822D94002
Volume I
. *•. . '' *
BRIEFING REPORT
to the
EPA SCIENCE ADVISORY BOARD MAR 2 11995
on the
EQUILIBRIUM PARTITIONING APPROACH
,*± TO PREDICTING METAt
IN SEDIMENTS AND TH&l>ERIVATlOfsi OF
SEDIMENT^QUALltY CRITlHIA FOR METALS
December 1994
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF WATER AND'
OFFICE OF RESEARCH AND DEVELOPMENT
Printed on Recycled Paper
-------
'••-•''<•" ' . c.,';- -••>,-• ',
-------
DISCLAIMER
This document is a compilation of data and analyses from scientific investigations
into the bioavailability of metals in sediments to benthic organisms with the intent of
proposing an approach to assessing metals contamination of sediments for the protection
of benthic organisms.
This document does not establish or affect legal rights or obligations. It does not
establish a binding norm and is not finally determinative of the issues addressed. Agency
decisions in any particular case will be made applying the law and regulations on the basis
of specific facts when permits are issued or regulations promulgated.
The mention of trade names or specific products does not constitute endorsement.
-------
-------
CADMIUM TITRATION OF IRON SULFIDE
0.0 0.5 1.O 1.K 2.0
CADMIUM ADDED (umol Cd/umol FeS)
Figure 4-1. Cadmium titrations of amorphous FeS. Abscissa Is cadmium added normalized
by FeS initially present. Ordinate is total dissolved cadmium. The lines connecting the
data points are an aid to visualizing the data.
-------
Results of Fes + Cd** = CdS + Fe**
Concentration of Fe**
Analysis of Filtrate
600
475
350
225
100-
-25
M M
0 02 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Cd'VFeS Molar Ratio
0*
0
Concentration of Cd**
Analysis of Filtrate
0.2 0.4 0.6 0.8 1 12 1.4 1.6 1.8
Cd'VFeS Molar Ratio
Theory x 1-14 * 1-15
a 1-18 A. M9.w/Buffer
Figure 4-2. Concentrations of Fe2+ and Cd2+ in supernatent from titration of FeS by
ud .
-------
CONTENTS
Chanter Pace
1 INTRODUCTION 1-1
Overview 1-1
Legal Basis 1-1
Definition of Sediment Criteria (aquatic life) 1-2
Methodology Selection 1-2
Application of Sediment Criteria 1-3
Document Outline , 1-6
2 EQUILIBRIUM PARTITIONING 2-1
Bioavailability 2-1
Pore Water Normalization 2-5
Sorption of Nonionic Organic Chemicals , 2-9
Effects Concentration 2-13
3 METAL TOXICITY IN WATER AND SEDIMENT EXPOSURES 3-1
Toxicity Correlates to Metal Activity 3-2
Interstitial Water And Metal Toxicity 3-4
4 METAL PARTITIONING 4-1
Metal Sorption Phases 4-1
Titration Experiments 4-2
Amorphous FeS 4-3
Sediments 4-4
Correlation to Sediment AVS 4-5
Solubility Relationships and Displacement Reactions 4-7
Application to Mixtures of Metals 4-10
5 LABORATORY SPIKING EXPERIMENTS 5-1
Results 5-2
Saltwater Amphipod Tests 5-2
Water-Only Tests 5-3
Spiked Sediment tests 5-4
Sediment Chemistry - 5-4
Day 0 vs Day 10 Chemistry Values - 5-4
Interstitial Water Metal Versus SEM/AVS 5-4
Sediment Toxicity 5-12
Discussion 5-20
6 FIELD COLLECTED SAMPLES 6-1
Saltwater Field Sites , 6-2
Description of Field Sites and Toxicity Test Results 6-2
Freshwater Field Sites 6-10
-------
Description of Field Sites and Toxicity Test Results 6-10
Discussion 6-17
Saltwater Field Sites 6-17
Saltwater and Freshwater Field Sites 6-23
Field Sites and Spiked Sediments 6-30
Summary 6-32
7 COLONIZATION EXPERIMENTS 7-1
Results 7-2
Exposure 7-2
Effects 7-8
Discussion 7-14
8 BIOACCUMULATION OF METALS 8-1
Laboratory Spiking Experiments-Freshwater 8-3
Laboratory Spiking-Marine 8-5
Field Sediments-Freshwater 8-9
Field Sediments-Marine 8-13
Field Spiking-Freshwater 8-14
Summary and Conclusions 8-17
9 AVS AND OTHER BINDING PHASES 9-1
Vertical and Seasonal AVS Distributions 9-1
The Correlation of AVS to Sediment Organic Carbon 9-9
Oxidation of Metal Sulfides 9-11
FeS(s) and CdS(s) Oxidation Kinetics 912
Sediment Metals Oxidation Model 9-16
Conclusions 9-28
Organic Carbon Binding 9-28
Analysis Framework 9-29
Sorption Isotherm Results 9-31
Least Sorptive Phase 9-44
Pore Water and AVS/SEM Sampling 9-47
Pore Water Sampling 9-47
AVS/SEM Sampling 9-50
10 CONSIDERATIONS FOR ASSESSING METAL BIOAVAILABILITY IN
SEDIMENTS 10-1
Interstitial Water 10-1
Acid Volatile Sulfide and SEM 10-2
11 PROPOSED SEDIMENT QUALITY CRITERIA 11-1
Introduction 11-1
Single Metal Sediment Quality Criteria 11-3
AVS Criteria 11-3
Interstitial Water Criteria 11-4
-------
AVS and Organic Carbon Criteria 11-6
AVS and Minimum Partitioning Criteria 11-7
Multiple Metals Criteria 11-8
AVS Criteria 11-8
Interstitial Water Criteria 11-9
AVS and Organic Carbon Criteria 11-10
AVS and Minimum Partition Coefficient Criteria 11-13
Criteria Summary 11-13
Sediment Quality Criteria Uncertainty 11-14
Research Recommendations 11-17
Conclusions 11-18
in
-------
CHAPTER 1
INTRODUCTION
Overview
This briefing document for the Science Advisory Board (SAB) describes the
methodology that the EPA is proposing to use to establish a national sediment quality
criteria for metals. It is based on the Equilibrium Partitioning (EqP) method. EqP was
originally developed and adopted to derive national sediment criteria for non-ionic organic
chemicals. This document presents the methodology and supporting information for
deriving sediment criteria for cadmium, copper, lead, nickel, and zinc.
Legal Basis
Responding to environmental problems with corrective action frequently requires
proving that something negative has occurred. "Innocent until proven guilty," the standard
of the American judicial system, works well for many legal activities; but applying this
logic to environmental protection efforts (an activity is environmentally acceptable until it
can be proven unacceptable) presents a unique set of strengths and limitations.
Regulatory agencies frequently are called upon to prove environmental or human health
degradation has or could occur prior to taking any corrective or preventive action.
Scientifically sound and legally defensible measures that demonstrate potential or actual
impacts are imperative. Fundamental to any effort to ensure environmental protection is
to define the party upon whom the burden of proof lies. Chemical specific criteria are one
tool developed by regulatory and non-regulatory agencies that is frequently used to meet
the burden of proof requirements. Criteria define when a release of a substance into the
environment is acceptable and when the release is causing or has potential to cause
adverse impacts to aquatic life, wildlife, or human health. EPA initiated efforts to develop
national sediment quality criteria under the authority of the Clean Water Act to protect the
chemical, physical, and biological integrity of the country's water resources, aquatic life,
and water dependent resources. The sediment criteria are intended to be used to prevent
-------
1-2
clean sediments from becoming contaminated and to assist in making regulatory and
remediation decisions on sediments that are already contaminated.
