasonal variations in AVS were unexpectedly low in
the zinc-spiked sediments. These observations
suggested that zinc sulfide, which composed the bulk
of AVS in the spiked sediments, was more stable than
the iron sulfide present in the control sediments. The
apparent stability of other metal sulfides versus iron
sulfide also has been noted in laboratory spiking
experiments with freshwater and saltwater sediments
(Leonard etal., 1995; De Witt etal., 1996;Hansenetal.,
1996b;Petersonetal., 1996; Sibley etal., 1996;
Boothman etal., 2001).
In support of these observations, metal-sulfide
oxidation experiments conducted by Di Toro et al.
(1996b) have confirmed that cadmium and zinc form
more stable sulfide solid phases than iron. If this is
also true for sulfide complexes of copper, nickel, silver,
and lead, the issue of seasonal/spatial variations in AVS
becomes of less concern because most of the studies
evaluating variations in AVS have focused on iron
sulfide (i.e., uncontaminated sediments). Thus, further
research concerning the differential stability of metal
sulfides, from both temporal and spatial perspectives, is
definitely warranted.
5.2.1 Sediments
At a minimum, sampling of the surficial 2.0 cm of
sediment between November and early May is
recommended. A sample depth of 2.0 cm is appropriate
for monitoring. However, for instances such as
dredging or in risk assessments where depths greater
than 2 cm are important, sample depths should be
planned based on particular study needs. Sediments
can be sampled using dredges, grabs, or coring, but
mixing of aerobic and anaerobic sediments must be
avoided because the trace metal speciation in the
sediments will be altered (see Bufflap and Allen, 1995,
for detailed recommendations to limit sampling
artifacts). Coring is generally less disruptive, facilitates
sampling of sediment horizons, and limits potential
metal contamination and oxidation if sealed PVC core
liners are used.
Sediments not immediately analyzed for AVS and
SEM must be placed in sealed airtight glass jars and
refrigerated or frozen. Generally, enough sediment
should be added to almost fill the jar. If sediments are
stored this way, there will be little oxidation of AVS
even after several weeks. Sampling of the stored
sediment from the middle of the jar will further limit
potential effects of oxidation on AVS. Sediments
experiencing oxidation of AVS during storage will
become less black or grey if oxidized. Because the rate
of metal-sulfide oxidation is markedly less than that of
iron sulfide, release of metal during storage is unlikely.
5.2.2 Interstitial Water
Several procedures are available to sample
interstitial water in situ or ex situ. Carignan et al. (1985)
compared metals concentrations in interstitial water
obtained by ex situ centrifugation at 11,000 rpm
followed by filtration (0.45 ,wmand 0.2 orO.03 ,wm) and
in situ diffusion samplers with 0.2 ^m polysulfone
membranes. For the metals of concern in this
benchmark document, concentrations of nickel and
cadmium were equivalent using both methods, and
concentrations of copper and zinc were higher and
more variable using centrifugation. They recommended
using in situ dialysis for studying trace constituents in
sediments because of its inherent simplicity and the
avoidance of artifacts that can occur with the handling
of sediments in the laboratory.
5-2
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
More recently, Bufflap and Allen (1995) reviewed
four procedures for collection of interstitial water for
trace metals analysis. These included ex situ
squeezing, centrifugation, in situ dialysis, and suction
filtration. These authors observed that each method
has its own advantages and disadvantages.
Importantly, interstitial water must be extracted by
centrifugation or squeezing in an inert atmosphere until
acidified, because oxidation will alter metal speciation.
Artifacts may be caused by temperature changes in ex
situ methods that may be overcome by maintaining
temperatures similar to those in in situ methods.
Contamination of interstitial water by fine particles is
important in all methods, because differentiation of
paniculate and dissolved metal is a function of the pore
size of the filter or diffusion sampler membrane. The
use of 0.45 jum filtration, although an often accepted
definition of "dissolved" metals, may result in
differences from laboratory to laboratory. Use of
suction filtration devices is limited to coarser
sediments, and they do not offer depth resolution.
Use of diffusion samplers is hampered by the time
required for equilibrium (7 to 14 days) and the need for
diver placement and retrieval in deep waters.
Acidification of interstitial water obtained by diffusion
or from suction filtration must occur immediately to limit
oxidation. Bufflap and Allen (1995) conclude that in
situ techniques have less potential for producing
sampling artifacts than ex situ procedures. They
concluded that, of the in situ procedures, suction
filtration has the best potential for producing artifact-
free interstitial water samples directly from the
environment. Of the ex situ procedures, they
concluded that centrifugation under a nitrogen
atmosphere followed immediately by filtration and
acidification was the simplest technique likely to result
in an unbiased estimate of metal concentrations in
interstitial water. At present, it is recommended
filtration of the surface water through 0.40 to 0.45 ,um
polycarbonate filters to better define that fraction of
aqueous metal associated with toxicity (Prothro, 1993).
This guidance applies to interstitial water. Thurman
(1985) equates the organic carbon retained on a 0.45 /j.m
glass-fiber filter to suspended organic carbon, so that
this filtration procedure under nitrogen atmosphere
followed immediately by acidification is acceptable for
interstitial waters. However, in studies comparing
collection and processing methods for trace metals,
sorption to filter membranes or the filtering apparatus
does occur (Schults et al., 1992). These authors later
presented a method combining longer centrifugation
times with a unique single-step interstitial water
withdrawal procedure that has potential for minimizing
metal losses by eliminating the need for filtration
(Ozretichand Schults, 1998).
Use of dialysis samplers to obtain samples of
interstitial water is recommended for comparison of
measured concentrations of dissolved metals with
WQC. This is primarily because diffusion samplers
obtain interstitial water with the proper in situ
geochemistry, thus limiting artifacts of ex situ sampling.
Furthermore, in shallow waters, where contamination of
sediments is most likely, placement of diffusion
samplers is easily accomplished and extended
equilibration times are not a problem. Second, use of
centrifugation under nitrogen and 0.45 ,wm filtration
using polycarbonate filters for obtaining interstitial
water from sediments in deeper aquatic systems. Care
must be taken to ensure that filters or the filter
apparatus do not remove metal from or add metal to the
interstitial water sample to be analyzed. Perhaps most
importantly, the extremely large database comparing
interstitial metals concentrations with organism
responses from spiked- and field-sediment experiments
in the laboratory has demonstrated that, where the
IWTU concept predicted that metals concentrations in
interstitial water should not be toxic, toxicity was not
observed when either dialysis samplers or
centrifugation were used (Berry et al., 1996; Hansen et
al., 1996a). Therefore, it is likely that when either
methodology is used to obtain interstitial water for
comparison with WQC, if metals concentrations are
below 1.0 IWBU, sediments should be acceptable for
protection of benthic organisms. The exception is for
some silver-spiked freshwater and saltwater sediments
that were toxic in spite of the absence of interstitial
silver. It is for this reason that IWBUs are not used as
ESBAVS:WQCs for silver (see Sections 4.2.1 and 4.2.2).
5.3 Analytical Measurements
An important aspect to deriving ESB values is that
the methods necessary to implement the approach must
be reasonably standardized or have been demonstrated
to produce results comparable to those of standard
methodologies. From the standpoint of the
ESB
AVS:WQC'
s, a significant amount of research has gone
into defining methodologies to obtain interstitial water
and sediments (see Section 5.2 above), to extract SEM
and AVS from sediments, and to quantify AVS, SEM,
and the metals in interstitial water.
5-3
-------
Sampling and Analytical Chemistry
5.3.1 Acid Volatile Sulflde
The SEM/AVS extraction method suggested is that
of Allen etal. (1993). In terms of AVS quantification, a
number of techniques have been successfully utilized,
including gravimetric (Di Toro et al., 1990; Leonard et
al., 1993), colorimetric (Cornwell and Morse, 1987), gas
chromatography- photoionization detection (Casas and
Crecelius, 1994; Slotton and Reuter, 1995), and specific
ion electrodes (Boothman and Helmstetter, 1992;
Brouwer and Murphy, 1994; Brumbaugh etal., 1994;
Leonard etal., 1996b). Allen etal. (1993) report a
detection limit for 50% accuracy of 0.01 ^mol/gfora 10
g sediment sample using the colorimetric method.
Based on several studies, Boothman and Helmstetter
(1992) report a detection limit of 1 ,wmol AVS, which
translates to 0.1 ^mol/g dry weight for a 10 g sediment
sample using the ion specific electrode method.
5.3.4 Interstitial Water Metal
Interstitial water can be analyzed for the metals
cadmium, copper, lead, nickel, silver, and zinc by routine
atomic spectrochemical techniques appropriate for
environmental waters (e.g., inductively coupled plasma
atomic emissionorGFAA) (U.S. EPA, 1994b). Because
of the need to determine metals at concentrations at or
below the threshold of biological effects (i.e., WQC
concentrations), additional consideration must be
given to preclude contamination during collection,
transport and analysis (U.S. EPA, 1995d,e,f; also see
guidance on clean chemistry techniques in U. S. EPA,
1994c). Generally, detection limits should be at <0.1
IWBU because the contributions of each of the metals
must be summed.
5.3.2 Simultaneously Extracted Metals
SEMs are operationally defined as metals
extracted from sediment into solution by the AVS
extraction procedure. The dissolved metals in this
solution are also operationally defined as the metal
species that pass through filter material used to remove
the residual sediment. Common convention defines
"dissolved" as metal species <0.45 ^m in size. SEM
concentrations measured in sediments are not
significantly different, however, using Whatman #1
filter paper alone (<11 ^m nominal interstitial size) or in
combination with a 0.45 ^m filter (W. Boothman, U.S.
EPA, Narragansett, RI, personal communication). SEM
solutions generated by the AVS procedure can be
analyzed for metals, commonly including cadmium,
copper, lead, nickel, silver, and zinc, by routine atomic
spectrochemical techniques appropriate for
environmental waters (e.g., inductively coupled plasma
atomic emission or graphite furnace atomic absorption
spectrophotometry [GFAA]) (U.S. EPA, 1994b).
Because of the need to determine metals at relatively
low concentrations, additional consideration must be
given to preclude contamination during collection,
transport, and analysis (U.S. EPA, 1995d,e,f).
5.3.3 Total Organic Carbon
Several methods for measuring organic carbon
exist and are reviewed by Nelson and Sommers (1996).
U.S. EPA (2001) summarizes the minimum requirements
of acceptable methods for quantifying total organic
carbon in sediments.
5-4
-------
Mixtures
Section 6
Sediment Benchmark Values:
Application and Interpretation
The procedures described in this document
indicate that, except possibly where a locally,
commercially, or recreationally important species is very
sensitive, benthic organisms should be acceptably
protected in freshwater and saltwater sediments if at
least one of the following two conditions are satisfied:
the sum of the molar concentrations of SEM cadmium,
copper, lead, nickel, silver, and zinc is less than or equal
to the molar concentration of AVS (Section 6. 1 ), or the
sum of the dissolved interstitial water concentration of
cadmium, copper, lead, nickel, and zinc divided by their
respective WQC FCV is less than or equal to 1.0
(Section 6.2). The AVS benchmark is intended to apply
to sediments having aO. 1 fanol AVS/g. The two
conditions for deriving ESBAVS.WQC are detailed in
Section 4.2 and are repeated below.
Consistent with the recommendations of EPA' s
Science Advisory Board, publication of these documents
does not imply the use of ESBs as stand-alone, pass-fail
criteria for all applications; rather, exceedances of
ESBs could trigger collection of additional assessment
data.
As discussed in Section 3.4, a more accurate
prediction of toxicity can be derived if the presence of
organic carbon is considered along with AVS. For the
multiple metals cadmium, copper, lead, nickel, silver and
zinc, the following assumptions are useful in deriving a
benchmark:
1 ) Any sediment in which (SEM - AVS)//^ < 1 30 Mmols/
g^ should pose low risk of adverse biological effects
due to cadmium, copper, lead, nickel and zinc.
2) Any sediment in which 1 SO^Hicls/g^ < (SEM - AVS)/
/oc < 3000 /^mols/goc may have adverse biological
effects due to cadmium, copper, lead, nickel or zinc.
3) In any sediment in which (SEM - AVS)//^. > 3000
jumols/g^, adverse biological effects due to cadmium,
copper, lead, nickel or zinc maybe expected.
4) Any sediment with AVS > 0.0 will not cause adverse
biological effects due to silver.
6.1 AVSESB
S,[SEMJ <; [AVS]
where
S, [SEM;] = [SEMCd] + [SEMCJ + [SEMpb] + [SEMNi]
+ [SEMZJ + l^tSEM^]
6.2 Interstitial Water ESB
;i.o
where
_ [Mcd.d] ,
''[FCVi>d] [FCVcd,d] [FCVcu.d'tFCVpb.d]
[MNJ.d] . [MZn,d]
[FCVNiid]
It is repeated here that the interstitial water
benchmark applies only to the five metals: cadmium,
copper, lead, nickel, and zinc. Silver is not included in
this benchmark because the FCV for silver is not
available.
Arguably, the most important additional data
needed for assessing contaminated sediments along
with ESBs are the results of toxicity tests. Sediment
toxicity tests provide an important complement to ESBs
in interpreting overall risk from contaminated
sediments. Toxicity tests have different strengths and
weaknesses compared to chemical-specific guidelines,
6-1
-------
Benchmark Statement
and the most powerful inferences can be drawn when
both are used together (see U.S. EPA 2003c,d for further
discussion of using toxicity testing with ESBs to
assess contaminated sediments).
The ESB approaches are intended to protect
benthic organisms from direct toxicity associated with
exposure to metal-contaminated sediments. They are
not designed to protect aquatic systems from metals
release associated, for example, with sediment
suspension, or the transport of metals into the food
web from either sediment ingestion or ingestion of
contaminated benthos. Furthermore, the ESBs do not
consider the antagonistic, additive or synergistic
effects of other sediment contaminants in combination
with metal mixtures or the potential for bioaccumulation
and trophic transfer of metal mixtures to aquatic life,
wildlife or humans.
6-2
-------
Section 7
References
References
Adams WJ, Kimerle RA, MosherRG. 1985. Aquatic
safety assessment of chemicals sorbed to sediments. In
Cardwell RD, Purdy R, Banner RC, eds, Aquatic
Toxicology and Hazard Assessment: Seventh
Symposium. STP 854. American Society for Testing and
Materials, Philadelphia, PA, pp 429-453.
Adams DA, O'Connor JS, Weisberg SB. 1996. Sediment
quality of the NY/NJ Harbor SystemAn investigation
under the Regional Environmental Monitoring and
Assessment Program (R-EMAP). Draft Final Report.
Division of Environmental Science and Assessment,
U.S. Environmental Protection Agency, Edison, NJ.
Allen HE, HalRH, BrisbinTD. 1980. Metal speciation
effects on aquatic toxicity. Environ Sci Technol
14:441443.
AllenHE, Fu G, DengB. 1993. Analysis of acid-volatile
sulfide (AVS) and simultaneously extracted metals
(SEM) for the estimation of potential toxicity in aquatic
sediments. Environ Toxicol Chem 12:1-13.
AndersonDM, Morel FM. 1978. Copper sensitivity of
Gonyaulax tamarensis. Limnol Oceanogr 23:283-295.
Ankley GT, Phipps GL, Leonard EN, Benoit DA,
Mattson VR, KosianPA, Cotter AM, Dierkes JR,
HansenDJ, Mahony JD. 1991. Acid volatile sulfide as a
factor mediating cadmium and nickel bioavailability in
contaminated sediments. Environ Toxicol Chem
10:1299-1307
Ankley GT, Mattson VR, Leonard EN, West CW,
Bennett JL. 1993. Predicting the acute toxicity of copper
in freshwater sediments: Evaluation of the role of acid-
volatile sulfide. Environ Toxicol Chem 12:315-320.