Definition of Sediment Criteria (aquatic life)
Sediment Quality Criteria are the Ij.S. Environmental Protection Agency's best
recommendation of the concentration of a substance jn sediment that will not
unacceotablv affect benthic organisms.
Methodology Selection
A detailed methodology has been developed to derive sediment quality criteria for
metals. This document presents the supporting logic and specifies the numerical
procedures to be used to calculate the criteria values.
The use of total sediment metal concentration as a measure of bioavailable - or even
potentially bioavailable - concentration is not supported by the available data.
Experimental results indicate that different sediments can differ in toxicity significantly for
the same total metal concentration. Without accounting for this difference one cannot set
a national sediment quality criteria that depends only on the total sediment metal
concentration. Therefore, the variation in the bioavailability of metals in various sediments
must be explicitly considered in the establishment of defensible sediment quality criteria.
This is a significant obstacle; since without some quantitative estimate of the bioavailable
metal concentration in a sediment, it is impossible to predict a sediment's toxicity based
on chemical measurements. This is true regardless of the methodology used to assess
biological impact - be it laboratory toxicity experiments or field data sets comprising
benthic biological and chemical sampling.
The EqP approach was selected to establish sediment quality criteria because it
incorporates the most useful technical aspects of a variety of approaches. Specific
reasons for selecting the approach are as follows:
-------
1-3
(1) It was likely that the EqP approach would yield sediment criteria that were
predictive of biological effects in the field and would be defensible when
used in a regulatory context. These criteria would directly address the issue
of bioavailability and for the most part are based on the extensive biological
effects data base used to establish national water quality criteria.
(2) Sediment criteria could be readily incorporated into existing regulatory
operations since a unique numerical sediment specific c'riteria can be
established for a chemical and compared to field measurements to assess
the likelihood of significant adverse effects.
(3) Sediment criteria could provide a simple and cost effective means of
assessing sediment measurements to identify areas of concern and could
quickly provide regulators with information on potential incremental impacts
on benthic organisms as a function of the extent of criteria excedences.
(4) The method takes advantage of the large amount of data and expertise that
went into the development of the National Water Quality Criteria.
(5) The methodology could be used as a regulatory predictive tool to link
sources to sediment sinks to ensure uncontaminated sites would be
protected from attaining unacceptable levels of contamination.
Application of Sediment Criteria
Persistent contaminants discharged into the surface waters of the U.S. end up
primarily in the water column and in sediments. Technology based controls, whole
effluent toxicity tests, and chemical specific water quality criteria provide the basis for
controlling water column contamination. For sediment contamination a similar approach
is being adopted. Technology based controls, sediment biological tests, and sediment
criteria are intended to control, prevent, and manage sediment contamination.
-------
1-4
Over the past two years EPA has been preparing a draft Agency wide sediment
management strategy to coordinate and focus EPA's resources on contaminated sediment
problems. An outline of this draft strategy has been released to the public for review and
a final draft was announced in the Federal Register this past summer as available for
formal public comment. The draft strategy is designed around three major principals:
1) In-place sediment should be protected from contamination to ensure that the
beneficial uses of the nation's surface waters are maintained for future
generations;
2) Protection of in-place sediment should be achieved through pollution
prevention and source controls;
3) Natural recovery is often the preferred remedial technique. In-place sediment
remediation will be limited to high risk sites where natural recovery will not
occur in an acceptable time period and where the cleanup process will not
cause greater problems than leaving the site alone.
The draft strategy has six components: assessment, prevention, remediation,
dredged material management, research, and outreach. Sediment quality criteria are
integrated into this strategy. However, their specific role will be outlined in a "User's
Guide" currently under development. The Guide will be based on the comments received
on the criteria and strategy and will incorporate the management strategies of a variety of
program offices on a program specific basis.
Even though the specific applications of sediment criteria are under development,
in general the primary use of sediment criteria will be to prevent sediment contamination
and assess risks associated with contaminant levels in sediments. Various EPA programs
concerned with contaminated sediment have different regulatory mandates and thus, have
different needs and areas for potential application of sediment criteria. Because each
-------
1-5
regulatory need is program specific, sediment criteria have to be implemented in a variety
of ways to meet the needs of diverse offices and programs.
A likely application of EqP sediment criteria would be in a tiered approach. In such
an application, when contaminants in sediments exceed the sediment quality criteria, the
sediments would be considered to be causing unacceptable impacts. Further testing may
or may not be required depending on site specific conditions and the degree to which a
criteria has been violated. (At locations where contamination significantly exceeds a
criterion, no additional testing would be required. Where sediment contaminant levels are
close to a criteria, additional testing may be necessary). Contaminants in a sediment at
concentrations less than the sediment criteria would not be of concern. However, the
sediment could not necessarily be considered acceptable for benthic organisms because
they may contain other contaminants above safe levels for which no sediment criteria
exist. In addition the synergistic, antagonistic, or additive effects of several contaminants
in the sediments may be of concern. Sediment criteria can provide a basis for determining
whether contaminants are accumulating in sediments to the extent that an unacceptable
contaminant level is being approached or has been exceeded. By monitoring sediment
contaminants in the vicinity of a discharge, contaminant levels can be compared to
sediment criteria to assess the likelihood of impact.
Sediment criteria will be particularly valuable in site monitoring applications where
sediment contaminant concentrations are gradually approaching a criteria over time.
Comparison of field measurements to sediment criteria will be a reliable method for
providing early warning of a potential problem. Such an early warning would provide an
opportunity to take corrective action before adverse impacts occur.
The safe removal and treatment or disposal of contaminated sediments can be
difficult, expensive, and in some case not the environmentally preferred option (e.g.,
resuspension of contaminated sediments may cause greater harm than leaving them where
they are). Leaving sediments in place and allowing source controls and natural processes
to reduce or remove the contamination has proven to be an effective option in some cases.
-------
1-6
In some situations, the high cost of clean-up activities has eliminated clean-up as an
option. In other situations the spatial extent of the site and site-specific conditions may
warrant remediation beyond identified clean-up levels, an additional margin of safety can
be provided at little or no extra cost. For these reasons it is not anticipated that
mandatory clean-up to a nationally designated level is appropriate in most situations.
Document Outline
This document contains ten additional chapters in which are presented the
experimental data and the methods to be employed in deriving sediment quality criteria.
Chapter 2 reviews the Equilibrium Partitioning methodology which is used to
understand the relationships between sediment and interstitial water concentrations and
observed biological effects.
Chapter 3 examines metal toxicity and bioavailability first in water only exposures
and then as a function of interstitial water concentrations. This chapter presents the data
that suggest that metal toxicity is best correlated to metal activity in water only
exposures. It also presents the correlations of organism response to interstitial water
concentrations in sediment exposures.
Chapter 4 reviews the state of the art for predicting metal partitioning in sediments.
This is important because it connects the solid phase sediment chemistry to the resulting
interstitial water concentrations - both of which set the metal activity of the sediment -
interstitial water system. The importance of the cold acid extractablesulfide in sediments,
which for historical reasons is called acid volatile sulfide (AVS), is demonstrated. An
equilibrium model is analyzed to establish the relationship between metal activity in the
sediment - interstitial water system, the extractable solid phase metal concentration -
which is called the SEM for reasons discussed below - and the ratio of the metal to iron
sulfide solubility constants.
-------
1-7
Chapter 5 presents the data from laboratory sediment spiking experiments.
Uncontaminated sediments are spiked with varying metal concentrations, and the resulting
toxicity is measured using sediment-dwelling animals. The results are analyzed using the
SEM and AVS method and the interstitial water concentrations.
Chapter 6 presents a similar set of results but these sediments are from field sites
with metal contamination. These experiments address the question: can the SEM, AVS,
and interstitial water methods explain the toxicity observed in field collected samples.
Chapter 7 presents the results of colonization experiments. These are designed to
mimic the field setting as closely as possible while still retaining a measure of laboratory
control over the exposures. A series of sediments are spiked with various concentrations
of metals. These are exposed either to raw flowing seawater or are placed on lake
bottoms for a period of time. The biological response that is monitored is the quality and
quantity of the benthic animals that colonize the sediments.
Chapter 8 examines the extent of metal bioaccumulation from sediment exposures.
The biological response in this case is different than that of the previous experiments. The
results are analyzed using the same methods based on SEM, AVS, and interstitial water
concentrations.
Chapter 9 presents information on a number of topics that influence the variation
of AVS, temporally, spatially, and with respect to organic carbon in sediments. The
oxidation kinetics of iron sulfide and, more importantly, cadmium and zinc sulfide are
examined in order to answer the question: what is the importance of seasonal variations
and how do they affect criteria. The experimental information related to organic carbon
binding and the resulting partition coefficients are presented.
Chapter 10 presents the strengths and limitations of the various methods that have
been discussed in the previous chapters. Recommendations are included for situations
where more study is required.
-------
1-8
Chapter 11 presents a proposed sediment quality criteria for the five metals: Cd,
Cu, Ni, Pb, and Zn. Single metal criteria are not technically supportable since the binding
of each of these metals interact with each other, thus affecting their mutual toxicity. A
discussion of the limitations of these criteria and their proposed utility is also included.
-------
CHAPTER 2
EQUILIBRIUM PARTITIONING
The development and application of the Equilibrium Partitioning (EqP) method for
deriving SQC for nonionic organic chemicals has previously been presented to the EPA
Science Advisory Board and has been published in the Federal Register for public
comment. A full discussion of this work is presented elsewhere [1,2] as well as the
derivation of SQC for five nonionic organic chemicals [3,7]. The development of sediment
quality criteria for metals also utilizes the EqP method. This chapter presents a summary
of the technical basis for establishing sediment quality criteria for nonionic organic
chemicals using EqP. Cadmium data is also presented. The purpose of presenting this
summary for nonionic organic chemicals is to briefly define the EqP method and to
demonstrate its utility in determining the bioavailability of sediment chemicals.
Bioavailability
Establishing a SQC requires a determination of the extent of the bioavailability of
sediment associated chemicals. It has frequently been observed that similar
concentrations of a chemical, in units of mass of chemical per mass of sediment dry
weight (e.g. micrograms chemical per gram sediment), can exhibit a range in toxicity in
different sediments. An example is presented in Figure 2-1. These are the results of two
sediment toxicity tests of cadmium using amphipods of similar sensitivity [8,9]. One
sediment is from Central Long Island Sound and the other is from Yaquina Bay, Oregon.
The LC50 for cadmium for these two sediments differ by approximately two orders of
magnitude. Because the purpose of SQC is to establish chemical concentrations that apply
to sediments of differing types, it is essential that the reasons for this varying
bioavailability be understood and be explicitly included in the criteria. Otherwise the criteria
cannot be presumed to be applicable across sediments of differing properties.
The importance of this issue cannot be overemphasized. For example, if 20 //g/g
of cadmium is the LC50 for an organism in one sediment and 1,500 //g/g is the LC50 in
-------
o
5
X
Dry Weight Normalization
Cadmium - Ampelisca
100
80
60
40
20
0
10 10' 10 10 10 10
Dry Weight Concentration (/tg/g)
Cadmium - Rhepoxynius
10 10' 10C 10° 10 10"
Dry Weight Concentration (uglg)
Figure 2-1. Comparison of percent mortality of Amoelisca abdita and Rhepoxvnius
abronius to concentrations of cadmium in bulk sediment.
-------
2-3
another sediment (Figure 2-1), then unless the cause of this difference can be associated
with some explicit sediment properties it is not possible to decide what would be the LC50
of a third sediment. The results of toxicity tests used to establish the toxicity of chemicals
in sediments would not be generalizable to other sediments. Imagine the situation if the
results of toxicity tests in water depended strongly on the particular water source - for
example, water from Lake Superior versus well water. Until the source of the differences
was understood, it would be fruitless to attempt to establish water quality criteria. It is
for this reason that understanding bioavailability is a principal focus in establishing
sediment quality criteria.
The observations that provided the key insight to the problem of quantifying the
bioavailability of chemicals in sediments were that the concentration-response curve for
the biological effect of concern could be correlated not to the total sediment chemical
concentration (micrograms chemical per gram sediment), but to the interstitial water (i.e.,
pore water) concentration (micrograms chemical per liter pore water). The results of
toxicity tests of kepone using Chironomus tentans in three sediments is shown in Figure
2-2a. The sediments have quite different LCSOs on a sediment dry weight basis: from
approximately 1 ug/g to approximately 35 ug/g. However if the mortality is examined as
a function of the concentration of kepone in the interstitial or pore water of the sediment,
Figure 2-2b, then the mortality-concentration responses are similar for the three sediments.
In addition, the LCBOs on a porewater basis are the same as the LC50 obtained from a
water only exposure toxicity test. Thus the water only LC50 can be used to predict the
toxicity of these sediments using their porewater concentrations.
Organism mortality, growth rate, and bioaccumulation data are used to demonstrate
this correlation, which is a critical part of the logic behind the EqP approach to developing
SQC. For nonionic organic chemicals, it is shown, subsequently, that the concentration
- response curves correlate equally well with the sediment chemical concentration on a
sediment organic carbon basis. Figure 2-2c presents the same mortality data as a function
of the organic carbon normalized sediment kepone concentration (ug kepone/gm sediment
organic carbon). The responses for the two sediments with the higher organic carbon
-------
X
£
"a
**
o
S
Acute Toxicity of Kepone
0.1 1.0 10.0 100.0
Dry Weight Concentration (ug/g)
1 WATER ONLY LC50
100
60
40
20
0
foe (X)
* • 0.09
. * 1.5
• 12
•
.
^ — *.
£"-*"- «r*
• "i •
1
/
/
•— '
*>*
* • • • - • -j • • »-..-.
f^ '-
SI t
'//
L
±
fm
B •
1 10 1,00 1000
Pore Water Concentration (ug/L)
100
3 80
x 60
**
« 40
20
0
o
5
10 100 1000 10000
Organic Carbon Normalized (ug/g oc)
Figure 2-2. Comparison of percent mortality of Chironomus tentans to kepone
concentration in bulk sediment (top), pore water (center) and bulk sediment using organic
carbon normalization (bottom) for three sediments with varying organic carbon
concentrations [9].
-------
2-5
fractions (foc) display the same response. The sandy sediment (with low organic carbon)
has a larger LC50, probably because the foc for this sediment is too small for organic
carbon to be the only sorption phase. The range of foc values for which carbon
normalization is appropriate has been examined in more detail [1].
These observations ( Figures 2-2b and 2-2c) can be rationalized by assuming that
the pore water and sediment carbon are in equilibrium and that the concentrations are
related by a partition coefficient, K^c, as shown in Figure 2-3 (right). The name
"Equilibrium Partitioning" (EqP) describes this assumption of partitioning equilibrium
between sediment carbon and pore water. The rationalization for the equality of
water-only and sediment-exposure effects concentrations on a pore water basis is that the
sediment - pore water equilibrium system (Figure 2-3, right) provides the same exposure
as a water-only exposure (Figure 2-3, left). The reason is that the chemical activity is the
same in each system at equilibrium. These results do not imply that pore water or
sediment organic carbon is the primary route of exposure because all exposure pathways
are at equal chemical activity in an equilibrium experiment and the route of exposure
cannot be determined.