Ankley GT, DiToro DM, HansenDJ, Mahony JC,
Swartz RC, Hoke RA, Thomas NA, Garrison AW, Allen
HE, Zarba CS. 1994. Assessing the potential
bioavailability of metals in sediments: Aproposed
approach. Environ Manage 18:331-337.
Ankley GT. 1996. Evaluation of metal/acid volatile
sulfide relationships in the prediction of metal
bioaccumulationby benthic macroinvertebrates. Environ
ToxicolChem 15:2138-2146.
Ankley GT, Di Toro DM, HansenD, Berry WJ. 1996.
Technical basis and proposal for deriving sediment
quality criteria for metals. Environ Toxicol Chem
15:2056-2066.
Barrow NJ, Ellis AS. 1986a. Testing a mechanistic model.
III. The effects of pH on fluoride retention by a soil. J
Soil Sci 37:287-295.
Barrow NJ, Ellis AS. 1986b. Testing a mechanistic model.
IV. Describing the effects of pH on zinc retention by a
soil. J Soil Sci 37:295-302.
Barrow NJ, Ellis AS. 1986c. Testing a mechanistic model.
V The points of zero salt effect for phosphate retention
and for acid/alkali titration of a soil. J Soil Sci 37:303-
310.
Bell RA, Kramer JA. 1999. Structural chemistry and
geochemistry of silver-sulfur compounds: Critical review.
Environ ToxicolChem 18:9-22.
Bergman HL, Dorward-KingEJ. 1997. Reassessment of
Metals Criteria for Aquatic Life Protection. Society of
Environmental Toxicology and Chemistry, Pensacola, FL.
BernerRA. 1967. Thermodynamic stability of
sedimentary ironsulfides.^4/w JSci 265:773-785.
Berry WJ, Hansen DJ, Mahony JD, Robson DL, Di Toro
DM, Shipley BP, Rogers B, Corbin JM, Boothman WS.
1996. Predicting the toxicity of metals-spiked laboratory
sediments using acid-volatile sulfide and interstitial
water normalization. Environ ToxicolChem 15:2067-2079.
Berry WJ, Cantwell M, Edwards P, Serbst JR, HansenDJ.
1999. Predicting the toxicity of sediments spiked with
silver.Environ ToxicolChem 18:40-48.
Besser JM, Ingersoll CG Giesy JP. 1996. Effects of spatial
and temporal variability of acid volatile sulfide on the
bioavailability of copper and zinc in freshwater
sediments. Environ Toxicol Chem 15:286-293.
7-1
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
BiesingerKE, Christensen GM, Fiandt JT. 1986. Effects
of metal salt mixtures on Daphnia magna reproduction.
Ecotoxicol Environ Saf 11:9-14.
Boothman WS, Helmstetter A. 1992. Vertical and
seasonal variability of acid volatile sulfides in marine
sediments. Research Report. Environmental Monitoring
and Assessment Program, U.S. Environmental
Protection Agency, Narragansett, RI.
Boothman WS, Hansen DJ, Berry WJ, Robson D,
Helmstetter A, Corbin J, Pratt SD. 2001. Variation of
acid volatile sulfides and metals and biological
response in field-exposed sediments. Environ Toxicol
Chem 20:264-272.
Boulegue J, Lord III CJ, Church TM. 1982. Sulfur
speciation and associated trace metals (Fe, Cu) in the
interstitial waters of Great Marsh, Delaware. Geochim
CosmochimActa 46:453-464.
BrouwerH, Murphy TP. 1994. Diffusion method for the
determination of acid-volatile sulfides (AVS) in
sediment. Environ Toxicol Chem 13:1273-1275.
Brumbaugh WG, Ingersoll CG Kemble ME, May TW,
ZajcekJL. 1994. Chemical characterization of sediments
and interstitial water from the upper Clark Fork River
and Milltown Reservoir, Montana. Environ Toxicol
Chem 13:1971-1983.
Bufflap WE, Allen HE. 1995. Sediment interstitial water
collection methods: Areview. Water Res 29:65-177.
Burgess RM, Scott KJ. 1992. The significance of in-
place contaminated marine sediments on the water
column: Processes and effects. In Burton GA, Jr., ed,
Sediment Toxicity Assessment. Lewis Publishing, Boca
Raton, FL,pp 129-154.
CarignanR. 1984. Interstitial water sampling by dialysis:
Methodological notes. Limnol Oceanogr 20:667-670.
CarignanR, RapinF, TessierA. 1985. Sediment
interstitial water sampling for metal analysis: A
comparison of techniques. Geochim CosmochimActa
49:2493-2497.
Carlson AR, Phipps GL, Mattson VR, KosianPA, Cotter
AM. 1991. The role of acid volatile sulfide in
determining cadmium bioavailability in sediments.
Environ Toxicol Chem 10:1309-1319.
Casas AM, CreceliusEA. 1994. Relationship between
acid volatile sulfide and the toxicity of zinc, lead and
copper in marine sediments. Environ Toxicol Chem
13:529-536.
Chapman GA. 1987. Establishing sediment criteria for
chemicalsRegulatory perspective. InDicksonKL,
Maki AW, Brungs WA, eds, Fate and Effects of
Sediment-Bound Chemicals in Aquatic Systems.
Pergamon Press, New York, NY, pp 355-376.
Cornwell JC, Morse JW 1987. The characterization of
iron sulfide minerals in anoxic marine sediments. Mar
Chem 22:193-206.
Davies-Colley RJ, NelsonPO, WilliamsonKJ. 1985.
Sulfide control of cadmium and copper concentrations
in anaerobic estuarine sediments. Mar Chem 16:
173-186.
DeWitt TH, Ozretich RJ, Swartz RC, Lamberson JO,
Shults DW, Ditsworth GR, Jones JKP, Hoselton L,
Smith LM. 1992. The influence of organic matter
quality on the toxicity and partitioning of sediment-
associated fluoranthene. Environ Toxicol Chem 11:
197-208.
DeWitt TH, Swartz RC, Hansen D J, Berry WJ,
McGovernD. 1996. Interstitial metal and acid volatile
sulfide predict the bioavailability of cadmium during a
full life-cycle sediment toxicity test using the estuarine
amphipod, Leptocheirus plumulosus. Environ Toxicol
Chem 15:2095-2101.
Di Toro DM, Harrison F, Jenne E, Karickoff S, Lick W
1987. Synopsis of discussion session 2: Environmental
fate andcompartmentalization. InDicksonKL, Maki
AW, Brungs WA, eds, Fate and Effects of Sediment-
Bound Chemicals in Aquatic Systems. Pergamon
Press, New York, NY, pp 136-147.
Di Toro DM, Mahony JJ, Hansen D J, Scott KJ, Hicks
MB, Mayr SM, Redmond MS. 1990. Toxicity of
cadmium in sediments: The role of acid volatile sulfide.
Environ Toxicol Chem 9:1487-1502.
Di Toro DM, Zarba CS, Hansen DJ, Berry WJ, Swartz
RC, Cowan CE, Pavlou SP, Allen HE, Thomas NA,
Paquin PR. 1991. Technical basis for establishing
sediment quality criteria for nonionic organic chemicals
using equilibrium partitioning. Environ Toxicol Chem
10:1541-1583.
7-2
-------
References
Di Toro DM, Mahony JD, HansenDJ, Scott KJ, Carlson
AR, Ankley GT. 1992. Acid volatile sulfide predicts the
acute toxicity of cadmium and nickel in sediments.
Environ Sci Technol26:96-101.
Di Toro DM, Mahony JD, Gonzalez AM. 1996a. Particle
oxidation model of synthetic FeS and sediment acid
volatile sulfide. Environ Toxicol Chem 15:2156-2167.
Di Toro DM, Mahony JD, HansenDJ, Berry WJ. 1996b.
A model of the oxidation of iron and cadmium sulfide in
sediments. Environ Toxicol Chem 15:2168-2186.
Di Toro DM, McGrath JA, Hansen DJ, Barry WJ,
PaquinPR, Mathew R, Wu, KB, Santore RC. Draft.
Predicting sediment metal toxicity using a sediment
biotic ligand model: (I) single metals.
Dzombak DA, Morel FMM. 1990. Surface
Complexation Modeling, Hydrous Ferric Oxide. John
Wiley, New York, NY.
Emerson S, Jacobs L,TeboB. 1983. The behavior of
trace metals in marine anoxic waters: Solubilities at the
oxygen-hydrogen sulfide interface. In Wong CS, ed,
Trace Metals in Sea Water. Plenum Press, New York,
NY
EnserinkEL, Maas-DiepeveenJL,vanLeeuwenCS.
1991. Combined toxicity of metals: Anecotoxicological
evaluation. Water Res 25:679-687.
GoldhauberMB, Kaplan IR. 1974. The sulfur cycle. In
Goldberg ED, ed, The SeaMarine Chemistry, Vol 5.
John Wiley, New York, NY, pp 569-655.
GreenAS, Chandler GT, Blood ER. 1993. Aqueous-,
interstitial water-, and sediment-phase cadmium:
Toxicity relationships for a meiobenthic copepod.
Environ Toxicol Chem 12:1497-1506.
Hansen, DJ. 1995. Assessment tools that can be used
for the National Sediment Inventory. Memorandum to
C. Fox, U.S. Environmental Protection Agency, Office of
Water, February 28,1994.
Hansen DJ, Berry WJ, Mahony JD, Boothman WS,
RobsonDL, Ankley GT,MaD, Yan Q, Pesch CE. 1996a.
Predicting toxicity of metal-contaminated field
sediments using interstitial water concentrations of
metals and acid-volatile sulfide normalization. Environ
Toxicol Chem 15:2080-2094.
HansenDJ, Mahony JD, Berry WJ, Benyi SJ, Corbin
JM, Pratt SD, Able MB. 1996b. Chronic effect of
cadmium in sediments on colonization by benthic
marine organisms: An evaluation of the role of
interstitial cadmium and acid-volatile sulfide in
biological availability. Environ Toxicol Chem 15:2126-
2137.
HareL, CarignanR, Huerta-Diaz MA. 1994. An
experimental study of metal toxicity and accumulation
by benthic invertebrates: Implications for the acid
volatile sulfide (AVS) model. Limnol Oceanogr
39:1653-1668.
Herlihy AT, Mills AL. 1985. Surfate reduction in
freshwater sediments receiving acid mine drainage.
Appl Environ Microbiol 49:179-186.
HessleinRH. 1976. An in situ sampler for close interval
interstitial water studies. Limnol Oceanogr 30:221-227.
Hoke RA, Ankley GT, Cotter AM, Goldenstein T,
KosianPA, Phipps GL, VandermeidenFM. 1994.
Evaluation of equilibrium partitioning theory for
protecting acute toxicity of field-collected sediments
contaminated with DDT, DDE, and ODD to the
amphipod Hyalella azteca. Environ Toxicol Chem
13:157-166.
HowardDE, EvansRD. 1993. Acid-volatile sulfide
(AVS) in a seasonally anoxic mesotrophic lake:
Seasonal and spatial changes in sediment AVS. Environ
Toxicol Chem 12:1051-1058.
Ingersoll CG, Haverland PS, BrunsonEL, Canfield TJ,
Dwyer FJ, Henke CE, Kemble ME, Mount DR, Fox RG.
1996. Calculation and evaluation of sediment effect
concentrations for the amphipod Hyalella azteca and
the midge Chironomus riparius. J Great Lakes Res
22:602-623.
Jenne EA. 1968. Controls onMn, Fe, Co, Ni, Cu, andZn
concentrations in soils and waterThe significant role
of hydrous Mn and Fe oxides. In Advances in
Chemistry. American Chemical Society, Washington,
DC, pp 337-387.
Jenne EA. 1977. Trace element sorptionby sediments
and soilSites and processes. In Chappell W, Petersen
K, eds, Symposium on Molybdenum in the
Environment, Vol 2. Marcel Dekker, New York, NY, pp
425-553.
7-3
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Jenne EA, Di Toro DM, AllenHE, Zarba CS. 1986. An
activity-based model for developing sediment criteria
for metals: Anew approach. Chemicals in the
Environment. International Conference,
Lisbon, Portugal.
Karickhoff S W, McDaniel VK, Melton C, Vellino AN,
NuteDE, CarreiraLA. 1991. Predicting chemical
reactivity by computer. Environ Toxicol Chem 10:
1405-1416.
Kemble NE, Brumbaugh WG, BrunsonEL, DwyerFJ,
Ingersoll CG, Mounds DP, Woodward DF. 1994. Toxicity
of metal-contaminated sediments from the upper Clark
Fork River, Montana, to aquatic invertebrates and fish
in laboratory exposures. Environ Toxicol Chem
13:1985-1997.
Kemp PF, Swartz RC. 1986. Acute toxicity of interstitial
and particle-bound cadmium to a marine infaunal
amphipod. Mar Environ Res 26:135-153.
Kraak MHS, Lavy D, Schoon H, Toussaint M, Peeters
WHM, vanStraalenNM. 1994. Ecotoxicity of mixtures
of metals to the zebra mussel Dreissena polymorpha.
Environ ToxicolChem 13:109-114.
Landers AH, David MB, MitchelMJ. 1983. Analysis of
organic and inorganic sulfur constituents in sediments,
soils, and water. IntJ Environ Anal Chem 14:245-256.
LarssonP. 1985. Contaminated sediments of lakes and
oceans act as sources of chlorinated hydrocarbons for
release to water and atmosphere. Nature 317:347-349.
Lee B-G, Griscom SB, Lee JS, Choi HJ, Koh C-H, Louma
SN, Fisher NS. 2000a. Influences of dietary uptake and
reactive sulfides on metal bioavailability from aquatic
sediments. Science 287:282-284.
Lee B-G, Lee JS, Luoma SN, Choi HJ, Koh CH. 2000b.
Influence of acid volatile sulfide and metal
concentrations on metal bioavailability to marine
invertebrates in contaminated sediments. Environ Sci
Technol 34:4517-4523.
Leonard EN, Mattson VR, Benoit DA, Hoke RA, Ankley
GT. 1993. Seasonal variation of acid volatile sulfide
concentration in sediment cores from three
northeastern Minnesota lakes. Hydrobiologia 271:
87-95.
Leonard EN, Mattson VR, Ankley GT. 1995. Horizon-
specific oxidation of acid volatile sulfide in relation to
the toxicity of cadmium spiked into a freshwater
sediment. Arch Environ Contain Toxicol 28:78-84.
Leonard EN, Ankley GT, Hoke RA. 1996a. Evaluation of
metals in marine and freshwater surficial sediments from
the Environmental Monitoring and Assessment
Program relative to proposed sediment quality criteria
for metals. Environ ToxicolChem 15:2221-2232.
Leonard EN, Cotter AM, Ankley GT. 1996b. Modified
diffusion method for analysis of acid volatile sulfides
and simultaneously extracted metals in freshwater
sediment.Environ ToxicolChem 15:1479-1481.
Liber K, Call D, Markee T, Schmude K, BalcerM,
WhitemanF, Ankley GT. 1996. Effects of acid volatile
sulfide on zinc bioavailability and toxicity to benthic
macroinvertebrates in spiked-sediment field
experiments.Environ ToxicolChem 15:2113-2125.
Long ER, Morgan LG. 1990. The potential for biological
effects of sediment sorbed contaminants tested in the
National Status and Trends Program. NOS OMA 52.
Technical Memorandum. National Oceanic and
Atmospheric Administration, Office of Oceanography
and Marine Assessment, Seattle, WA.
LongER,MacDonaldDD, Smith SL, CalderFD. 1995a.
Incidence of adverse biological effects within ranges of
chemical concentrations in marine and estuarine
sediments. Environ Manage 19:81-97.
LongER, Wolfe DA, Scott KJ, Thursby GB, Stern EA,
Peven C, Schwartz T. 1995b. Magnitude and extent of
sediment toxicity in the Hudson-Raritan Estuary. NOS
ORCA88. Technical Memorandum. Office of Ocean
Resources, Conservation, and Assessment, National
Oceanic and Atmospheric Administration, Silver Spring,
MD.