It should be pointed out that the EqP assumptions are only approximately true and,
therefore, the predictions from the model have an inherent uncertainty. A discussion and
quantification of uncertainty is found in the EPA SQC Technical basis document [2].
Pore Water Normalization
A substantial amount of data has been assembled that addresses the relationship
between toxicity and pore water concentration [1,2]. The data presented below examines
the use of the water-only LC50 to predict the pore water LC50. Figure 2-4 presents
mortality data for various chemicals and sediments compared to pore water concentrations
when normalized on a toxic unit basis. Three different sediments are tested for each
chemical as indicated. Predicted pore water toxic units are the ratio of the measured pore
water concentration to the LC50 obtained from an independent water only toxicity test.
-------
£
jp
a.
£ Q>
o co
O)
o
0
0
"•5
o
CO
x
111
I
s
o
CO
a.
s
CO
O"
LU
j£ J2
O) 3
'= y
=i
3 el-
's W
co "5
co .2
•o E
c »
"•g
I*
gi
SB
D. 0)
X -O
O in
o
o
«
2r
c 'x *i
2;l
® E o
*™ ^ *™
CO ~ 0
8 «c
sa a
25«
0> C I—
"^ £
(0
i- O
s e-
»s <>
,2
O
25
CO
Q. t tJ
X 3 5
CO « ^
£ CO £
.2 o o
c .c a
52 5 c
^ tS «
H- »- C
O> *S o
.2 $ |
O ID re
d. ra
co 5 °
i > *->
e* -Si c
•=P
U. Uj CO
-------
o
c
o
0>
o c _ c £ 2
C m C (Q (0 «e
o c ~ fe c . g
a o -5 2 • t 9
o.c c 2, o Q <
* a. w u. < a O
• • • * * A *
O a O O > A
O
o
I
o
to
§N
To
<
a>
o
Q.
i
i
i
1
O
q
d
0
o
o
T-
d
0)
'E
D
O
'x
£
o
"5
d>
o
Q.
•o
o
•o
CD
O
o
o
eo
o
CO
o
CM
q
d
O)T3 i
C QJ QJ
•R 55 co
c
O OT 2
-Q r, O
vj fc—
—^ »- O
.« £ O *3
O O ~
o to
a> c
CD
a>
§• '21 i
m c fl>
1 P8 s
re o ex
*^ iT CD
(0 (A O r-
"co ® -g ^
•p o S "5
S = O .2
c •£ O) o
fll O i- .-
OT « b
».. cp
O "O >
•*• Q)
"2 == «
•" S
£ -a a.
__ CD —^ **
^* ?? +^ C3
0^*0
I £ ° w
« =5 § a>
= a> S to
J2 CO •£ (a
m CO 4-^
>•§ ^-g
S'l'S SP
§2 SicM
i 2 °- c
Sills
^) AUTOJOIN
3 .= Q) O ;,,
.S'"? C Q) "E
u. to s -a *
-------
2-8
The EqP model predicts that the pore water LC50 will equal the water only LC50. Define:
predicted pore water toxic unit = (P°re water concentration) (2.1}
(water only LC50)
A toxic unit of one occurs when the pore water concentration equals the water-only LC50,
at which point it would be predicted that 50 percent mortality would be observed. In
addition, normalization of these data to pore water toxic units allows all chemical-to-
chemical differences to be removed so that all data from these tests with these seven
chemicals can be coplotted. The correlation of observed mortality to predicted pore water
/
toxic units in Figure 2-4 demonstrates (a) the efficacy of using pore water concentrations
to remove sediment to sediment differences and (b) the applicability of the water-only
effects concentration and, by implication, the validity of the EqP model. By contrast, it
has been shown [1,2], that the mortality versus sediment chemical concentration on a dry
weight basis varies dramatically from sediment to sediment.
The equality of the effects concentration on a pore water basis suggests that the
route of exposure is via pore water. However, the equality of the effects concentration
on a sediment organic carbon basis, which will be presented in Figure 2-5, suggests that
the ingestion of sediment organic carbon is the primary route of exposure. Neither of
these references can be supported by the data presented in Figures 2-4 and 2-5
subsequently. It is important to realize that if the sediment and pore water are in
equilibrium, then the effective exposure concentration is the same regardless of exposure
route. Therefore, it is not possible to determine the primary route of exposure from
equilibrated experiments.
Whatever the route of exposure, the correlation of toxicity to pore water suggests
that if it were possible to either measure the pore water chemical concentration, or predict
it from the total sediment chemical concentration and the relevant sediment properties
such as the sediment organic carbon concentration, then that concentration could be used
to quantify the exposure concentration for an organism. Thus, understanding the
-------
2-9
partitioning of chemicals between the solid and the liquid phase in a sediment becomes a
necessary component for establishing SQC for any substance, including metals.
Sorption of Nonionic Organic Chemicals
For nonionic hydrophobia organic chemicals sorbing to natural soils and sediment
particles, a number of empirical models have been suggested [10]. The chemical property
that indexes hydrophobicity is the octanol/water partition coefficient, Kow. The important
particle property is the weight fraction of organic carbon, foc. Another important
environmental variable appears to be the particle concentration itself [11]. A detailed
discussion of the particle concentration effect has been presented [1,2].
A number of explanations have been offered for the particle concentration effect.
However it is not necessary to decide which of these mechanisms is responsible for the
effect if all the possible interpretations yield the same result for sediment/pore water
partitioning. Each suggest that the proper partition coefficient to be used in order to relate
the free dissolved chemical concentration to the sediment concentration is the partition
coefficient to sediment organic carbon, Koc, which is approximately equal to the octanol-
water partition coefficient Kow, that is, KOC«KOW by the following equation [12].
Log10 KOC • 0.00028 + 0.983 Log10 Kow (2-2)
The unifying parameter that permits the development of SQC for nonionic
hydrophobic organic chemicals that are applicable to a broad range of sediment types is
the organic carbon content of the sediments. This can be shown as follows. The
sediment/pore water partition coefficient, Kd, is given by
Kd = foe Koc <2'3)
and the solid phase concentration is given by
-------
2-10
Cs=focKocCd (2-4)
where Cs is the concentration on sediment particles. An important observation can be
made that leads to the idea of organic carbon normalization. Equation 2-3 indicates that
the partition coefficient for any nonionic organic chemical is linear in the organic carbon
fraction, foc. Data have presented [1,2,12] to support the linearity of partitioning above
a value of foc = 0.2 percent. This result and the toxicity experiments [1 ,2] suggest that
for foc > 0.2 percent, organic carbon normalization is valid.
As a consequence of the linear relationship of Cs and foc, the relationship between
sediment concentration, Cs, and free dissolved concentration, Cd, can be expressed as
= KOC Cd (2-5)
Toc
If we define
C = ^*s (2-6)
*-s,oc f — ^ °'
'oc
as the organic carbon normalized sediment concentration (ug chemical/g organic carbon),
then from Equation (2-5):
Cs,oc - Koc Cd (2-7)
Therefore, for a specific chemical with a specific Koc, the organic carbon normalized total
sediment concentration, Cs oc, is proportional to the dissolved free concentration, Cd, for
any sediment with foc > 0.2 percent. This latter qualification is judged to be necessary
because at foc < 0.2 percent the other factors that influence partitioning (e.g., particle size
and sorption to nonorganic mineral fractions) become relatively more important [8J. Using
the proportional relationship given by Equation 2-7, the concentration of free dissolved
chemical can be predicted from the normalized sediment concentration and K^. The free
concentration is of concern as it is the form that is bioavailable.
-------
2-11
As discussed above, hydrophobia chemicals also tend to partition to colloidal-sized
organic carbon particles that are commonly referred to as dissolved organic carbon, or
DOC. Although DOC affects the apparent pore water concentrations of highly hydrophobic
chemicals, the DOC-bound fraction of the chemical appears not to be bioavailable [1].
Therefore, we expect that toxicity in sediment can be predicted from the water-only
effects concentration and the Koc of the chemical. The utility of these ideas can be tested
with the same mortality data as these in Figure 2-4 but restricted to nonionic organic
chemicals for which organic carbon normalization applies. The concept of sediment toxic
units is useful in this regard. These are computed as the ratio of the organic
carbon-normalized sediment concentrations, Cs/foc, and the predicted sediment LC50 using
Koc and the water-only LC50. That is:
C /f
predicted sediment toxic unit = ——; 5_2£— (2-8)
Koc (water only LC50)
Figure 2-5 presents the percent mortality versus predicted sediment toxic units.
The correlation is similar to that obtained using the pore water concentrations in Figure 2-
4. The predicted sediment toxic units for each chemical follow a similar concentration-
response curve independent of sediment type. The data demonstrate that 50 percent
mortality occurs at about one sediment toxic unit, independent of chemical, organism or
sediment type, as expected if the EqP assumptions are correct. If we know the
appropriate normalizing phase then the same can be done for metals.
If the assumptions of EqP were exactly true and there were no experimental
variability or measurement error, then the data in Figures 2-4 and 2-5 should all predict 50
percent mortality at one toxic unit. There is an uncertainty of approximately a factor of
two in the results. This uncertainty associated with sediment quality criteria was obtained
from a quantitative estimate of the degree to which the data in Figure 2-5 support the
assumptions of EqP [21. This variation reflects inherent variability in these experiments as
-------
o
O
O
O C
C u)
S £
•a
"3
Q
o
c
o
= ~ c 2-
(0 C (0 TO
Jills
Q. HI U. < Q
OT OJT3
=
£°°
cv = Q .2
i S *|£ *°r
»- C M O 05 n
< O n O O > A
c
to
*3
(0
i
1
i
o
q
6
o
o
o
o
o
X
,2
4^
c
O
E
5
o
CO
o
^
o
CL
o
o
o
CO
o
(O
o
CM
q
6
o c
P o
o O
0
c
05 O
05 J3 c
£ — . - CO f-
nj
£ ill?S";tI
C ™i??^coCDTJ^=C
;^^;x«wSTj£c
.yj2°.2cD^-ot-o)
2* "P ^- "co S 05 ZT 5
C c 2. •- CO O
.^ en f*\ ^ /i* ^> r ^
>£*->»- ^O)COo
flj (Q CD * f ^™ ^^ .•.
?A 03 d^ ^^^ ™* ^j ~jjj ^^
^j ~r * * ^T ^~ ^
Q.
O
O _ *-•
•*-> C _
O -° C *^
"^ ^T^ "• ^^
o £3 w S ^J Z
•^ co cz *- .§
c -2 ^ | S 2
£ ^ = 05 IZ 05
p .b •*- w .E =
«•-=
=0 £ T5 T3
c.
O
TO
05 -. •- "° 9
g £ O — 05
I'lSl
U CO
05
O
CO Jr ^
i_ 05 M—
2 w -S 3 o o E
05 .*"fO+J— C05
> 0 CO CD .2 -£ ~ S5
•- 'E 'E OT m "* ^ «
—;c£a5t°00c:>
C05;C02:v..LLI«
a>
If)
CN
05 05 C ~ HI
^1 HgW
S S H S fe =
£ E ^ 'c 'i "^ .c
~.'-o " w CD 5*^
1 05 C 05 J= O >.
U. V) ••=• -O O ±£ £)
• to
-------
2-13
well as phenomena that have not been accounted for in the EqP model. This appears to
be the limit of the accuracy and precision to be expected.
Effects Concentration
The development of SQC requires an effects concentration for benthic organisms.
Since many of the organisms used to establish the water quality criteria (WQC) are
benthic, perhaps the WQC are adequate estimates of the effects concentrations for benthic
organisms. To examine this possibility, the acute toxicity data base, which is used to
establish the WQC was segregated into benthic and water column species, and the relative
sensitivities of each group was compared. The data are from the 40 freshwater and 30
saltwater U.S. Environmental Protection Agency (EPA) criteria documents. If it were true
that benthic organisms are as sensitive as water column organisms, then SQC could be
established using the final chronic value (FCV) from these WQC documents as the effects
concentration for benthic organisms. The apparent equality between the effects
concentration as measured in pore water and in water-only exposures (Figure 2-4) or as
predicted from organic .carbon normalization (Figure 2-5) supports using an effects
concentration derived from water only exposures.
This use of WQC assumes that (a) the sensitivities of benthic species and species
tested to derive WQC predominantly water column species, are similar and (b) the levels
of protection afforded by WQC are appropriate for benthic organisms. The assumption of
similarity of sensitivity using a comparative toxicological examination of the acute
sensitivities of benthic and water column species was presented [1,2]. A comparison of
the FCVs and the chronic sensitivities of benthic saltwater species in a series of sediment
colonization experiments was done [1,2]. Although there is considerable scatter, these
results, a more detailed analysis of all the acute toxicity data, and the results of benthic
colonization experiments support the contention of equal sensitivity [21.
The final validation of SQC will come from field studies that are designed to
evaluate the extent to which biological effects can be predicted from SQC. Sediment
-------
2-14
quality criteria for nonionic organic chemicals can possibility be validated more easily than
WQC because determining .the organism exposure is more straightforward. The benthic
population exposure is quantified by the organic carbon normalized sediment
concentration.
-------
REFERENCES
1. Di Toro, D.M., Zarba, C., Hansen, D.J., Swartz, R.C., Cowan, C.E., Allen, H.E.,
Thomas, N.A., Paquin, P.R., and Berry, W.J. 1991. Technical basis for
establishing sediment quality criteria for non-ionic organic chemicals using
equilibrium partitioning. Environ. Toxicol. Chem. 10:1541-1583.
2. U.S. Environmental Protection Agency. 1993a. Technical basis for deriving
sediment quality criteria for nonionic organic contaminants for the protection of
benthic organisms by using equilibrium partitioning. EPA 822-R-93-011. U.S.
Environmental Protection Agency, Office of Water. Washington, D.C.
3. U.S. Environmental Protection Agency. 1993b. Sediment quality criteria for the
protection of benthic organisms: ACENAPHTHENE. EPA 822-R-93-013. U.S.
Environmental Protection Agency, Office of Water. Washington, D.C.
4. U.S. Environmental Protection Agency. 1993c. Sediment quality criteria for the
protection of benthic organisms: DIELDRIN. EPA 822-R-93-015. U.S.
Environmental Protection Agency, Office of Water. Washington, D.C.
i
5. U.S. Environmental Protection Agency. 1993d. Sediment quality criteria for the
protection of benthic organisms: ENDRIN. EPA 822-R-93-016. U.S. Environmental
Protection Agency, Office of Water. Washington, D.C.
6. U.S. Environmental Protection Agency. 1993e. Sediment quality criteria for the
J protection of benthic organisms: FLUORANTHENE. EPA 822-R-93-012. U.S.
Environmental Protection Agency, Office of Water. Washington, D.C.
7. U.S. Environmental Protection Agency. 1993f. Sediment quality criteria for the
protection of benthic organisms: PHENANTHRENE. EPA 822-R-93-014. U.S.