Long ER, Sloane GM, CarrRS, Scott KJ, Thursby GB,
WadeTL. 1996. Sediment toxicity in Boston Harbor:
Magnitude and extent, and relationships with chemical
toxicants. NOS ORCA88 (REMAP). Technical
Memorandum. Office of Ocean Resources,
Conservation, and Assessment. National Oceanic and
Atmospheric Administration, Silver Spring, MD.
Luoma SN, Bryan W. 1981. A statistical assessment of
the form of trace metals in oxidized sediments
employing chemical extractants. Sci Total Environ
17:165-196.
7-4
-------
References
Luoma SN. 1989. Can we determine the biological
availability of sediment-bound trace elements?
Hydrobiologia 176/177:379-396.
Luoma SN, Carter JL. 1993. Understanding the toxicity
of contaminants in sediments: Beyond the bioassay-
basedparadigm. Environ Toxicol Chem 12:793-796.
MacDonaldDD, CarrRS, CalderFD, LongER, Ingersoll
CG. 1996. Development and evaluation of sediment
quality guidelines for Florida coastal waters.
Ecotoxicol 5:253 -278.
Mahony JD, Di Toro DM, Koch R, Berry W, Hansen D.
1993. Vertical distribution of AVS and SEM in bedded
sediments, biological implications and the role of metal
sulfide oxidation kinetics. Abstract. 14th Annual
Meeting. Society of Environmental Toxicology and
Chemistry, Houston, TX, November 14-18,1993, p 46.
Mahony JD, Di Toro DM, Gonzalez AM, Curto M, Dilg
M. 1996. Partitioning of metals to sediment organic
carbon. Environ Toxicol Chem 15:2187-2197.
Meyer JS, Davison W, Sundby B, Oris JT, LaurenDJ,
ForstnerU, Hong J, Crosby DG. 1994. The effects of
variable redox potentials, pH and light onbioavail-
ability in dynamic water-sediment environments.
In Hamelink JL, Landrum PF, Bergman HL, Benson WH,
eds, Bioavailability: Physical, Chemical and
Biological Interactions. CRC Press, Boca Raton, FL,
pp 155-170.
Morse JW, Millero FJ, Cornwell JC, Rickard D. 1987.
The chemistry of the hydrogen sulfide and iron sulfide
systems in natural waters. Earth Sci Rev 24:1-42.
NelsonDW, Sommers LE. 1996. Total carbon, total
organic carbon and organic matter. In Sparks D, ed,
Methods of Soil Analysis: Part 3, Chemical Methods.
Soil Science Society of America and American Society
of Agronomy, Madison, WI.
Oakley SM, WilliamsonKJ, NelsonPO. 1980. The
geochemical partitioning and bioavailability of trace
metals in marine sediments. PhD Thesis. Water
Research Institute, Oregon State University,
Corvallis, OR.
OzretichRJ, SchultsDW. 1998. Acomparisonof
interstitial water isolation methods demonstrates
centrifugation with aspiration yields reduced losses of
organic constituents. Chemosphere 36:603-615.
Pankow JF. 1979. The dissolution rates and mechanisms
of tetragonal ferrous sulfide (mackinawite) in anoxic
aqueous systems. PhD thesis. California Institute of
Technology, Pasadena, CA.
PersaudD, JacqumagiR, HaytonA. 1989. Development
of provincial sediment quality guidelines. Draft Report.
Water Resources Branch, Ontario Ministry of the
Environment. Toronto, ON, Canada.
Pesch CE, Hansen D J, Boothman WS, Berry WJ,
Mahony JD. 1995. The role of acid volatile sulfide and
interstitial water metal concentrations in determining
bioavailability of cadmium and nickel from
contaminated sediments to the marine polychaete
Neanthes arenaceodentata. Environ Toxicol Chem
14:129-141.
Peterson GS, Ankley GT, Leonard EN. 1996. Effect of
bioturbation on metal-sulfide oxidation in surficial
freshwater sediments. Environ Toxicol Chem 15:2147-
2155.
Phillips HO, Kraus KA. 1965. Adsorption on inorganic
materials VI. Reaction of insoluble sulfides with metal
ions in aqueous media. JChrom 17:549-557.
ProthroMG. 1993. Office of Water policy and technical
guidance on interpretation and implementation of
aquatic life metals criteria. Letter to Water Management
Division directors. Office of Water, U.S. Environmental
Protection Agency, October 1,1993.
Salomons W, De Rooij NM, Kerdijk H, Bril J. 1987.
Sediments as sources of contaminants? Hydrobiol
149:13-30.
SchoonenMAA, Barnes HL. 1988. An approximation of
the second dissociation constant for H2S. Geochim
CosmochimActa 52:649-654.
Schults DW, Ferraro SP, Smith LM, Roberts FA,
Poindexter CK. 1992. A comparison of methods for
collecting interstitial water for trace organic compounds
and metals, Wat Res 26:989-995.
Sibley PK, Ankley GT, Cotter AM, Leonard EN. 1996.
Predicting chronic toxicity of sediments spiked with
zinc: An evaluation of the acid volatile sulfide (AVS)
model using a life-cycle test with the midge
Chironomus tentans. Environ Toxicol Chem
15:2102-2112.
7-5
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Slotton DG, Reuter JE. 1995. Heavy metals in intact and
resuspended sediments of a California reservoir, with
emphasis on potential bioavailability of copper and
zinc. Mar Freshwater Res 46:257-265.
SpeharRL, Fiandt JT. 1986. Acute and chronic effects
of water quality criteria-based metal mixtures on three
aquatic species. Environ Toxicol Chem 5:917-931.
Sposito G, De Wit JCM, Neal RH. 1988. Selenite
adsorption on alluvial soils: III. Chemical modeling.
JSoilSci SocAm 52:947-950.
Stephan CE, Mount DI, Hansen D J, Gentile JH,
Chapman GA, Brungs WA. 1985. Guidelines for
deriving numerical national water quality criteria for the
protection of aquatic organisms and their uses. PB85-
227049. National Technical Information Service (NTIS),
Springfield, VA.
Strobel CJ, BuffumHW, Benyi SJ, Petrocelli EA,
ReifstickDR, Keith DJ. 1995. Statistical summary
EMAP-Estuaries: Virginian Province 1992-1993. EPA/
620/R-94/-26. U.S. Environmental Protection Agency,
Narragansett, RI.
Stumm W, Morgan JS. 1981. Aquatic Surface
ChemistryAn Introduction Emphasizing Chemical
Equilibria in Natural Waters. John Wiley,
New York, NY.
Stumm W. 1987. Aquatic Surface ChemistryChemical
Processes at the Particle-Water Interface. John Wiley,
New York, NY.
Stumm W, Morgan JS. 1996. Aquatic Surface
ChemistryChemical Equilibria and Rates in Natural
Waters. 3rd ed. John Wiley, New York, NY.
Suedel BC, Boracxek JA, Peddicord RK, Clifford PA,
Dillon TM. 1994. Trophic transfer and biomagnification
potential of contaminants in aquatic ecosystems.
Environ Toxicol Chem 136:21-89.
Sullivan J, Ball J, Brick E, Hausmann S, Philarski G,
SopcichD. 1985. Report of the technical subcommittee
on determination of dredge material suitability for in-
water disposal. Wisconsin Department of Natural
Resources, Madison, WI.
Sunda WG, GuillardRRL. 1976. The relationship
between cupric ion activity and the toxicity of copper
to phytoplankton. J Mar Res 34:511-529.
Sunda WG, Lewis JM. 1978. Effect of complexation by
natural organic ligands on the toxicity of copper to a
unicellular alga, Monochrysis lutheri. Limnol
Oceanogr 23:870-876.
Sunda WG, Engel DW, Thuotte RM. 1978. Effect of
chemical speciation of toxicity of cadmium to grass
shrimp, Palaemonetespugio: Importance to free
cadmium ion. Environ Sci Technol 12:409-413
Swartz RC, Dittsworth GR, Schults DW, Lamberson JO.
1985. Sediment toxicity to a marine infaunal amphipod:
Cadmium and its interaction with sewage sludge. Mar
Environ Res 18:133-153.
Swartz RC, Schults D W, DeWitt TH, Dittsworth GR,
Lamberson JO. 1990. Toxicity of fluoranthene in
sediment to marine amphipods: Atest of the equilibrium
partitioning approach to sediment quality criteria.
Environ Toxicol Chem 9:1071-1080.
Swartz RC. 1991. Acenaphthene and phenanthrene files.
Memorandum to D. Hansen, HydroQual, Inc., Mahwah,
NJ, June 26,1991. 160pp.
Swartz RC, Cole FA, Lamberson JO, Ferraro SP, Schults
DW, Deben WA, Lee IIH, OzretichRJ. 1994. Sediment
toxicity, contamination, and amphipod abundance at a
DDT- and dieldrin-contaminated site in San Francisco
Bay. Environ Toxicol Chem 13:949-962.
TessierA, Campbell PGC,BissonM. 1979. Sequential
extraction procedure for the speciation of paniculate
trace metals. Anal Chem 51:844-851.
TessierA, Campbell PGC. 1987. Partitioning of trace
metals in sediments: Relationships with bioavailability.
Hydrobiologia 149:43-52.
TessierA, CouillardY, Campbell PGC, AuclairJC. 1993.
Modeling Cd partitioning in oxic lake sediments and Cd
concentrations in the freshwater bivalve Andonata
grandis. Limnol Oceanogr 38:1-17.
ThurmanEM. 1985. Organic Geochemistry of Natural
Waters. Kluwer Academic Publishers Group, Martinus
Nijhoff/Dr. W. Junk Publishers, Boston, MA, pp 497.
TroupBN. 1974. The interaction of iron with
phosphate, carbonate, and sulfide in Chesapeake Bay
interstitial waters: Athermodynamic interpretation. PhD
thesis. Johns Hopkins University, Baltimore, MD.
7-6
-------
References
U.S. Environmental Protection Agency. 1980. Ambient
water quality criteriafor silver1980. EPA-440/5-80-
071. Office of Water Regulations and Standards,
Washington, DC.
U.S. Environmental Protection Agency. 1985a. Appendix
BResponse to public comments on "Guidelines for
deriving numerical national water quality criteria for the
protection of aquatic organisms and their uses." July
19,1985. Federal Register 50:30793-30796.
U.S. Environmental Protection Agency. 1985b. Ambient
water quality criteria for cadmium1984. EPA440/5-85-
032. Office of Water Regulations and Standards,
Washington, DC.
U.S. Environmental Protection Agency. 1985c. Ambient
water quality criteriafor copper1984. EPA440/5-85-
031. Office of Water Regulations and Standards,
Washington, DC.
U.S. Environmental Protection Agency. 1985d. Ambient
water quality criteriafor lead1984. EPA 440/5-85-027.
Office of Water Regulations and Standards,
Washington, DC.
U.S. Environmental Protection Agency. 1986. Ambient
water quality criteria for nickel1986. EPA 440/5-85-
004. Office of Water Regulations and Standards,
Washington, DC.
U.S. Environmental Protection Agency. 1987. Ambient
water quality criteriafor zinc1987. EPA 440/5-85-032.
Office of Water Regulations and Standards,
Washington, DC.
U.S. Environmental Protection Agency. 1989. Sediment
classification methods compendium. PB92-231679.
NTIS, Springfield, VA.
United States Environmental Protection Agency. 1992.
Sediment classification methods compendium. EPA
823-R-92-006, Office of Water, Washington, DC.
U.S. Environmental Protection Agency. 1994a.
Equilibrium partitioning approach to predicting metal
bioavailability in sediments and the derivation of
sediment quality criteria for metals, Vol 1. EPA822-D-94-
002. Briefing Report to the Science Advisory Board.
Office of Water, Office of Research and Development,
Washington, DC.
U.S. Environmental Protection Agency. 1994b. Methods
for the determination of metals in environmental
samples. Supplement I. EPA-600-R-94-111. Office of
Research and Development, Washington, DC.
U.S. Environmental Protection Agency. 1994c. Interim
guidance on determination and use of water-effect
ratios for metals. EPA-823-B-94-001. Office of Water,
Washington, DC.
U.S. Environmental Protection Agency. 1995a. Review
of the agency's approach for developing sediment
criteria for five metals. EPA-SAB-EPEC-95-020. SAB
Report. Office of Water, Washington, DC.
U.S. Environmental Protection Agency. 1995b. Water
quality criteria for dissolved metals. Federal Register
60(86):22231-22237.
U.S. Environmental Protection Agency. 1995c. Ambient
water quality criteriaSaltwater copper addendum.
Office of Water, Office of Science and Technology,
Washington, DC. p 12.
U.S. Environmental Protection Agency. 1995d. Method
1669: Sampling ambient water for trace metals at EPA
water quality criteria levels. EPA-821-R-95-034. Office of
Water, Engineering and Analysis Division (4303),
Washington, DC.
U.S. Environmental Protection Agency. 1995e. Method
1638: Determination of trace metals in ambient waters
by inductively coupled plasma-mass spectrometry.
EPA-821-R-95-031. Office of Water, Engineering and
Analysis Division (4303), Washington, DC.
U.S. Environmental Protection Agency. 1995f. Method
1639: Determination of trace metals in ambient waters
by stabilized temperature graphite furnace atomic
absorption. EPA-821-R-95-032. Office of Water,
Engineering and Analysis Division (4303),
Washington, DC.
U.S. Environmental Protection Agency. 1996. EMAP-
Estuaries Virginian Province Data 1990-1993. Available
from: EMAP home page, http://www.epa.gov/emap.
Accessed 1997 Sept 17.
U.S. Environmental Protection Agency. 1997a. The
incidence and severity of sediment contamination in
surface waters of the United States, Vol 1: National
sediment quality survey. EPA823-R-97-006. Office of
Water, Washington, DC.
7-7
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
U.S. Environmental Protection Agency. 1997b. The
incidence and severity of sediment contamination in
surface waters of the United States, Vol 2: Data
summaries for areas of probable concern. EPA 823-R-97-
007. Office of Water, Washington, DC.
U.S. Environmental Protection Agency. 1997c. The
incidence and severity of sediment contamination in
surface waters of the United States, Vol 3: National
sediment contaminant point source inventory. EPA 823-
R-97-008. Office of Water, Washington, DC.
U.S. Environmental Protection Agency. 1999.
Integrated approach to assessing the bioavailability
and toxicity of metals in surface waters and sediments.
EPA-822-E-99-001. Report to the U. S. EPA Science
Advisory Board. April 6-7,1999. Office of Water and
Office of Research and Development, Washington, DC.
U.S. Environmental Protection Agency. 2001. Methods
for collection, storage and manipulation of sediments
for chemical and lexicological analyses: technical
manual. EPA-823-B-01-002. Office of Water,
Washington, D.C.
U.S. Environmental Protection Agency. 2003a. Technical
basis for the derivation of equilibrium partitioning
sediment benchmarks (ESBs) for the protection of
benthic organisms: Nonionic organics. EPA-600-R-02-
014. Office of Research and Development,
Washington, DC (draft).
U.S. Environmental Protection Agency. 2003b.
Procedures for the derivation of site-specific
equilibrium partitioning sediment benchmarks (ESBs)
for the protection of benthic organisms: Nonionic
organics. EPA-600-R-02-012. Office of Research and
Development, Washington, DC (draft).
U.S. Environmental Protection Agency. 2003c.
Procedures for the derivation of equilibrium partitioning
sediment benchmarks (ESBs) for the protection of
benthic organisms: Endrin. EPA-600-R-02-009. Office
of Research and Development, Washington, DC.
U.S. Environmental Protection Agency. 2003d.
Procedures for the derivation of equilibrium partitioning
sediment benchmarks (ESBs) for the protection of
benthic organisms:Dieldrin. EPA-600-R-02-010. Office
of Research and Development, Washington, DC.
U.S. Environmental Protection Agency. 2003e.