Environmental Protection Agency, Office of Water. Washington, D.C.
-------
8. Swartz, R.C., Ditsworth, G.R., Schults, D.W. and Lamberson, J.O. 1985.
Sediment toxicity to a marine infaunal amphipod: Cadmium and its interaction with
sewage sludge. Mar. Environ. Res. 18:133-153.
9. Di Toro, D.M., Mahony, J.J., Hansen, D.J. Scott, K.J., Hinks, M.B., Mayr, S.M. and
Redmond, M.S. 1990. Toxicity of cadmium in sediments: The role of acid volatile
sulfide. Environ. Toxicol. Chem. 9:1487-1502.
10. Karickhoff, S.W. 1984. Organic pollutant sorption in aquatic systems. J. Hydraul.
Div. ASCE 110:707-735.
11. O'Connor, D.J. and Connolly, J. 1980. The effect of concentration of adsorbing
solids on the partition coefficient. Water Resour. 14:1517-1523.
12. Di Toro, D.M. 1985. A particle interaction model of reversible organic chemical
sorption. Chemosphere 14:1503-1538.
13. Abernethy, S. and Mackay, D. 1987. A discussion of correlations for narcosis in
aquatic species. In K.L. Kaiser, ed. QSARin Environmental Toxicology II. D. Reidel
Publishing Company, Dordrecht, The Netherlands. 1-16.
14. Leo, A. and Hansch, C. eds. 1986. Log (P) Database and Related Parameters.
Pomona College, Clarmont, California.
-------
CHAPTER 3
METAL TOXICITY IN WATER AND SEDIMENT EXPOSURES
The equilibrium partitioning methodology for establishing sediment quality criteria
requires that the chemical concentration be measured in the bioavailable phase and that
the chemical potential of the chemical be determined. In this chapter, two questions are
addressed. The first concerns the forms of metals that are bioavailable. The question is
addressed using water only exposures. The second concerns the observation that the
biological response is the same for water only exposures and for sediment exposures using
the pore water concentrations. This equality was demonstrated for non-ionic organic
chemicals and it is a fundamental tenent of the Equilibrium Partitioning model.
The data presented below demonstrate that biological effects correlate to the metal
activity, and that water only exposures and sediment exposures are equivalent. Therefore
for both metals and nonionic chemicals this EqP requirement is satisfied.
A direct approach to establishing sediment quality criteria for metals would be to
apply the water quality criteria to measured pore water concentrations. The validity of this
approach depends on the degree to which pore water concentration and represents free
metal activity and can be accurately measured in both systems. For most metals, free
metal activity can not be measured at water quality criteria concentrations and present
water quality criteria are not based on activity. Metals readily bind to dissolved (actually
colloidal) organic carbon (DOC), and DOC complexes do not appear to be bioavailable.
Hence the direct use of pore water concentration is precluded for metals with significant
DOC complexing.
By implication this difficulty extends to any complexing ligand that is present in
sufficient quantity. The decay of sediment organic matter can cause substantial changes
in interstitial water chemistry. In particular bicarbonate increases due to sulfate reduction.
This increases the importance of the metal-carbonate complexes and further complicates
the question of the bioavailable specie.
-------
3-2
The sampling of sediment interstitial water is not a routine procedure. The least
invasive technique employs a diffusion sampler which has cavities covered with a filter
membrane [2,3,4,51. The sampler is inserted into the sediment and the concentrations on
either side of the membrane equilibrate. When the sampler is removed the cavities contain
filtered pore water samples. The time required for equilibration depends on the pore size
of the membrane and the geometry of the cavity and usually exceeds one day.
An alternate technique is to obtain a sediment core, slice it, and filter or centrifuge
the slice to separate the pore water. For anaerobic sediments this must be done in a
nitrogen atmosphere to prevent the precipitation of iron hydroxide which would scavenge
the metals and yield artificially low dissolved concentrations [5,6].
Although either of these techniques are suitable for research investigations they
require more than the normally available sampling capabilities. If solid phase chemical
measurements were available from which pore water metal activity could be deduced it
would obviate the need for pore water sampling and analysis and it would circumvent the
need to deal with complexing ligands.
Toxicity Correlates to Metal Activity
A substantial number of water only exposure experiments disscussed below point
to the fact that biological effects can be correlated to the divalent metal activity {M2 + }.
The claim is not that the only bioavailable form is M2+ - for example MOH+ may also be
bioavailable - but that the DOC and certain other ligand complexed fractions are not
bioavailable.
The acute toxicity of cadmium to grass shrimp (Palaemonetes) has been determined
at various concentrations of chloride and NTA, both of which form cadmium complexes
[7]. The results are shown in Figure 3-1. The top panels are concentration response
curves as a function of total cadmium. The response is quite different at different
concentrations of chloride, indexed by salinity, and NTA. However if the concentration
-------
H7
S «J
Z3 Al
lii
So
u. "5 <
•J Q. OJ
?•- •** ii
bo- o
o 3E i—
ACUTE TOXI
GRASS SHRI
EFFEC
CD
O
7
e
4g
O
z
(0
(AFTER W.G.
l!\\\N
l/C •""**"*--•*.
{ ' **»,
/\/ "••--;-,
r*--V. '
I "^^--.
~~*~~~— ^ •««»_
*\ *
^
5 e> 0^0,^*1
£ "2s gS
J^ inn
4 • » • •
,1 1
t-
U
§
1
*"
"• 9
I
«
U
«
«
i___ »
•i i ' > >-
22220
>
£
u
<
IVAIAint 1N13IM
IVAIAini 1N13H4
sSg
3 w <
i§2
^3 E J
< W Q.
°2 2
li. o o
O O. O
HO. £
CD
O>
(O
O
220
• • •
O
O
o
<
u
... « » «
* i 2 !2 2
2 o2«2
Z 4 • • »
i—Hr
9 <
« u
IVAIAlHf 1N13II4
o 6 * " 2 ° °
e • « t «
IVAIAinS 1N13I14
C
0)
Q)
O
i
.Q
U
(0
I
i
"S "(5
O W
^ «
i-
.2 2
E"
•o
(0
o
_ -o
(0 C
(A <
flj ^^
±J ^^
|z
o ^2
P
a. o>
P 'x
•« E
•Si
8.
0 ~
So
< g
.
0) Q)
o>
ii
-------
3-4
response is evaluated with respect to Cd2 + activity in the solution, then the curves all
collapse into the same single curve (bottom panels). Comparable results have been
reported for copper-EDTA complexes [8] for which concentration response correlates to
Cu2+ activity (Figure 3-2, left top and bottom).
Chronic toxicity of zinc, with phytoplankton growth as the endpoint, has also been
examined. The results of an experiment in which the metal concentration is held constant
and the complexing ligand is varied are shown in Figure 3-2, right top and bottom [9]. As
NTA is added the toxicity of zinc to M/crocystis decreases. The cell density increases
rather than decreases in time and reaches control levels at the highest NTA concentration
(left top and bottom panel). The data can all be correlated to free zinc activity as shown
(right top and bottom panel). Similar results for diatoms exposed to copper and the
complexing ligand Tris are shown in Figure 3-3 (top) [10]. Variations in Tris concentrations
and pH produce markedly different growth rates (left top and bottom) which can all be
correlated to the Cu2+ activity (right). A similar set of results have been obtained by
Sunda and Lewis [11] with DOC from river water as the complexing ligand, Figure 3-3
(right top and bottom).
Metal bioavailability as measured by organism uptake can also be examined [12].
Uptake of copper by oysters is correlated not to total copper concentration (Figure 3-4 top)
but to copper activity (bottom).