Procedures for the derivation of equilibrium partitioning
sediment benchmarks (ESBs) for the protection of
benthic organisms: PAH mixtures. EPA-600-R-02-013.
Office of Research and Development, Washington, DC.
U.S. Environmental Protection Agency. 2003f.
Procedures for the derivation of equilibrium partitioning
sediment guidelines (ESGs) for the protection of
benthic organisms: Nonionics compendium. EPA-600-
R-02-016. Office of Research and Development,
Washington, DC (draft).
Wolfe DA, Bricker SB, LongER, Scott KJ, Thursby GB.
1994. Biological effects of toxic contaminants in
sediments from Long Island Sound and environs.
Technical Memorandum. NOS ORCA 80. National
Oceanic and Atmospheric Administration. Office of
Ocean Resources Conservation and Assessment, Silver
Spring, MD.
Zamuda CD, Sunda WG. 1982. Bioavailability of
dissolved copper to the American oyster Crassostrea
virginica'. Importance of chemical speciation. MarBiol
66:77-82.
7-8
-------
Appendix A
Lake Michigan EMAP Sediment Monitoring Database
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Concentrations of SEM, AVS, TOC, and IWBU for cadmium, copper, lead, nickel, and zinc in 46 surficial
samples from Lake Michigan
TOC
Sample (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
0.18
4.63
3.36
4.89
0.92
4.37
5.27
0.08
4.27
2.11
1.89
0.41
2.87
3.68
0.28
0.07
3.51
0.40
1.73
0.69
2.51
1.17
0.13
1.03
0.63
0.30
0.29
0.21
0.11
0.05
0.27
4.95
0.54
6.75
0.18
0.15
0.56
0.10
0.06
2.68
0.16
1.80
1.29
0.05
0.14
0.57
SEM
0/mol/g)
0.53
3.46
2.78
3.55
0.14
2.82
1.20
0.17
1.47
0.25
1.12
0.74
1.17
1.56
1.32
0.17
0.75
0.97
1.74
0.70
0.19
0.59
0.21
0.62
0.13
0.15
0.25
0.12
0.20
0.04
0.85
1.17
0.44
1.37
0.26
0.06
0.17
0.22
0.06
5.83
0.16
0.56
1.02
0.06
0.16
0.66
AVS
Omol/g)
0.03a
0.35
0.06
0.05
a
0.03
1.13
0.13a
0.03
4.49
a
0.03
a
0.03
0.07
0.18
a
0.03
0.44
0.05
0.08
a
0.03
0.15
a
0.03
0.05
a
0.03
a
0.03
a
0.03
0.20
a
0.03
0.03*
a
0.03
0.06
a
0.03
a
0.03
1.66
0.12
0.09
a
0.03
0.05
0.05
0.12
a
0.03
a
0.03
0.07
a
0.03
2.25
a
0.03
0.05
a
0.03
SEM-AVS
Oimol/g)
0.51
3.11
2.72
3.50
0.12
1.69
1.07
0.15
-3.02
0.23
1.10
0.67
0.99
1.54
0.88
0.12
0.67
0.95
1.59
0.68
0.14
0.57
0.19
0.60
-0.07
0.13
0.23
0.10
0.14
0.02
0.83
-0.49
0.32
1.28
0.24
0.01
0.12
0.10
0.04
5.81
0.09
0.54
-1.23
0.04
0.11
0.64
Cadmium
b
0.029
0.018
0.018
c
0.0002
0.024
0.029
0.115
0.050
C
0.0002
c
0.0002
c
0.0002
0.018
0.079
b
c
0.0002
c
0.0002
0.012
0.018
0.003
0.006 c
0.0002°
Copper
d
0.003
0.308
0.266
0.034
0.049d
0.003d
0.003
0.034
0.070
0.003
0.119
0.060
0.013
0.155
0.003
0.036
0.041
0.119
d
0.003
0.028
IWBU
Lead
0.00004
0.002
0.0004
0.0008
0.0002e
o.ooo ie
0.001
0.0008
0.002
0.0004
0.0002
0.0008
0.0008
e
0.0001
0.0004
0.0004
0.0002
0.001
0.0006
0.002
% Survival
Nickel
0.005
0.003
0.003
0.006
0.004
0.006
0.006
0.004
f
0.0005
0.006
0.004
0.008
0.010
0.011
0.007
0.002
0.017
f
0.0005
0.008 f
0.0005
Zinc
0.003
0.029
0.006
0.032
0.020
0.020
0.055
0.026
0.001
0.015
0.050
0.058
0.020
0.0003
0.0003
0.020
0.012
0.020
0.015
0.044
Hyalella Chironomus
Sum azteca tentans
0.040
0.360
0.293
0.073
0.097
0.058
0.180
0.115
0.074
0.025
0.173
0.145
0.123
0.167
0.011
0.070
0.088
0.144
0.033
0.075
92.5
90
92.5
100
0
97.5
92.5
95
95
77.5
97.5
97.5
96.5
90
100
100
95
97.5
97.5
75
97.5
57.5
72.5
95
35
75
80
97.5
97.5
97.5
100
95
95
95
60
97.5
90
62.5
75
100
82.5
70
40
90
90
97.5
90
100
100
87.5
100
87.5
100
97.5
92.5
87.5
100
100
100
97.5
97.5
92.5
100
65
57.5
90
35
72.5
82.5
100
97.5
95
100
90
100
92.5
55
100
95
65
95
55
72.5
67.5
a AVS Limit of Detection =0.03 ,um S/g.
b Insufficient interstitial water volume for metals analysis.
c Cadmium LOD=0.01 Mg/L (0.0002 IWBU).
d Copper LOD=0.2 Mg/L (0.0003 IWBU).
e Lead LOD=0.1
f
(0.0001 IWBU).
Nickel LOD=0.5 Mg/L (0.0005 IWBU).
Source: Columns for Sample, TOC, SEM, AVS, SEM-AVS, and IWBU taken directly from Leonard et al. (1996a).
Column for survival from personal communication with E.N. Leonard, U.S. EPA, Duluth, Minnesota.
A-l
-------
Appendix B
Saltwater Sediment Monitoring Database
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Concentrations of SEM, AVS, toxicity, and TOC for EMAP, NOAA NST, and REMAP databases
Study3
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
SEM
Cumol/g)
0.289
1.500
0.066
0.134
0.266
0.266
1.292
0.347
0.750
0.212
0.497
0.624
0.032
0.988
0.604
0.031
1.597
1.065
0.189
0.018
0.079
0.421
0.798
0.903
1.202
0.159
0.246
0.687
0.699
1.663
0.083
0.740
0.878
0.044
0.910
0.567
0.734
2.171
3.423
0.197
0.162
2.803
0.472
2.079
0.445
2.228
0.847
1.402
1.425
0.263
2.936
0.394
3.074
2.555
0.452
0.173
0.578
AVS
Cwmol/g)
1.400
0.742
0.029
0.028
3.740
1.080
1.230
0.087
0.948
0.283
0.490
13.400
0.024
81.100
3.340
0.331
72.400
8.480
6.460
0.034
0.976
3.210
68.000
3.150
67.700
3.310
4.870
2.420
0.430
116.000
1.300
0.976
1.220
0.025
3.430
0.621
25.000
5.610
138.000
0.892
3.590
11.900
12.500
26.600
0.056
15.100
17.300
52.700
22.300
0.079
29.600
0.031
10.400
0.402
0.480
0.201
0.257
SEM-AVS
Cumol/g)
-1.111
0.758
0.037
0.106
-3.474
-0.814
0.062
0.260
-0.198
-0.071
0.007
-12.776
0.008
-80.112
-2.736
-0.300
-70.803
-7.415
-6.271
-0.016
-0.897
-2.789
-67.202
-2.247
-66.498
-3.151
-4.624
-1.733
0.269
-114.337
-1.217
-0.236
-0.342
0.019
-2.520
-0.054
-24.266
-3.439
-134.577
-0.695
-3.428
-9.097
-12.028
-24.521
0.389
-12.872
-16.453
-51.298
-20.875
0.184
-26.664
0.363
-7.326
2.153
-0.028
-0.028
0.321
Survival15
%
100
98
99
103
99
102
107
102
99
108
103
113
101
101
107
98
102
93
103
99
97
111
104
99
105
104
106
93
91
100
99
101
98
106
104
104
107
102
100
107
82
101
101
94
106
103
99
109
88
84
100
87
104
96
100
98
101
Significance0
%
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
TOC
%
0.60
2.68
0.17
0.14
0.49
0.56
1.80
0.30
0.95
0.37
1.00
1.58
0.11
3.36
1.38
0.09
4.19
3.17
0.32
0.15
0.14
0.49
2.84
2.85
2.28
0.51
0.71
1.70
2.05
4.12
0.14
2.30
2.84
0.15
3.00
0.76
2.21
2.57
4.14
0.37
0.81
2.36
2.77
3.18
0.20
2.92
2.38
2.70
3.14
0.27
4.15
0.18
2.47
2.18
1.07
0.22
0.65
B-l
-------
Appendix B
Study3
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
EMAP-VA
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
SEM
Cumol/g)
0.209
5.411
1.298
1.039
0.960
7.369
1.380
4.259
8.229
3.535
2.543
2.124
0.188
0.229
1.820
3.468
1.622
0.693
0.294
0.178
0.223
0.239
0.801
0.751
0.299
0.341
0.205
2.415
0.632
1.516
3.249
0.462
0.043
0.050
1.177
0.624
0.799
0.020
0.088
2.220
0.813
0.851
0.701
1.113
0.601
1.505
0.701
0.717
2.163
0.616
2.368
1.278
2.253
0.865
0.950
1.113
AVS
Gwmol/g)
3.460
17.800
0.228
0.705
12.900
3.460
2.270
54.600
68.000
61.800
35.600
35.600
0.836
0.692
0.227
14.600
6.080
1.200
0.026
0.074
0.087
1.120
5.120
0.090
0.090
0.174
0.611
4.050
28.200
52.700
12.300
6.140
0.024
0.025
3.460
6.210
29.700
0.259
4.150
59.600
0.381
0.029
3.600
3.510
6.440
18.730
5.630
13.090
65.310
6.940
19.990
4.710
59.590
3.880
16.520
14.950
SEM-AVS
Cumol/g)
-3.251
-12.389
1.070
0.334
-11.940
3.909
-0.890
-50.341
-59.771
-58.265
-33.057
-33.476
-0.648
-0.463
1.593
-11.132
-4.458
-0.507
0.268
0.104
0.136
-0.881
-4.319
0.661
0.209
0.167
-0.406
-1.635
-27.568
-51.184
-9.051
-5.678
0.019
0.025
-2.283
-5.586
-28.901
-0.239
-4.062
-57.380
0.432
0.822
-2.899
-2.397
-5.839
-17.225
-4.930
-12.373
-63.147
-6.324
-17.622
-3.432
-57.337
-3.015
-15.570
-13.837
Survival15
%
96
100
100
102
94
87
97
76
43
99
33
0
108
95
104
102
102
99
95
81
104
88
92
102
104
105
95
100
88
85
103
108
100
102
100
104
100
96
100
74
93
87
100
96
96
93
93
93
92
92
91
91
91
91
90
89
Significance0
%
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
TOC
%
0.36
2.78
0.51
0.30
1.91
1.86
0.25
2.47
4.98
3.19
2.50
2.15
0.35
0.46
1.90
2.08
2.02
1.11
0.38
0.42
0.43
0.31
1.88
0.66
0.43
0.99
0.71
2.25
3.35
7.01
3.29
2.19
0.18
0.17
1.83
2.25
4.10
0.30
0.25
2.18
0.98
0.57
0.74
1.12
1.43
2.56
0.77
2.05
3.22
0.81
3.02
1.81
2.51
1.32
1.52
2.00
B-2
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Study3
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- LI
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
SEM
Cumol/g)
1.026
1.446
2.777
0.211
2.665
2.813
1.235
2.198
3.624
3.594
1.342
2.462
0.964
0.332
2.311
0.623
0.896
0.544
0.641
0.355
0.222
2.262
1.307
1.963
2.785
4.333
1.927
0.004
3.831
0.808
1.783
2.622
0.597
1.181
1.862
2.726
2.102
2.471
1.870
1.607
4.942
2.705
2.087
1.514
2.629
3.194
0.872
1.080
0.123
2.914
2.218
2.609
3.650
1.634
1.267
2.892
AVS
Gwmol/g)
0.850
12.480
29.720
0.090
78.900
35.050
2.080
14.690
21.800
27.410
37.970
46.450
1.000
4.010
79.890
6.610
16.370
2.170
2.060
1.390
4.180
39.960
0.380
51.820
61.020
16.080
3.710
24.580
9.250
0.960
40.630
61.840
1.090
3.730
50.390
62.760
33.630
7.220
17.120
17.810
100.800
83.010
26.730
30.880
32.050
35.390
25.810
11.300
5.310
2.893
2.369
43.959
101.984
5.237
3.256
80.584
SEM-AVS
Oumol/g)
0.176
-11.034
-26.943
0.121
-76.235
-32.237
-0.844
-12.492
-18.176
-23.816
-36.628
-43.988
-0.036
-3.678
-77.579
-5.987
-15.475
-1.626
-1.419
-1.035
-3.958
-37.698
0.927
-49.857
-58.235
-11.747
-1.783
-24.576
-5.419
-0.152
-38.847
-59.218
-0.493
-2.549
-48.528
-60.034
-31.528
-4.749
-15.250
-16.203
-95.858
-80.305
-24.643
-29.366
-29.421
-32.196
-24.938
-10.220
-5.187
0.021
-0.151
-41.350
-98.334
-3.603
-1.989
-77.692
Survival15
%
88
88
87
87
87
86
84
84
83
82
82
82
81
81
81
80
80
79
79
79
77
77
76
76
76
75
75
74
73
71
70
70
69
68
67
67
64
63
61
59
54
53
47
42
39
37
34
16
10
8
15
26
29
36
52
83
Significance0
%
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
TOC
%
1.63
2.05
2.81
0.54
3.33
3.83
1.58
2.80
2.48
2.59
1.85
3.18
1.60
1.29
3.69
0.67
1.11
0.27
1.56
0.64
0.45
2.67
1.56
3.46
3.81
3.48
1.60
2.87
3.08
1.19
2.50
3.49
0.76
0.91
2.81
2.81
3.42
2.80
3.29
2.07
3.15
3.62
3.45
2.69
2.68
3.17
1.83
1.91
0.22
3.05
2.89
3.74
1.83
1.72
1.53
6.98
B-3
-------
Appendix B
Study3
NOAA- BO
NOAA- BO
NOAA- BO
NNOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA- BO
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
SEM
Cumol/g)
2.511
0.661
2.458
1.872
0.959
2.480
0.784
0.943
1.683
1.753
2.447
1.839
1.296
1.697
1.390
2.310
0.399
2.481
1.736
0.958
9.192
1.525
0.678
5.037
4.202
1.174
1.855
3.092
2.997
2.581
2.869
5.442
2.618
5.061
2.376
6.998
4.480
4.662
5.896
3.103
1.662
3.512
0.273
0.335
1.664
2.674
5.532
4.029
4.614
3.379
4.240
4.303
5.209
4.801
4.697
2.600
AVS
Gwmol/g)
2.241
13.490
23.077
48.062
53.288
7.599
22.486
8.831
42.399
17.697
10.958
68.306
56.838
9.089
43.801
51.857
3.899
19.604
148.969
18.622
120.622
81.842
5.679
69.320
21.980
27.540
14.170
51.770
79.710
61.050
28.080
25.900
1.080
12.240
4.390
63.450
20.780
23.720
51.580
59.780
7.230
25.840
0.050
0.036
18.760
3.630
29.210
18.440
20.530
30.120
19.320
22.570
14.570
35.370
54.710
56.730
SEM-AVS
Cumol/g)
0.270
-12.829
-20.619
-46.190
-52.329
-5.119
-21.702
-7.888
-40.716
-15.944
-8.511
-66.467
-55.542
-7.392
-42.411
-49.547
-3.500
-17.123
-147.233
-17.664
-111.430
-80.317
-5.001
-64.283
-17.778
-26.366
-12.315
-48.678
-76.713
-58.469
-25.211
-20.458
1.538
-7.179
-2.014
-56.452
-16.300
-19.058
-45.684
-56.677
-5.568
-22.328
0.223
0.299
-17.096
-0.956
-23.678
-14.411
-15.916
-26.741
-15.080
-18.267
-9.361
-30.569
-50.013
-54.130
Survival15
%
86
87
87
89
90
90
91
91
92
94
94
95
96
97
97
97
99
99
99
99
100
102
103
0
41
11
18
101
112
119
81
95
109
97
108
0
20
14
2
77
19
0
91
93
69
3
96
51
91
88
101
102
101
70
38
37
Significance0
%
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
0
0
1
1
0
1
0
0
0
0
0
1
1
1
TOC
%
2.12
1.00
3.15
3.25
2.39
4.45
1.88
1.78
3.41
1.41
4.45
2.54
3.05
2.68
3.27
3.35
0.80
3.31
2.94
1.77
4.61
2.96
1.45
5.02
3.47
1.88
4.44
3.86
3.09
2.86
2.50
2.20
2.67
2.98
2.49
1.98
2.98
3.19
4.78
3.99
2.61
4.44
0.07
0.07
0.69
1.00
3.18
2.20
1.94
2.80
3.15
3.02
3.21
2.98
3.47
1.47
B-4
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Study3
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA-HR
NOAA- BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-BA
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
SEM
Cumol/g)
1.013
1.527
0.505
3.341
3.449
0.270
0.341
0.888
0.722
0.362
2.138
3.008
0.151
0.115
0.543
0.103
0.167
0.073
0.294
0.120
0.109
0.185
0.120
0.347
0.120
2.275
0.344
0.258
0.119
0.258
0.494
0.109
0.266
0.327
0.230
2.026
14.550
3.332
3.763
0.357
0.524
0.244
1.247
2.478
1.744
0.131
0.846
4.399
3.884
0.673
3.150
0.270
0.162
2.880
0.323
0.413
AVS
Gwmol/g)
10.160
15.130
0.630
43.920
37.860
0.950
0.156
12.971
4.948
0.936
3.295
3.941
0.555
0.156
0.156
0.156
0.932
0.156
0.156
0.156
0.156
0.156
0.156
0.156
0.156
16.592
0.012
0.343
0.156
0.156
0.156
0.156
0.156
0.393
6.400
47.793
389.857
243.322
201.687
10.923
3.974
4.502
48.130
47.376
0.156
1.184
0.927
116.954
237.650
21.769
43.975
4.491
0.873
153.755
1.684
3.056
SEM-AVS
Oumol/g)
-9.147
-13.603
-0.125
-40.579
-34.411
-0.680
0.185
-12.083
-4.226
-0.574
-1.157
-0.933
-0.404
-0.041
0.387
-0.053
-0.765
-0.083
0.138
-0.036
-0.047
0.029
-0.036
0.191
-0.036
-14.317
0.332
-0.085
-0.037
0.102
0.338
-0.047
0.110
-0.066
-6.170
-45.767
-375.307
-239.990
-197.924
-10.566
-3.450
-4.258
-46.883
-44.898
1.588
-1.053
-0.081
-112.555
-233.766
-21.096
-40.825
-4.221
-0.711
-150.875
-1.361
-2.643
Survival15
%
29
68
105
86
76
96
84
92
85
98
95
95
96
99
94
85
97
99
91
84
92
90
88
89
81
69
91
94
84
91
86
89
86
93
83
51
0
37
79
95
98
84
91
36
69
94
73
93
89
77
91
91
98
92
93
94
Significance0
%
1
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
1
1
1
0
0
0
0
1
1
0
1
0
0
1
0
0
0
0
0
0
TOC
%
0.77
0.95
0.25
2.55
3.63
0.26
0.06
4.05
0.40
0.26
0.43
0.18
0.15
0.08
0.07
0.05
0.16
0.05
0.34
0.83
0.92
4.48
0.83
1.26
0.62
1.81
3.85
0.77
2.23
0.88
2.10
4.07
1.06
0.29
0.19
0.77
1.52
0.83
0.97
0.26
0.35
0.27
0.54
1.12
1.14
0.21
1.58
6.55
8.45
4.11
5.47
0.74
1.40
7.70
0.20
1.20
B-5
-------
Appendix B
Study3
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-JB
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-LS
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
SEM
Cumol/g)
0.377
0.099
1.100
0.209
0.213
0.954
2.759
0.711
1.915
2.186
2.480
0.606
3.289
3.241
0.616
1.506
2.485
1.894
3.149
0.632
1.057
0.638
1.087
3.711
2.990
8.894
1.277
3.925
5.632
6.809
7.645
4.012
3.905
0.942
3.515
2.216
3.323
3.391
3.443
2.466
2.294
5.768
1.013
2.479
0.554
5.222
5.116
14.791
4.917
0.398
4.855
3.290
5.822
9.167
6.214
0.794
AVS
Gwmol/g)
3.056
0.686
58.945
1.466
0.780
1.542
6.498
10.240
12.596
17.605
23.523
2.501
91.773
56.100
1.070
26.201
28.248
25.394
64.643
1.310
4.647
0.218
0.312
17.184
59.256
60.816
23.266
42.727
114.770
135.354
150.012
43.663
26.229
6.531
7.134
11.243
7.573
4.820
3.982
20.273
11.046
5.028
11.079
25.687
2.634
22.617
7.352
109.780
0.530
0.218
9.606
10.105
51.460
93.563
42.415
2.651
SEM-AVS
Cumol/g)
-2.679
-0.587
-57.845
-1.257
-0.567
-0.588
-3.739
-9.529
-10.681
-15.419
-21.043
-1.895
-88.484
-52.859
-0.454
-24.695
-25.763
-23.500
-61.494
-0.678
-3.590
0.420
0.775
-13.473
-56.266
-51.922
-21.989
-38.802
-109.138
-128.545
-142.367
-39.651
-22.324
-5.589
-3.619
-9.027
-4.250
-1.429
-0.539
-17.807
-8.752
0.740
-10.066
-23.208
-2.080
-17.395
-2.236
-94.989
4.387
0.180
-4.751
-6.815
-45.638
-84.396
-36.201
-1.857
Survival15
%
92
93
96
93
95
83
96
97
97
95
99
98
95
97
95
96
96
93
93
87
90
92
90
88
80
85
92
90
86
91
92
86
89
84
87
86
85
83
95
82
84
75
90
83
84
83
9
8
89
94
83
60
41
25
68
93
Significance0
%
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
1
1
0
0
0
1
1
1
1
0
TOC
%
1.30
0.75
3.86
0.58
0.69
0.26
0.45
0.56
0.21
0.27
0.32
0.25
0.77
1.14
0.15
0.95
0.25
0.98
0.90
1.51
2.44
3.52
7.36
3.99
5.24
3.63
3.18
3.85
4.29
4.36
6.04
3.73
3.93
0.67
0.75
1.22
1.25
1.05
0.88
1.40
0.95
1.77
0.76
0.99
0.60
1.48
1.45
9.15
3.10
2.42
2.62
5.70
2.22
6.48
3.24
2.36
B-6
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Study3
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-NB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-RB
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
SEM
Cumol/g)
4.985
5.280
2.268
6.678
2.833
0.333
0.756
0.582
1.012
1.596
0.326
2.709
5.485
3.596
5.329
0.337
0.986
0.856
5.364
1.706
0.371
0.193
0.869
1.288
1.650
2.422
0.512
4.198
5.081
6.095
8.471
3.370
1.198
2.127
1.360
1.197
1.975
2.829
2.830
1.385
1.519
3.186
2.086
1.799
0.930
0.459
0.889
0.833
1.317
2.480
0.626
1.500
0.723
4.158
2.241
2.907
AVS
Gwmol/g)
43.663
1.934
6.300
17.559
45.222
22.315
1.216
0.821
0.567
0.447
0.156
3.120
14.666
19.503
4.321
2.901
0.156
0.156
39.700
23.515
4.210
0.156
19.617
0.593
0.624
0.156
0.156
4.086
36.490
5.957
8.078
17.247
0.156
12.446
1.790
3.373
17.136
25.189
56.401
44.588
11.549
86.235
11.713
12.631
10.093
0.156
2.623
2.464
15.563
32.123
9.949
5.427
1.341
13.504
27.788
29.285
SEM-AVS
Oumol/g)
-38.678
3.346
-4.032
-10.881
-42.389
-21.982
-0.460
-0.239
0.445
1.149
0.170
-0.411
-9.181
-15.907
1.008
-2.564
0.830
0.700
-34.336
-21.809
-3.839
0.037
-18.748
0.695
1.026
2.266
0.356
0.112
-31.409
0.138
0.393
-13.877
1.042
-10.319
-0.430
-2.176
-15.161
-22.360
-53.571
-43.203
-10.030
-83.049
-9.627
-10.832
-9.163
0.303
-1.734
-1.631
-14.246
-29.643
-9.323
-3.927
-0.618
-9.346
-25.547
-26.378
Survival15
%
53
83
16
77
54
93
92
94
94
95
93
70
92
62
91
97
96
96
91
93
91
92
85
92
91
98
93
90
89
4
91
94
94
83
99
92
45
84
96
88
82
93
82
37
89
98
95
86
88
87
97
89
89
96
70
95
Significance0
%
1
0
1
1
1
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
TOC
%
3.90
6.10
1.99
15.20
2.02
1.23
0.33
0.30
0.30
0.17
0.08
0.42
2.29
0.88
0.97
0.53
0.12
0.51
1.17
3.21
3.54
2.52
2.39
2.44
2.68
2.60
0.42
2.63
2.08
3.03
5.30
3.91
1.03
3.43
1.26
5.85
2.33
0.91
1.21
1.03
1.06
1.39
0.79
1.06
0.43
0.13
0.21
4.96
2.56
3.06
2.58
2.71
3.89
4.78
2.66
5.15
B-7
-------
Appendix B
Study3
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
REMAP-UH
SEM
Cumol/g)
0.852
2.294
2.995
2.981
0.677
AVS
Gwmol/g)
1.591
53.955
33.995
44.910
10.323
SEM-AVS
Cumol/g)
-0.739
-51.661
-31.000
-41.929
-9.646
Survival15
%
93
15
88
94
91
Significance0
%
0
1
0
0
0
TOC
%
2.03
4.37
3.55
2.97
3.32
^Sources: EMAP-VAis U.S. EPA, 1996. NOAA-LI is Wolfe etal., 1994. NOAA-BO is Long etal., 1996. NOAA-HR is
Long et al., 1995b. REMAP is Adams et al, 1996.
bConclusion of significance varies for three databases. EMAP significance based on percent survival of control. NOAA
significance based on percent survival less than 80%. REMAP significance based on percent survival less than 80%.
0 Significance: 0, no significant toxicity; 1, significant toxicity.
B-8
-------
Appendix C
Quality Assurance Summary for the ESB Document:
Procedures for the derivation of equilibrium
partitioning sediment benchmarks (ESBs)
for the protection of benthic organisms: Metal Mixtures
(Cadmium, Copper, Lead, Nickel, Silver, and Zinc)
c-i
-------
Appendix C
All data were obtained either from the WQC document for the metals cadmium, copper, lead, nickel, silver, and zinc
(USEPA, 1980,1985b, c, d, 1986,1987) or from a comprehensive literature search completed in 1999 and updated
in 2004. Data for the chromium appendix was obtained from a comprehensive literature search completed in 2004.
All data used in the example benchmark calculations were evaluated for acceptability using the procedures outlined in
the Stephan et al. (1985): Guidelines for deriving numerical national water quality criteria for the protection of
aquatic organisms and their uses. Data not meeting the criteria were rejected. The approach for deriving the values
in thisdocument were also reviewed by the U.S. EPA SAB (U.S. EPA, 1994a; 1995a; 1999). All calculations were
made using the procedures in Stephan et al. (1985). This document was reviewed for scientific quality assurance by
U.S. EPA Office of Water and Office of Research and Development scientists.
Hard copies of all literature cited in this document reside at ORD/NHEERL Atlantic Ecology Division - Narragansett,
Rhode Island.
C-2
-------
Appendix D
Procedures for the Derivation of Equilibrium Partitioning
Sediment Benchmark (ESBs) for the Protection
of Benthic Organisms: Chromium
D-l
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Executive Summary
Chromium exists in sediments primarily in
two oxidation states: Cr(III), which is relatively
insoluble and nontoxic, and Cr(VI), which is
much more soluble and toxic. Cr(VI) is
thermodynamically unstable in anoxic sediments
and AVS is formed only in anoxic sediments;
therefore sediments with measurable AVS
concentrations should not contain toxic Cr(VI). If
this "chromium hypothesis" holds true, measuring
AVS could form the basis for an ESB for
chromium in sediments.
A review of the literature and recently
performed experiments with both freshwater and
saltwater sediments support the chromium
hypothesis.
In saltwater:
1) Survival of amphipods was decreased by
waterborne Cr(VI), with a 10-day median lethal
concentration (LC50) of 1850 |ig/L Cr(VI).
2) Survival of amphipods was not decreased by
waterborne Cr(III) at concentrations well above
saturation.
3) In both laboratory-spiked sediments with
Cr(III) and Cr(VI) and field-contaminated
sediments, in sediments where detectable AVS
was present, chromium concentrations in
interstitial water were very low (<100 |lg/L)
and no significant lethality to A. abdita was
observed. In sediments in which AVS was not
significantly greater than zero, chromium
concentrations in interstitial waters increased
significantly, with greater than 90% of the
chromium present as Cr(VI), and the mortality
of A. abdita was elevated.
In freshwater:
1) Survival of amphipods was decreased by
waterborne Cr(VI), with a 42-day LC50 of 40
jlg/L.
2) Cr(VI) spiked into test sediments with differing
levels of AVS resulted in graded decreases in
AVS.
3) Sediments with low AVS concentrations (<1
|imol/g) after spiking with Cr(VI) caused 100%
mortality of amphipods, but no toxic effects
were observed in Cr(VI)-spiked sediments that
maintained higher AVS concentrations.
4) Waterborne Cr(III) levels near solubility limits
caused decreased survival of amphipods at pH
7 and pH 8, but not at pH 6.
5) Sediments spiked with high levels of Cr(III)
had no effect on amphipod survival, but caused
significant decreases in reproduction and/or
growth.
6) Interstitial waters of some Cr(III)-spiked
sediments contained measurable concentrations
of Cr(VI), but observed toxic effects did not
correspond closely to concentrations of
aqueous Cr species.
Thus, although both Cr(VI) and Cr(III) could
be toxic to H. azteca in water and sediment, risks
of Cr toxicity were low in sediments containing
substantial concentrations of AVS. Results
presented in this appendix suggest that
measurements of AVS and interstitial water
chromium can be useful in predicting the absence
of acute effects from chromium contamination in
both freshwater and saltwater sediments. In
sediments with substantial AVS, risks of
chromium toxicity should be low, because the
chromium will be present in the form of Cr (III).
This should apply to any sediment with SEM-AVS
< 0.0. Sediments with SEM-AVS > 0.0, but which
have substantial AVS present may be toxic due to
copper, cadmium, lead nickel, or zinc, but should
not be toxic due to chromium or silver. The
relationship, (SEM-AVS)//^, should be used with
caution (with regard to chromium toxicity) in
sediments with little or no AVS, because a
sediment with no appreciable AVS or SEM and
substantial chromium might be toxic due to
chromium, even though no toxicity due to the
other metals would be expected. These findings
form the basis for a chromium ESB.