The implication to be drawn from these experiments is that the partitioning model
required for establishing sediment quality criteria should predict {M2 + } in the pore water.
The following section examines the utility of this idea.
Interstitial Water And Metal Toxicity
This section presents some early data that first indicated the equivalence of pore
water concentrations and water only exposures. Much more data of this sort are
presented subsequently in Chapters 5 and 6. Swartz [13] tested the acute toxicity of
-------
ACUTE TOXICITY OF COPPER
TO A DINOFLAGELLATE
(FROM ANDERSON AND MOREL. 1976 )
CHRONIC TOXICITY OF ZINC
ON MICROCYSTIS AERUGINOSA
(FROM ALLEN, et.al., 19€0)
10 •
56 7 8 9
TOTAL COPPER (-LOG(CuTJ)
O A AP •/! 1 M F« EOT* •/. T M
4 I id' M Zn
.1.0.10 M NTA
. JS.10"*M
. «0.10 M
.1.0.10* M
. SO.IO' M
.1.0.K M
20
100
90
60
ro
§ 60
so
40
30
20
10
0
10'H
e 10*-
u
o
10
K* •! pH 7.0
10 II 12 13
COPPER ACTIVITY (pCu)
1.0 2.0 3.0 4.0 5.0
Free Zinc (moles/liter xlO7)
Figure 3-2. Acute toxicity to a dinof lagellate (left) of total copper (top) and copper activity
(bottom), with and without EDTA. Chronic toxicity of zinc to Microcvstis aeruoinosa
(right) showing growth as cells/ml versus time with different levels of EDTA and NTA (top)
and number of cells at five days as a function of free zinc concentration (bottom).
-------
^
t
•
r
tr
LU
0. •-> a
Q. DC I"
O IL) Ol
CJI -
**• ^ in
v en 5
"ir ?
X 7C '•
O 1 ' -•.
0 ° io
§ 0 g
tr*- a
•Ji u.
o —
1
\^^s^
\\,
\ ^x.
^l ^x^_
~ it ^( ~
X. \
^V \
^v \
^v \
>. \
\
^
"ii i
03 —
CJ «_
03 a o
Q.
> a
« °
>
" "co
CJ *^
< O
0 :- w
0 < •
in o
o
o
111 tlf Ul
t~ ^ ^ ^^
552 \
SEE
sis
• • 4
=i c- w
r» CJ gj
£
.S5
CO
o ^
o J
o o o • o o o •
O O O^ O O O «B (/)
ni »» o cu « o (/)
AVQ/SNOISIAIO AVQ/SNOISIAIQ •£
NI 31VH KlMObS OIJI03dS NI 31Vb HIMObS OIdI03dS O
hb ^
F
IT
Ul
a.
Q.
o
o
si
ts
0°
x£
•"
u
•~
z
o
u
to
Ot
0*
e
<
_i
j
5
o
o
z
<
<
o
z
3
M
~"
i
i
i
i
i
i '
^»j* \
i x \
1 x^ \ '
i « \
— •— — 1- »*• o
i x*
i %
| N
1 N^
t »x
-L ^v
i ^
Nl ./~>^
S z * a a ~*"" * "
tt\^*O^®^ ** 1 ^ t ^^1^
O.X04CO- o|c> a O &(•
i "t
7"
%
a
in O
C9
O
_J
1
cc
Ul
a.
&.
o
j
4
t-
0
H
10
al-
^T!
1
1
i
JL i
" 'Llf
! rt
it
— *-o —
Jfl
~""""_jy—
Li •
~Or~
-4r-
i ,
"r" •* Q-J- I
'
<£
T3
CO
__^
^j
• H—
~
f**
u re
^ T3
o> ro
> "o
\ ~
< CO
K «-
A ^
•^ ^—
tf> O §
O O
J^
O
O
Q
"co
*-•
CO
c
CO
•—
1-
c
CO
O)
o>
c
'x
CD
o
o
03
r*
4-"
"o
CO
ntration
f conce
o
03
o>
c
CO
0) (0
o «-
o ^
03 ^
Q. C
• ^^
t
1 i !
° B e
9
o
*-<
.Q
AVQ/SNOISIAIO NI
HlMO«OOUID3dS
£ •-
.
LL CO
-------
UPTAKE OF COPPER BY OYSTERS
ISO
si
2 100
»
0 »
OC a
u o
01
0.01
*
A
I
O.I 1.0
TOTAL COPPER CONC.
10.0
x
Ul
5
g
f "8
Z ^5
H
j »
1 *
o
5
a.
8-
V
200
180
160
140
120
100
80
60
40
20
I
TOTAL NTA (uM)
Winter
A 1.0
o 10.0
Summer
A 1.0
• 8.1
- 1
' \
r 1 U
. , 1 r*
i
"t
v!
' ! it IT
t
i i i
9 10 11
COPPER ACTIVITY (pCu)
AFTER C.D. ZAMUDA AND W.G. SUNDA
USING (Crassostrea virginca ), I982
Figure 3-4. Body burden of copper in oysters versus total copper (top) and copper activity
(bottom) with different levels of the complexing ligand NTA.
-------
3-8
cadmium to the marine amphipod, Rhepoxvnius abronius in sediment and seawater. An
objective of the study was to determine the contributions of interstitial and particle-bound
cadmium to toxicity. Figure 3-5 presents mean survival versus dissolved cadmium
concentration for 4-day toxicity tests in seawater and interstitial water. A comparison of
the 4-day LC50 of cadmium in interstitial water (1.42 mg/L) with the 4-day LC50 of
cadmium in seawater without sediment (1.61 mg/L) resulted in no significant difference
between the two.
Experiments were performed to determine the role of acid volatile sulfides in
cadmium spiked sediments using the amphipods Amoelisca abdita and Rhepoxvnius
hudsoni [141. Three sediments were used, a Long Island Sound sediment with high AVS,
Ninigret Pond sediment with low AVS concentration and a 50/50 mixture of the two
sediments. Figure 3-6 presents a comparison of the observed mortality to the observed
interstitial water cadmium activity, measured with a specific ion electrode, for the three
sediments. Four-day water only and 10-day exposure sediment toxicity tests were
performed. The water-only response data iorAmpe/isca and Rhepoxynius are included for
comparison although they represent a shorter duration exposure.
An elegant experimental design was employed by Kemp and Swartz [15] to examine
the relative acute toxicity of particule bound and dissolved interstitial cadmium. They
circulated water of the same cadmium concentration through different sediments. This
result in differing bulk sediment concentrations but the same interstitial water
concentrations. They found no statistically significant difference in organisim response for
1 the different sediments. Since the interstitial water concentrations were the same in each
A
treatment - the circulating water concentrations established the interstitial water
concentrations - these experiments confirmed the equal concentration hypothesis.
A series of 10-day toxicity tests using the amphipod Hvalella azteca were performed
to evaluate the bioavailability of copper in sediments from two sites highly contaminated
i with this metal: Steilacoom Lake, Washington, and Keweenaw Watershed, Michigan [16],
A water-only, 10-day copper toxicity test was also conducted with the same organism.
-------
-COMPARISON OF WATER AND
SEDIMENT EXPOSURE
( AFTER R.C.SWARTZ etal., 1985)
20
rr
CO
<
LL)
2
CO
X
o
a.
LJ
X
-------
o
o
DC
LLJ
J
<
CO
DC
LU
CO
cc
o
O
o
o
o>
o
o
•
I
o
q
co
O
CM
•f
•o
O
O
o
o
o
o
CO
o
(O
o
CM
o
q
in
i
(%) AinVlHOlAI
CO
4->
CO
•o
03
3
U)
O
a
x
V
c
o
l_
0)
•M
(0
CO
£
V
E
0)
(0
0)
2 «
f- 4-"
•p CD
U. «
0 CD
•*- 2
> CO
O
co r:
Si
CO
CO O
.E co
CO
a>
w .t
0)
CD
!