D-2
-------
Section 1
Introduction
Appendix D
Chromium is often found in contaminated
sediments (Pawlitz et al., 1977). Elevated
chromium concentrations in sediments are usually
associated with tanneries, smelters, and plating
facilities. However, without a good understanding
of the adverse biological effects of chromium in
sediments, it is difficult to know what
concentration of chromium in sediment may
present ecological risk to benthos.
Although there have been several studies on
the bioaccumulation of chromium from
laboratory-spiked sediments (Wang et al., 1997;
Griscom et al., 2000; Fan and Wang., 2001), there
are few published studies on biological effects of
chromium in laboratory-spiked sediments other
than uptake of chromium. There are also very few
reports on effects of chromium in field sediments.
Leslie et al. (1999) found that a tributary below a
chromium salt processing plant was incapable of
supporting benthic macrofauna, presumably
because of chromium leaching from stock piles
along the banks of the tributary, but concluded
that much of the chromium might be coming from
the water rather than the sediment. In a study of
sediments associated with a tannery, some toxicity
was observed in ten-day static toxicity tests with
several sediments with chromium in excess of
4000 |ig/g; the same sediments, however,
exhibited no toxicity in 28-day flow-through tests,
suggesting that the toxicity observed in the 10-day
static test was related to test conditions and
duration and not sediment chromium (HydroQual,
1994). Several other studies have found elevated
chromium concentrations in the tissues of benthos
from sediments contaminated with high levels of
chromium from mining activities (Bervoets et al.,
1998) or tannery wastes (Catsiki et al., 1994), but
these tissue concentrations were not linked to
biological effects.
Part of the difficulty in understanding the
biological effects of chromium in sediment is that
chromium exists in sediments in two oxidation
states, Cr(III) and Cr(VI), each with very different
geochemical properties and toxicological effects.
Cr(VI) is highly oxidized and unstable in reducing
and even moderately oxidizing environments
(DeLaune et al., 1998, Masscheleyn et al., 1992).
Cr(VI) is also very soluble and highly toxic, while
Cr(III) has very low solubility at environmentally
relevant pH (DeLaune et al., 1998; Barnhart.,
1997) and is generally thought to have relatively
low toxicity (Wang et al., 1997; Thompson et al.,
2002). For example, Leslie et al. (1999) assumed
that the effects they saw due to chromium must
have been caused by Cr(VI). However, they did
not measure the chromium speciation.
This appendix provides the technical basis for
the derivation of an ESB for chromium analogous
to the ESB for the cationic metals cadmium,
copper, lead, nickel, silver, and zinc discussed
earlier. Determining the relationship between
AVS and chromium in sediments would extend
the utility of AVS measurements as a part of
sediment assessments. Chromium should not
necessarily be included among the SEM metals
because its interaction with AVS is not via
formation of an insoluble sulfide, but rather
oxidation of sulfide and concomitant reduction of
chromium. However, the geochemical relationship
between AVS and chromium and the toxicological
differences between oxidation states of chromium
might be used to develop a theoretically-derived
benchmark through what is called the "chromium
hypothesis." The hypothesis is based on the
concepts that Cr(III) is much less soluble and
toxic than Cr(VI) and that Cr(VI) is not stable in
reducing environments such as anoxic sediments
in which AVS is formed. Thus, in a sediment
where AVS is present, chromium will exist solely
as Cr(III), and therefore the interstitial water
should contain little chromium and the sediment
should not be toxic due to chromium.
Although there is literature discussing
chromium toxicity and geochemistry, no studies
were available which had tested the "chromium
hypothesis" directly. To this end, recently,
D-3
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
experiments have been carried out with both
freshwater (Besser et al., in press) and saltwater
sediments (Berry et al., in press) to verify the
chromium hypothesis. In the saltwater test series
ten-day water-only and ten-day spiked sediment
toxicity tests with the amphipod Ampelisca abdita
were performed with Cr(VI) and Cr(III). Ten-day
sediment tests with saltwater sediments collected
from a site contaminated with high concentrations
of chromium were also performed. In freshwater
sediments, chronic (28- to 42-d) water-only and
spiked sediment toxicity tests with the amphipod
Hyalella azteca were performed with Cr(VI) and
Cr(III).
D-4
-------
Appendix D
Section 2
Chemistry of Chromium in Sediment
2.1 Valence States of Chromium in
Sediments
Studies by many researchers have provided a
generalized model of the cycling of chromium
between redox states in various aquatic soils/
sediments and their interstitial and overlying
waters (Masscheleyn et al., 1992; Kozuh et al.,
2000; Hassan and Garrison., 1996; Mattuck and
Nikolaidis., 1996). This model is characterized by
the relative stability of Cr(VI) in oxygenated
overlying waters, particularly in marine waters,
and rapid removal of Cr(III) through precipitation
of the insoluble hydroxide and adsorption onto
particulate matter. In freshwater systems with
elevated dissolved organic carbon concentrations,
such as in wetland soils and waters, a considerable
amount of Cr(III) may be organically complexed,
which slows the rate of removal to the particulate
matter. In some circumstances, Cr(III) may be
oxidized to Cr(VI) by Fe/Mn-rich films on the air-
water interface where reduced Mn(II) and Fe(II)
diffuses from sediments into oxic overlying water
(Masscheleyn et al., 1992).
In sediments and soils, the reactivities of
Cr(III) and Cr(VI) are somewhat reversed. Cr(III)
may be oxidized to Cr(VI) in soils or sediments
with high concentrations of MnO2 and low organic
content apparently by oxidation at MnO2 surfaces
(Hassan and Garrison., 1996). Similar oxidation
by resuspended sediments rich in manganese
oxides has also been postulated as the cause for
relatively higher concentrations of Cr(VI) in deep
ocean seawater relative to seawater overlying
reduced coastal sediments (Nakayama et al.,
1981). If organic content is elevated, however,
Cr(III) is not oxidized, even in highly oxidizing
sediments (Masscheleyn et al., 1992; Kozuh et al.,
2000). On the other hand, Cr(VI) is reduced to
Cr(III) and almost completely removed from
solution in even moderately oxidizing sediments
(redox potential Eh < 300 mV). In more reducing
sediments (Eh < 200 mV), reduction is
significantly more rapid due to reaction with
ferrous ionic Fe(II). In such reduced sediments,
very high partitioning constants indicate that
almost all chromium is bound to the sediment,
presumably as Cr(III), with very little mobile in
interstitial waters (Mattuck and Nikolaidis.,
1996). In wetland sediments, much of the
dissolved chromium may be organically
complexed Cr(III) (Icopini and Long, 2002). Once
the reductive capacity of soils or sediments is
exceeded, concentrations of dissolved Cr(VI)
increase sharply and remain stable. Although the
reductive capacity of sediments is generally
proportional to organic content, the primary
reductant is more likely Fe(II) or, in sulfidic
sediments, sulfide.
2.2 Geochemical Distribution of
Chromium in Toxicological Exposures
The geochemical distributions found in the
recent experiments with marine sediments (Berry
et al., in press) amended with Cr(VI) and Cr(III)
reflected the behavior described in the previous
section. Although the sediments had differing
characteristics such as silt/clay and organic
contents, they were both reducing sediments, as
evidenced by the presence of AVS. Cr(III) added
to these sediments in massive quantities was
essentially inert to redox transformation: no
Cr(VI) was evident in sediments, interstitial
waters or overlying waters throughout the
experiment, as was the case with organic-rich peat
soils and wetland sediments. Cr(VI) added to the
sediments was reduced completely and rapidly
(<1 day), regardless of the amount added, up to
the reductive capacity of the sediments. No
significant amount of chromium was evident in
either interstitial or overlying waters, indicating
that Cr(III) complexed by dissolved organic
carbon was not important in these sediments.
Once the capacity of the sediments were
exceeded, very high concentrations of chromium,
almost entirely Cr(VI), were evident in interstitial
D-5
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
waters (Figure D-l). Concentrations of chromium
in overlying waters decreased throughout the
experiment and remained primarily as Cr(VI).
These geochemical controls on the concentrations
and redox speciation of chromium in sediments
constrain exposure and consequent biological
effects of the chromium on benthic organisms,
such as the amphipods used in the saltwater tests.
In a recent study, spiking of freshwater
sediments (Besser et al., in press) with several
levels of Cr(VI) resulted in graded decreases in
AVS concentrations and changes in POC
concentrations in sediment and interstitial water.
For example, mean AVS concentrations decreased
by up to 97% in study sediments on day 0 of the
test. Sediment TOC also decreased slightly in
Cr(VI)-spiked sediments. Cr(VI) spikes were
associated with increased DOC, increased
alkalinity, and decreased hardness in interstitial
waters.
Overlying and interstitial water Cr(VI)
concentrations reflected differences in AVS
concentrations among treatments. Initial
concentrations of Cr(VI) in interstitial water
samples were greater than 10,000 |ig/L in
treatments with the lowest AVS concentrations
while interstitial water Cr(VI) concentrations
remained low (<20 |lg/L) in treatments with
higher AVS concentrations. In all three treatments
with quantifiable Cr(VI) in interstitial water,
concentrations decreased during the test. Cr(VI)
concentrations in overlying water were much
lower than those in interstitial water, but followed
similar trends among treatments and over time.
Decreases in Cr(VI) concentrations during the
course of the study may have resulted from
reactions with AVS and POC and from dilution
due to replacement of overlying water. The
smallest proportional decrease of Cr(VI) in
interstitial and overlying water occurred in the
treatment which had no AVS and low POC.
These results are consistent with the
hypothesis that Cr(VI) concentrations remain low
in sediments containing substantial concentrations
of AVS. Berry et al. (impress), in studies with
Cr(VI)-spiked marine sediments, did not detect
Cr(VI) in interstitial waters of sediments spiked
with Cr(VI) at CrAVS ratios of 2.2 or less. In
contrast, substantial Cr(VI) concentrations were
measured in sediments spiked at 3:1 CrAVS
ratios (Besser et al., in press). AVS may have
persisted in these treatments due to regeneration
of AVS during the test, at least in some treatments.
However, the data also suggested that some added
Cr(VI) reacted with sediment POC, as has been
reported in several previous studies (Wittbrodt
and Palmer., 1995; Elovitz and Fish., 1995;
U.S.EPA., 2002; Poleo., 1995). The relationship
between Cr(VI) spikes and AVS depletion in one
freshwater sediment was similar to the 2:1 ratio
reported by Berry et al. (impress). However, AVS
concentrations in a high-POC sediment decreased
in a proportion of about one mole of AVS per
eight moles of added Cr(VI). Reaction of Cr(VI)
spikes with sediment POC is also suggested by
decreases in organic carbon in several sediments
and increases in interstitial water DOC in all three
sediments. These results suggest that sediment
POC also provides protection against Cr(VI)
lethality in benthic environments, although Cr(VI)
lethality occurred in some of the spiked sediments
despite high levels of OM.
D-6
-------
Appendix D
Sediment 1
Sediment 2
V V
Nominal Cr:AVS
IAVSDTotalCrD[Cr(VI)]
1,000,000
- 800,000 ^
5*
- 600,000 5.
- 400,000 j|
O
- 200,000 "
1,000,000
O)
O
Figure D-l. Concentrations of AVS in sediment, and total Cr and Cr(VI) in interstitial
waters of two saltwater sediments spiked with Cr (VI).
D-7
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Section 3
Chromium Toxicity in Water and
Sediment
3.1 Chromium Toxicity in Water-only
Tests
Water-only tests were performed to
complement the sediment toxicity tests conducted
by Berry et al. (in press) and Besser et al. (in
press). Survival of amphipods was decreased by
Cr(VI) in water-only tests in saltwater, calculated
10-day LCSOs were 1980 and 1854 |ig Cr/L based
on dissolved and hexavalent concentrations,
respectively (Berry et al., in press). Survival of
amphipods was not decreased by Cr(III) in water-
only tests in saltwater at concentrations well
above saturation.
Exposure to water-only Cr(VI) in freshwater
caused decreased survival offf. azteca (Besser et
al., impress) (Table D-l). Identical LCSOs (40 |ig/
L) were determined for the 28- and 42-d exposure
periods suggesting that lethal effects of Cr(VI)
occurred early in the exposure. Evidence of
sublethal effects of Cr(VI) on amphipods was less
conclusive. Amphipod growth was not
significantly decreased at any Cr(VI) exposure
level, but reproduction in all Cr(VI) treatments
was at least one-third less than controls. These
results indicate that H. azteca is highly sensitive
to chronic toxicity of Cr(VI). Excluding the
reproduction data, the threshold for chronic
Cr(VI) toxicity to H. azteca was 15 |ig/L
(geometric mean of Cr concentrations bracketing
the lowest significant toxic effect) slightly greater
than the current U.S.EPA water quality criterion
for Cr(VI) of 10 |ig/L (Richard and Bourg., 1991,
U.S.EPA., 1995). Previous studies have reported
chronic values for Cr(VI) between 6 |ig/L and 40
|ig/L for crustacean zooplankton and between 264
M£/L and 1987^ig/L for fish (U.S.EPA, 1986).
Toxicity of Cr(III) in freshwater water-only
tests was measured at 3 pHs: 6,7, and 8 (Besser et
al., impress). Cr concentrations in the Cr(III)
water-only test were less than the nominal
concentration of 100 |lg/L, indicating that Cr(III)
concentrations were limited by solubility (Table
D-2). Filterable Cr concentrations were highest at
Table D-l. Results of a toxicity test with the amphipod H. azteca exposed to Cr(VI) in water.
Means with standard error in parentheses. Asterisks indicate significant
difference between treatment and control (p<0.05; ANOVA and Dunnett's test with
log-transformed data). From Besser et al. (in press).
Cr (ug/L)
(n=4)
<2.0
2.0 (0.3)
4.7(1.1)
10(1.0)
18(6)
48(2)
Survival (%)
Day 28
(n=12)
100 (0)
90(7)
95(3)
98(3)
88 (5)*
38 (5)*
Survival (%)
Day 42
(n=8)
100 (0)
90(7)
95(5)
95(3)
80 (4)*
40 (9)*
Length (mm)
Day 42
(n=8)
4.87(0.11)
4.90 (0.07)
4.84 (0.09)
5.27 (0.08)
5.05 (0.07)
5.22(0.17)
Reproduction
(young per female)
(n=8)
8.4 (2.4)
2.3(1.1)*
2.9(1.1)
5.4(1.5)
3.3(1.0)
1.6(1.0)*
D-8
-------
Appendix D
pH 6 and lowest at pH 8. Amphipod survival was
high (>90%) in controls at all three pH levels, but
control growth and reproduction were
significantly lower in pH 6 and pH 8 as compared
to pH 7. Poor performance of amphipods in the
pH 6 controls may indicate that this is near the
lower limit of pH tolerance for this species, but
growth and reproduction were also significantly
decreased at pH 8, relative to the pH 7 control.
3.2 Spiked Sediments: Saltwater
Mortality of amphipods exposed to Cr(VI) in
saltwater sediments increased with increasing
chromium concentration, but the response was
sediment dependent (Figure D-2a) (Berry et al., in
press). In sediments where detectable AVS was
present, chromium concentrations in interstitial
water were very low (<100 |lg/L). No significant
lethality to A. abdita was observed in sediments
with less than 0.5 interstitial water toxic units
(IWTU) (Figure D-2b). In sediments in which
AVS was not significantly greater than zero,
chromium concentrations in interstitial waters
increased significantly, with greater than 90% of
the chromium present as Cr(VI), and A. abdita
mortality was elevated (Figure D-2c). In a single
treatment spiked with a high concentration of
Cr(III) there was no chromium in the interstitial
water, and the sediment was not toxic (Figures D-
2a and D-2b). The results in these tests are
consistent with the chromium hypothesis, and are
similar to those for the other metals discussed in
the main document.
3.3 Field Sediments: Saltwater
Berry et al. (in press) exposed amphipods for
ten days to field sediments collected from
Shipyard Creek, a tidal creek adjacent to a former
ferrichromium alloy production facility in
Charleston, SC, USA (Breedlove et al., 2002).
The relationship between geochemical fractions
and amphipod mortality in the field sediments was
similar to that found with spiked sediments. AVS
was measured at concentrations well above
detection limits in all sediments, and despite some
exceptionally high concentrations of total
chromium (> 3000 |ig Cr/g), only traces of Cr(VI)
were detected (<4 |ig/g) in sediments, and these
concentrations were likely artifacts of the Cr(III)/
Table D-2. Results of toxicity test with the amphipod H. azteca exposed to Cr(III) in water at
three pHs. Means with range (for pH) or standard error in parentheses. Within a pH
level, asterisks indicate significant decreases in test endpoints in the Cr(III) treatment,
relative to the control. For control sediments, means followed by the same letter are
not significantly different (p^O.05; ANOVA and Fisher's LSD test with log-
transformed data). From Besser et al. (In press).
Treatment
Survival (%)
Length (mm)
Chromium
(Mg/L)
pH
Day 28
(n=12)
Day 42
(n=8)
Day 28
(n=4)
Day 42
(n=8)
Reproduction
(young/female)
Control - pH 6 <2 6.44 (6.00-7.12) 94 (1) ab 93 (2)
Cr(III)-pH6 76(63-90) 6.41(6.00-7.00) 98(2) 95(3)
3.8 (0.1) b 3.3 (0.03) c 0 (0) c
4.3(0.4) 4.3(0.4) 1.0(0.3)
Control-pH 7 <2 7.11(6.90-7.34) 90 (2) b 93(3)
Cr(III)-pH7 48(38-54) 7.24(6.96-7.42) 63(5)* 60(7)*
4.4 (0.1) a 3.9 (0.04) a 1.4 (0.2) a
4.1(0.1) 4.1(0.1) 1.3(0.6)
Control-pH 8 <2 7.98(7.79-8.20) 95 (3) a 93(4)
Cr(III)-pH8 29(23-35) 7.94(7.81-8.12) 63(4)* 53(5)*
4.0 (0.2) b 3.6 (0.04) b 0.8 (0.2) b
3.9(0.1) 3.9(0.1) 2.0(0.6)
D-9
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
100n
_ 75-
£ 50-
ro
o
5 25-
(
*
|
i
m^fcj
Mi T ^W
D 1 100
t a 100-
/ 80-
T 60-
f .-&1
| 40-
20-
. d
%
0
-------
Appendix D
Cr(VI) separation technique (Berry et al., in
press). No metals, including Cr(VI), were detected
in interstitial waters of any of the sediments,
which was expected given the large excess of AVS
over SEM measured in all of the sediments.
Despite concentrations of chromium exceeding
1700 to 3000 |ig/g in some Shipyard Creek
sediments, amphipod mortality in those sediments
(5-25%) was no greater than in sediments from
reference sites (5-20%) or a control sediment
performed in conjunction with them (5-15%)
(Berry et al., in press). These results are also
consistent with the chromium hypothesis.
3.4 Spiked Sediments: Freshwater
The 28 and 42-day mortality results from the
freshwater Cr(VI)-spiked sediment tests from
Besser et al. (in press) were very similar to those
from the 10-day saltwater Cr(VI) and Cr(III)-
spiked sediment tests described by Berry et al. (in
press). Mortality of amphipods exposed to Cr(VI)
in freshwater sediments increased with increasing
chromium concentration, but the response was
sediment dependent (Figure D-3a) (Besser et al.,
in press). In sediments where detectable AVS was
present, chromium concentrations in interstitial
water were generally very low (Besser et al., in
press). No significant toxicity to H. azteca was
observed in sediments with less than 0.5
interstitial water toxic units (IWTU) (Figure D-
3b). In sediments in which AVS was not
significantly greater than zero, chromium
concentrations in interstitial waters increased
significantly (Besser et al., in press}, with greater
than 90% of the chromium present as Cr(VI), and
mortality ofH. azteca was elevated (Figure D-3c).
Growth and reproduction were not significantly
affected in any Cr(VI) -spiked treatment that did
not show significant effects on survival (Besser et
al., in press) (Figures D-3a ,D-3b, and D-3c)
The 28 and 42-day mortality results from the
freshwater spiked sediment tests from Besser et
al. (in press) with Cr(III) were also similar to
those from the 10-day saltwater spiked sediment
tests described by Berry et al. (in press) in that
there was no increased mortality, even at high
concentrations of Cr(III). However, the chemistry
and sublethal results from the freshwater spiked
sediment tests were different from the 10-day
saltwater spiked sediment tests and the exposures
with Cr(VI)-spiked freshwater sediment in several
important respects. First, there was measurable
chromium in the interstitial water of all three
sediments spiked with a high concentration of
Cr(III) (Figure D-3b). Also, there was
significantly reduced growth in three of these
sediments (Figures D-4a, D-4b, and D-4c) and
reduced reproduction in one (Figures D-5a, D-5b,
and D-5c). Finally, the reduced growth and
reproduction was seen in some sediments which
had less than 0.5 IWTU and/or significant
amounts of AVS.
Besser et al. (in press) concluded that it was
difficult to ascribe growth and reproductive
effects in the Cr(III) - spiked sediments to
chromium toxicity, because t he measured effects
did not correspond with dissolved chromium
concentrations, or with amphipod mortality. They
hypothesized that the effects may have been a
result of the physical effect of large amounts of
chromium (presumably hydroxide) precipitate
which forms when the Cr(III) solutions are pH-
neutralized, prior to spiking (Besser et al., in
press).
D-ll
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
100-
_ 80-
£
>, 60-
"ro
o 40-
20-
n
\j -
a 100-
/ / - 8°"
/ /
/ / ^ 60"
/ / ^
/ / ~^-
/ / o 40
/ /
/ O ~f 20-
W^ o
Td
^
o° \ ^a i
1 10 100 1000 10000 100000 -100 -50 0
Total Chromium (ug/g) SEM-AVS (umoles/g)
100n
80-
^ 60-
"ro
o 40-
5
20-
n
\j -
0
+. 100-,
b
80-
~ 60-
"ro
| 40-
O "*" 20-
» e
1
01 0.1 1 10 100 1000 10000 -4000 -3000 -2000 -1000 0 1000
IWTU (SEM-AVS)/foc
100<
80-
£
>, 60-
i
° 40-
20 <
<
r\<
- Sedl-Cr(VI)
-«-Sed2-Cr(VI)
- SedS-Cr(VI)
o Sedl-Cr(lll)
>
n Sed2-Cr(lll)
>... ° o Sed3-Cr(lll)
0 20 40 60 80 100
AVS (umoles)
Figure D-3. Mortality in chromium-spiked freshwater sediment experiments (Besser et al., in
press) vs. total chromium (a), IWTU (b), AVS (c), SEM - AVS (d), and SEM -AVS//OC
(e). Where IWTU = interstitial water toxic units, AVS = acid volatile sulfide, SEM =
simultaneously extracted metal, and/oc = fraction of organic carbon. For illustrative
purposes, sediments which caused greater than 24% mortality were classified as toxic
(horizontal line) (Mearns et al., 1986). Vertical lines are drawn at 0.5 IWTU (b), 0.0
SEM-AVS (d), and 130 (SEM-AVS)//OC (e).
D-12
-------
Appendix D
E
51
4-
3-
D)
D)
!= 9-
0) ^
1 -
0
20 40 60 80
AVS (umoles)
100
+ Sed1-Cr(VI)
Sed2-Cr(VI)
SedS-Cr(VI)
0Sed1-Cr(lll)
n Sed2-Cr(lll)
o Sed3-Cr(lll)
Figure D-4 Growth (length in mm) in chromium-spiked freshwater sediment experiments
(Besser et al., in press) vs. total chromium (a), IWTU (b), AVS (c), SEM -AVS (d),
and SEM -AVS//OC (e). Where IWTU = interstitial water toxic units, AVS = acid
volatile sulfide, SEM = simultaneously extracted metal, and/oc = fraction of organic
carbon. Treatments significantly different from control are indicated with an
asterisk. Vertical lines are drawn at 0.5 IWTU (b), 0.0 SEM-AVS (d), and 130 (SEM-
AVS)//oc(e).
D-13
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
10-,
Young/Female
Ol
0-
1
10-,
a
* ' ° 1
o ^ 5"
0 §
D °
D*
o -
°° ..
1 d
1
1
i
i
.$
D i
1
D*i
i
10 100 1000 10000 100000 -100 -80 -60 -40 -20 0 20
Total Chromium (ug/g dry wt) SEM-AVS (umoles/g
10-,
Young/Female
Ol
0-
0.
b1 0 "
^B ^ ^B (D
^P (0
0) e
0 * I
D >-
D*
n
>° .
01 0.1 1 10 100 1000 10000 -4000 -3000 -2000 -1000
dry wt)
i e
i
:
i
i
*ni
i
0 1000
IWTU (SEM-AVS)/foc
101
4
0)
CO
E
0)
LL
"5) 5
£= <
=3 <
O
>
0-
(
c
. .
>
> 0
+ Sed1-Cr(VI)
Sed2-Cr(VI)
SedS-Cr(VI)
oSedl-Cr(lll)
n nSed2-Cr(lll)
n* oSed3-Cr(lll)
D 20 40 60 80 100
AVS (umoles/g dry wt)
Figure D-5. Reproduction (young per female) in chromium-spiked freshwater sediment experiments
(Besser et al., in press) vs. total chromium (a), IWTU (b), AVS (c), SEM-AVS (d), and
SEM -AVS//OC (e). Where IWTU = interstitial water toxic units, AVS = acid volatile
sulfide, SEM = simultaneously extracted metal, and/oc = fraction of organic carbon.
Treatments significantly different from control are indicated with an asterisk. Vertical
lines are drawn at 0.5 IWTU (b), 0.0 SEM-AVS (d), and 130 (SEM-AVS)//OC (e).
D-14
-------
Appendix D
Section 4
Derivation of ESB for Chromium
4.1. General Information
Mortality results of the toxicity tests
conducted in both fresh and saltwater, with both
spiked and field sediments, were generally
consistent with the chromium hypothesis. They
indicated that sediments with measurable amounts
of AVS will not have acute lexicologically
significant concentrations of chromium in the
interstitial water, and that the sediments will not
be acutely toxic due to chromium. Therefore, if
measured sediment chemistry is being used as part
of a sediment assessment, the presence of
measurable AVS could be used to rule out
chromium as the cause of observed acute toxicity.
The chromium hypothesis can also serve as a
foundation for a theoretically-derived sediment
ESB for chromium.
The growth and reproduction results of the
chronic tests conducted in freshwater with Cr(VI)
were also consistent with the chromium
hypothesis. The growth and reproduction results
of the chronic tests conducted in freshwater with
Cr(III) were more ambiguous. It is possible that
these effects were observed as a result of the
unrealistic conditions in the Cr(III)-spiked
sediments, but more testing may have to be
performed before the presence of growth and
reproductive effects in sediments with large
amounts of Cr(III) present can be ruled out.
4.2 Limitations of the chromium
hypothesis
For the chromium hypothesis to work, and a
Cr ESB to be useful, Cr(III) must not be toxic in
interstitial water; however, many studies have
reported on the toxic effects of Cr(III). Both the
U.S. and Canada have water quality criteria
(WQC) for Cr(III), although the criteria for Cr(III)
are much higher than those for Cr(VI) (Pawlitz et
al., 1997; U.S.EPA., 1985). Confounding factors
in many of the tests used to develop these criteria,
such as pH values outside of the tolerance range
of test organisms (Dorfman., 1997) or reported
LCSOs orders of magnitude above limits of
solubility for Cr(III) (Calabrese et al., 1973),
make interpretation of the results of these tests
difficult. Nonetheless, some of the tests used to
develop the criteria demonstrate biological effects
of Cr(III) at environmentally reasonable pH values
and within limits of solubility (e.g., Stevens and
Chapman (1984) showed chronic effects of
Cr(III) on salmonid larvae).
Several recent studies have also shown
biological effects due to Cr(III), including DNA
damage and other sublethal effects associated with
exposure to sediments from some of the same
field sites from which Berry et al., (in press)
collected sediments (Breedlovee et al., 2002),
reduced growth of cyanobacteria (Thompson et
al., 2002) and reduction in population growth rate
of polychaetes (Mauri et al., 2002). Lastly, Besser
et al. (2002) report reduced survival of the
amphipod Hyalella azteca after 28 days in water-
only exposures to Cr(III) at concentrations below
solubility limits at a range of environmentally
reasonable pHs. All of these reported effects are
either sublethal or occurred after 10 days, so none
of them would be expected to occur in the acute
assays of Berry et al., (impress).
Another important fact to consider when
deriving an ESB for chromium is that benthic
animals, particularly tube and burrow dwellers
such as A. abdita, modify the sediment around
them by irrigation of their tubes and burrows,
leading to changes in the sediment environment,
and particularly in the redox condition of
sediments near the animal (Wang et al., 2001).
Thus, bulk sediment might have measurable AVS,
while Cr(VI) might be present in oxic
microenvironments within the sediment. The
geochemistry of chromium argues against this,
D-15
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
however, because direct oxidation of Cr(III) to
Cr(VI) by dissolved oxygen is slow (DeLaune et
al., 1998), significant oxidation of Cr(III) to
Cr(VI) occurs only in soils and sediments with
elevated concentrations of manganese oxides and
low organic content, conditions under which AVS
would not be formed (Masscheleyn et al., 1992,
Kozuh et al., 2000), and Cr(III) is very slow to
react even in environments where it is thermo-
dynamically unstable (Barnhart, 1997).
4.3 Incorporation into Multiple Metals
Benchmark
In sediments where chromium is the only
major metal of concern the AVS and interstitial
water ESBs may be used as listed below.
However, in many cases chromium will be present
along with other metals, and will need to be
evaluated along with them. One of the major
objectives of this appendix is to expand the utility
of the AVS methodology used with cadmium,
copper, lead, nickel, silver, and zinc to include
chromium. See sections 4 and 6 of the metals
ESB for more detail on the benchmarks for
cadmium, copper, lead, nickel, silver and zinc, and
exact definitions of the AVS and interstitial water
benchmarks.
Any sediment in which SEM -AVS < 0.0
should have low risk of adverse biological effects
due to chromium, because measurable AVS must
be present for this to be true (Figures D-2d, D-3d,
D-4d, and D-5d). It should also have low risk of
adverse biological effects due to cadmium,
copper, lead, nickel, and zinc. Any sediment in
which SEM - AVS > 0.0, but AVS > 0.0 should
have low risk of adverse biological effects due to
chromium or silver, but may have adverse
biological effects due to cadmium, copper, lead,
nickel or zinc. Sediments with SEM - AVS > 0.0
in which AVS does not exceed 0.0 may have
adverse biological effects due to cadmium,
copper, lead, nickel, silver, zinc and chromium.
The use of the (SEM - AVS)//OC benchmark in
sediments contaminated with chromium is
complicated slightly by the fact that a sediment
with a slight excess of SEM - AVS may have AVS
= 0.0, and thus be at risk to adverse biological
effects of chromium, while at the same time not
posing a risk due to cadmium, copper, lead,
nickel, or zinc because of organic carbon binding.
However, with an understanding of the chemistry
of AVS, organic carbon, and metals it is possible
to use the benchmark in sediment containing a
mixture of metals including chromium. The
interpretation of the benchmark with respect to
cadmium, copper, lead, nickel, silver, zinc, and
chromium is driven by four assumptions:
1) Any sediment with AVS > 0.0 will not cause
adverse biological effects due to chromium or
silver.
2) Any sediment in which (SEM -AVS)//OC < 130
|imols/goc should pose low risk of adverse
biological effects due to cadmium, copper, lead,
nickel and zinc.
3) Any sediment in which 130 |imols/goc < (SEM
- AVS)//OC < 3,000 |imols/goc may have
adverse biological effects due to cadmium,
copper, lead, nickel or zinc.
4) In any sediment in which (SEM - AVS)//OC >
3,000 |imols/goc adverse biological effects due
to cadmium, copper, lead, nickel or zinc may be
expected.
D-16
-------
Appendix D
Section 5
Sampling and Analytical Chemistry
5.1 General Information
All of the issues regarding proper sampling
and analytical methods described for other metals
(e.g., sampling biologically active zone, seasonal
variation) are equally pertinent when chromium is
an analyte of interest. Therefore, the guidance
given on these topics earlier in this document are
similarly appropriate. However, the differing
physical and chemical characteristics of
chromium in various oxidation states create
additional concerns, both in sampling and
analysis. For example, samples need to be
collected and stored to preserve and minimize
disturbance of existing redox conditions and
thereby retain the distribution of solid and
aqueous phase Cr(III) and Cr(VI) as much as
possible. These chromium specific concerns are
discussed below.
5.2 Sampling Sediment and IW
Normal procedures used to collect and
preserve sediments for analysis of AVS and SEM
are sufficient to preserve chromium speciation as
well. Potential artifacts of sample handling and
storage might include reduction of Cr(VI) or
oxidation of Cr(III). The former should be
addressed by keeping a sample cold, or even
frozen, to inhibit in situ microbial reduction,
while isolating a sediment sample from air, as
well as chilling and freezing the sample, should
eliminate the likelihood of oxidation of Cr(III) to
Cr(VI) in sediments. Preservation of redox
conditions in water samples, however; is
significantly more problematic, and requires
greater diligence.
As with sediment sampling, the guidance
provided earlier in this document regarding
collection of interstitial water is appropriate for
samples in which chromium is an analyte of
interest. Because of the potential of reduction of
Cr (VI) to insoluble Cr(III) species within the
sampler during the course of the experiment (Berry
et al., impress), interstitial water samples should be
filtered immediately after removal from the
sampler, whether collected using centrifugation or
in situ diffusion samplers (Berry et al., 1996). If
centrifugation is used to isolate interstitial water,
temperature should be kept low and the overlying
atmosphere rendered inert to prevent possible
oxidation of Cr(III) to Cr(VI) by Fe/Mn-rich films
at the air-water interface (Masscheleyn et al.,
1992).
Techniques that separate Cr species of different
redox states should be applied to water samples as
soon after collection as possible; if such separation
cannot be obtained rapidly, samples should be
frozen to preserve chemical speciation until such
time as separation is practical. For example, Cr(III)
and Cr(VI) species in overlying and interstitial
water samples can be separated using a modified
Fe(OH)3 coprecipitation technique (Berry et al., in
press, Cranston and Murray, 1978) within hours of
collection. Treatment with ion exchange resins to
isolate Cr species has also been used (Besser et al.,
in press).
5.3 Chemical Analyses
5.3.1 Sediment Analysis
Techniques recommended for analysis of AVS
in sediment samples are appropriate when
chromium is a concern, with only slight
modification of techniques for analyzing
simultaneously extracted metals (SEM). As with
water samples, if Cr(VI) is to be measured in the
SEM solution, separation of redox species should
be conducted as soon after filtration of the extract
as possible. If Cr(VI) is expected to be a problem,
it should be determined in an aliquot of the SEM
extracts by using a modified Fe(OH)3 coprecipita-
tion technique to remove Cr(III) (Wang et al., 1997,
Berry et al., impress) and analyzing Cr in the
D-17
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
supernatant by atomic spectrochemical means
(e.g., inductively coupled plasma atomic emission
spectrometry (ICP-AES) and graphite furnace
atomic absorption spectrophotometry (GFAAS)).
5.3.2 Water analysis
Interstitial waters and overlying seawater
from sediment tests may be analyzed for total and
dissolved chromium and Cr(VI); however,
analysis of chromium in saline waters at low
concentrations can be problematic, so separation
of redox species should only be conducted when
evidence suggests the presence of Cr(VI) (i.e.,
AVS concentrations are near or below detection
limits). Aliquots of water samples to be analyzed
for dissolved metals should be filtered through a
0.4-micron polycarbonate membrane and then
acidified with concentrated nitric acid (1% v/v),
with Cr(VI) determined in subsamples of the
dissolved sample using a modified Fe(OH)3
coprecipitation technique (Cranston and Murray,
1978) as appropriate. Analysis of the various
fractions may be conducted by GFAAS.
D-18
-------
Appendix D
Section 6
Benchmark Sediment Values:
Application and Interpretation
6.1 AVS Benchmark
The AVS benchmark for chromium is
different from the SEM-AVS used for cadmium,
copper, lead, nickel, silver and zinc because
chromium does not form an insoluble sulfide.
However, an AVS measurement is still useful in
predicting the toxicity of chromium in a sediment,
because sediments which have measurable AVS
should be reducing in nature; therefore, most
chromium should be present in the form of Cr(III),
and the risk from acute toxicity due to chromium
exposure should be low.
6.2 Interstitial Water Benchmark
The interstitial water benchmark is similar to
that for cadmium, copper, lead, nickel, silver and
zinc. If the interstitial water concentration of
chromium does not exceed the chronic WQC FCV
for Cr(VI) (10 |ig/L in freshwater and 50 |ig/L in
saltwater (U.S.EPA., 1995)), the risk from
chromium exposure should be low. The Cr(VI)
WQC is used because most of the dissolved
chromium in sediments should be in the form of
Cr(VI), the freshwater benchmark for Cr(VI) is
lower than that for Cr(III), and there is no chronic
benchmark for Cr(III) in saltwater.
6.3 Incorporation into Multiple Metals
Benchmark
The metals benchmark with respect to
cadmium, copper, lead, nickel, silver, zinc and
chromium is driven by four assumptions:
1) Any sediment with AVS > 0.0 will not cause
adverse biological effects due to chromium or
silver.
2) Any sediment in which (SEM -AVS)//OC < 130
|imols/goc should pose low risk of adverse
biological effects due to cadmium, copper, lead,
nickel and zinc.
3) Any sediment in which 130 |imols/goc < (SEM
- AVS)//OC < 3,000 |imols/goc may have
adverse biological effects due to cadmium,
copper, lead, nickel or zinc.
4) In any sediment in which (SEM - AVS)//QC >
3,000 |imols/goc adverse biological effects due
to cadmium, copper, lead, nickel or zinc may be
expected.
These four assumptions should prove useful in
the application of the chromium ESB in sediment
assessments. However, the relationship (SEM -
AVS)//QC should be used with caution (with
regard to chromium toxicity) in sediments with
little or no AVS. This is because a sediment with
no appreciable AVS or SEM and substantial
chromium might be toxic due to chromium, even
though no toxicity due to other metals would be
expected. Other potential limitations to the use
of the chromium ESB are outlined in Section 4 of
this appendix.
Use of an AVS based benchmark for assessing
and predicting mortality in sediments due to
chromium was successful in both freshwater and
saltwater. Assessing and predicting sublethal
toxicity in freshwater sediments was hindered by
the observation of significant growth and
reproductive effects in treatments where such
effects were not expected. Causes of these effects
remain ambiguous and may reflect sublethal
chromium toxicity or experimental artifacts.
Further study is needed to resolve these questions.
Consequently, consistent with the
recommendations of EPA's Science Advisory
Board, publication of this document does not
D-19
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
imply the use of ESBs as stand-alone, pass-fail
criteria for all applications; rather, exceedances of
ESBs could trigger collection of additional
assessment data.
Arguably, the most important additional data
needed for assessing contaminated sediments
along with ESBs are the results of toxicity tests.
Sediment toxicity tests provide an important
complement to ESBs in interpreting overall risk
from contaminated sediments. Toxicity tests have
different strengths and weaknesses compared to
chemical-specific guidelines, and the most
powerful inferences can be drawn when both are
used together (see U.S. EPA 2003a,b for further
discussion of using toxicity testing with ESBs to
assess contaminated sediments).
The ESB approaches are intended to protect
benthic organisms from direct toxicity associated
with exposure to metal-contaminated sediments.
They are not designed to protect aquatic systems
from metals release associated, for example, with
sediment suspension, or the transport of metals
into the food web from either sediment ingestion
or ingestion of contaminated benthos.
Furthermore, the ESBs do not consider the
antagonistic, additive or synergistic effects of
other sediment contaminants in combination with
metal mixtures or the potential for
bioaccumulation and trophic transfer of metal
mixtures to aquatic life, wildlife or humans.
D-20
-------
Appendix D
Section 7
References
Barnhart J. 1997. Chromium chemistry and
implications for environmental fate and toxicity.
JSoilContam 6:561-568.
Berry WJ, Hansen DJ, Mahony JD, Robson DL,
Di Toro DM, Shipley BP, Rogers B, Corbin JM,
Boothman WS. 1996. Predicting the toxicity of
metal-spiked laboratory sediments using acid-
volatile sulfide and interstitial water
normalizations. Environ Toxicol Chem 15:2067-
2079.
Berry, WJ, Boothman WS, Serbst JR, and
Edwards PA. In press. Predicting the toxicity of
chromium in sediments. Environ Toxicol Chem.
Bervoets L, Solis D, Romero AM, Van Damme
PA, Ollevier F. 1998. Trace metal levels in
chironomid larvae and sediments from a Bolivian
river: impact of mining activities. Ecotoxicol
Environ Safety 41:275-283.
Besser JM, Brumbaugh WG, Hardesty DK, Ivey
CD, Kunz JL, Kemble NE, Ingersoll CG. 2002.
Toxicity of chromium(III) and chromium(VI) to
the amphipod, Hyalella azteca, in water and
sediment. Interagency Agreement #
DW14780901. U.S. Geological Survey.
Washington, D.C., USA.
Besser JM, Brumbaugh WG, Kemble NE, May
TW, and Ingersoll CG. In press. Effects of
sediment characteristics on the toxicity of
chromium (III) and chromium (VI) to the
amphipod, Hyalella azteca. Environ Sci Tech.
Breedlove D, Young, and Associates. 2002.
Baseline 4.0 ecological risk assessment report.
Macalloy Corporation Site and Shipyard Creek.
Breedlove, Dennis, Young, and Associates.
Franklin, TN, USA.
Calabrese A, Collier RS, Nelson DA, Maclnnes
JR. 1973. The toxicity of heavy metals to embryos
of the american oyster Crassostrea virginica. Mar
5/0/18:162-166.
Catsiki VA, Katsilieri C, Gialamas V 1994.
Chromium distribution in benthic species from a
gulf receiving tannery wastes (Gulf of Geras -
Lesbos island, Greece). Sci Total Environ
145:173-185.
Cranston RE, Murray JW. 1978. The
determination of chromium species in natural
waters. Anal ChimActa 99:275-282.
DeLaune RD, William H. Patrick J, Guo T 1998.
The redox-ph chemistry of chromium in water and
sediment. In Allen HE, Garrison AW, Luther III
GW eds, Metals in Surface Waters. Ann Arbor
Press, Chelsea, MI, pp 241-255.
Dorfman D. 1977. Tolerances ofFundulus
heteroclitus to different metals in saltwaters. Bull
New Jersey Acad Sci 22:21-23.
Elovitz, MS, Fish W. 1995. Redox reactions of
Cr(VI) and substituted phenols: Products and
mechanism. Environ Sci Technol 29:1933-1943.
Fan W, Wang W-X. 2001. Sediment geochemical
controls on Cd, Cr, and Zn assimilation by the
clam Ruditapes philippinarnum. Environ Toxicol
Chem 20:2309-2317.
Griscom SB, Fisher NS, Luoma SN. 2000.
Geochemical influences on assimilation of
sediment-bound metals in clams and mussels.
Environ Sci Technol 34:91-99.
Hassan SM, Garrison AW. 1996. Distribution of
chromium species between soil and porewater.
Chem Speciation Bioavail 8:85-103.
HydroQual. 1994. Remedial design pre-design
studies, report 94-4, Cannelton Industries site,
Sault Ste. Marie, Michigan, October 1994.
Icopini GA, Long DT 2002. Speciation of
aqueous chromium by use of solid-phase
extractions in the field. Environ Sci Technol
36:2994 -2999.
D-21
-------
Equilibrium Partitioning Sediment Benchmarks (ESBs): Metal Mixtures
Kozuh N, Stupar J, Gorenc B. 2000. Reduction
and oxidation processes of chromium in soils.
Environ Sci Technol 34:112 -119.
Leslie HA, Pavluk TI, bij deVaate A, Kraak MRS.
1999. Triad assessment of the impact of chromium
contamination on benthic macroinvertebrates in
the Chusovaya River (Urals, Russia). Arch
Environ Contam Toxicol 37:182-189.
Masscheleyn PH, Pardue JH, Delaune RD,
William H. Patrick J. 1992. Chromium redox
chemistry in a lower Mississippi valley
bottomland hardwood wetland. Environ Sci
Technol 26:1217-1226.
Mattuck R, Nikolaidis NP 1996. Chromium
mobility in freshwater wetlands. J Contam Hydrol
23:213-232.
Mauri M, Simonini R, Baraldi E. 2002.
Demographic responses of the polychaete
Dinophilus gyrociliatus to chromium exposure.
Environ Toxicol Chem 21:1903-1907.
Mearns AJ, Swartz RC, Cummins JM, Dinnel PA,
Plesha P, Chapman PM. 1986. Inter-laboratory
comparison of a sediment toxicity test using the
marine amphipod, Rhepoxynius abronius. Mar
Environ Res 19:13-37.
Nakayama E, Tokoro H, Kuwamoto T, Fujinaga T.
1981. Dissolved state of chromium in seawater.
Nature 290:768-770.
Pawlitz AV, Kent, RA, Schnieder UA, Jefferson C.
1977. Canadian water quality guidelines for
chromium. Environ Toxicol Water Qual 12:185-
193.
PoleoABS. 1995. Aluminum polymerization: a
mechanism of acute toxicity of aqueous aluminum
to fish. Aqua Toxicol 31:347-356.
Richard FC, Bourg ACM. 1991. Aqueous
geochemistry of chromium: A Review. Wat. Res.
25:807-816.
Stevens DG, Chapman GA. 1984. Toxicity of
trivalent chromium to early life stages of
steelhead trout. Environ Toxicol Chem 3:125-133.
Thompson SL, Manning FCR, McColl SM. 2002.
Comparison of the toxicity of chromium III and
chromium VI to cyanobacteria. Bull Environ
Contam Toxicol 69:286-293.
U.S. Environmental Protection Agency. 1985.
Ambient water quality criteria for chromium -
1984. EPA 440/5-84-029. Office of Water.
Washington, DC.
U.S. Environmental Protection Agency. 1986.
Ambient water quality criteria for chromium -
1986. EPA 440/5-86-001. Office of Water.
Washington, DC.
U.S. Environmental Protection Agency. 1995.
Federal Register. 60(86):22231-22232.
U.S. Environmental Protection Agency. 2002.
National recommended water quality criteria:
2002. EPA 822/R-02-047. Office of Water.
Washington, DC.
U.S. Environmental Protection Agency. 2003a.
Procedures for the derivation of equilibrium
partitioning sediment benchmarks (ESBs) for the
protection of benthic organisms: Endrin. EPA-
600-R-02-009. Office of Research and
Development, Washington, DC.
U.S. Environmental Protection Agency. 2003b.
Procedures for the derivation of equilibrium
partitioning sediment benchmarks (ESBs) for the
protection of benthic organisms: Dieldrin. EPA-
600-R-02-010. Office of Research and
Development, Washington, DC.
Wang F, Tessier A, Hare L. 2001. Oxygen
measurements in the burrows of freshwater
insects. Freshwat Biol. 46:317-327.
Wang W-X, Griscom SB, Fisher NS. 1997.
Bioavailability of Cr(III) and Cr(VI) to marine
mussels from solute and particulate pathways.
Environ Sci Technol 31:603-611.
Wittbrodt PR, Palmer CD. 1995. Reduction of
Cr(VI) in the presence of excess soil fulvic acid.
Environ Sci Technol 29:255-263.
D-22
------- |