CO
LL (0
-------
3-11
Figure 3-7 presents a comparison of the resulting mortality from the water-only test and
interstitial water from the sediment tests. The LCSOs for the water-only and the average
of day 0 and day 10 pore-water concentration were 31 and 28 ug/L respectively showing
strong agreement in predicting toxicity.
The data presented in this chapter demonstrate that in water-only exposures metal
activity and concentration can be used to predict toxicity. The results of four experiments
demonstrate that mortality data from water-only exposures can be used to predict
sediment toxicity using pore water concentrations. The metal activity or concentration in
interstitial water therefore would be an important component of a partitioning model that
is needed to establish sediment quality criteria. The following chapters present the current
developments in determining partitioning and hence bioavailability for sediment metals.
-------
Percent Mortality
§
o
o
•o
IS *»
CO o
-1 O
?
OS
o
o
-------
References
1. Berner, R.A. 1980. Early Diagenesis. A Theoretical Approach. Princeton Univ.
Press, Princeton, N.J.
2. Hesslein, R.H. 1976. An in situ sampler for close interval pore water studies.
Limnol. Oceanogr. 21:912-914.
3. Carignan, R. 1984. Interstitial water sampling by dialysis: Methodological notes.
Limnol. Oceanogr. 29(3):667-670.
4. Carignan, R., Rapin, F. and Tessier, A. 1985. Sediment porewater sampling for
metal analysis: a comparison of techniques. Geochimica et Cosmochem. Acta
49:2493-2497.
5. Allen, H.E., Fuj, G., Deng, Baolin. 1993. Analysis of acid-volatile sulfide (AVS) and
simultaneously extracted metals (SEM) for the estimation of potential toxicity in
aquatic sediments. Environ. Toxicol. Chem. 12:001-013.
6. Troup, B.N. 1974. The Interaction of Iron with Phosphate, Carbonate, and Sulfide
in Chesapeake Bay Interstitial Waters: A f hermodynamic Interpretation Ph.D Thesis.
Johns Hopkins University, Baltimore, Maryland. 1-114.
7. Sunda, W.G., Engel, D.W. and Thuotte, R.M. 1978. Effect of chemical speciation
of toxicity of cadmium to grass shrimp, Palaemonetes pugio: Importance to free
cadmium ion. Environ. Sci. Tech. 12:409-413.
8. Anderson, D.M. and Morel, F.M.M. 1978. Copper sensitivity of Gonvaulax
tamarensis. Limnol. Oceanogr. 23:283-295.
9. Allen, H.E., Hall, R.H. and Brisbin, T.D. 1980. Metal speciation. effects on aquatic
toxicity. Environ. Sci. Technol 14:441- 443.
-------
10. Sunda, W. and Guillard, R.R.L. 1976. The relationship between cupric ion activity
and the toxicity of copper to phytoplankton. J. Mar. Res. 34:511-529.
11. Sunda, W.G. and Lewis, J.M. 1978. Effect of complexation by natural organic
ligands on the toxicity of copper to a unicellular alga, Monochrvsis lutheri. Limnol.
Oceanogr. 23:870-876.
12. Zamuda, C.D. and Sunda, W.G. 1982. Bioavailability of dissolved copper to the
American ovster Crassostrea virainica. Importance of chemical speciation. Marine
Biology 66:77-82.
13. Swartz, B.C., Ditsworth, G.R., Schults, D.W. and Lamberson, J.O. 1985.
Sediment toxicity to a marine infaunal amphipod: Cadmium and its interaction with
sewage sludge. Mar. Environ. Res. 18:133-153.
14. Di Toro, D.M., Mahony, J.D., Hansen, D.J. Scott, K.J., Hinks, M.B., Mayr, S.M.
and Redmond, M.S. 1990. Toxicity of cadmium in sediments: The role of acid
volatile sulfide. Environ. Toxicol. Chem. 9:1487-1502.
15. Kemp, P.P., and Swartz, R.C. 1986. Acute toxicity of interstitial and particle-
bound cadmium to a marine infaunal amphipod. Mar. Environ. Res. 26:135-153.
16. Ankley, G.T., Mattson, V.R., Leonard, E.N., West, C.W., and Bennett, J.L. 1993.
Predicting the acute toxicity of copper in freshwater sediments: Evaluation of the
role of acid-volatile sulfide. Environ. Toxicol. Chem. 12:315-320.
-------
CHAPTER 4
METAL PARTITIONING
The state of the art of modeling metal sorption to oxides in laboratory systems is
well developed and detailed models are available for cation and anion sorption [see the
articles in Stumm, [1] and Dzombak and Morel, 12] for recent summaries]. The models
consider surface complexation reactions as well as electrical interactions via models of the
double layer. Models for natural soil and sediment particles are less well developed.
However, recent studies suggest that similar models can be applied to soil systems
13,4,5,6,7,8,91. Since the ability to predict partition coefficients is required if pore water
metal concentration is to be inferred from the total concentration, some practical model
is required. This chapter presents the theoretical development of metals partitioning in
sediments.
Metal Sorption Phases
The initial difficulty that one confronts in selecting an applicable sorption model is
that the available models are quite complex and many of the parameter estimates may be
specific to individual soils or sediments. However the success of organic carbon based
non-ionic chemical sorption models suggests that some model of intermediate complexity
that is based on an identification of the sorption phases may be more generally applicable.
A start in this direction was made during a recent conference [10]. A more formal
presentation is available [11]. The basic idea was that instead of considering only one
sorption phase as is assumed for non-ionic hydrophobic chemical sorption, multiple
sorption phases were considered. The conventional view of metals speciation in aerobic
soils and sediments is that metals are associated with the exchangable, carbonate, and Fe
and Mn oxide forms, as well as that associated with the organic matter and stable metal
sulfides, and a residual phase. In oxic soils and freshwater sediments sorption phases
have been identified as paniculate organic carbon (POO and the oxides of iron and
manganese [12,13,14,15]. These phases are important because they have a large
-------
4-2
sorptive capacity. Further they appear as coatings on the particles and occlude the other
mineral 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 labile metal sulfide interactions which dominate the
speciation in the anaerobic layers of the sediment.
Titration Experiments
The importance of sulfide in the control of metal concentrations in the interstitial
water of marine sediments is well documented [16,17,18,19]. Metal sulfides are very
insoluble and the equilibrium interstitial water metal concentrations in the presence of
sulfides are small. If the interstitial water sulfide concentration in sediments is large, then
as metal is added to the sediment, metal sulfide would precipitate following the reaction:
M2+ + S2--»MS(s) <4'1)
This appeared to be happening during a spiked cadmium sediment toxicity test [20] since
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 indicated that there was little or no free sulfide in the unspiked sediment. This
was, at the time, a most puzzling result.
The lack of 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 this solid phase sulfide initially present. As cadmium is added to
the sediment it causes the solid phase iron sulfide to dissolve releasing sulfide which is
available for the formation of cadmium sulfide. The reaction is:
-------
4-3
Cd2+ + FeS(s) - CdS(s) + Fe2* <4-2>
Cadmium titrations with amorphous FeS and with sediments were performed to examine
this possibility.
Amorphous FeS
A direct test of the extent to which this reaction takes place was performed [20].
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, [Cd]A, to the amount of FeS initially present, [FeS(s)]j, is shown
in Figure 4-1. The plot of dissolved cadmium versus cadmium added illustrates the
increase in dissolved cadmium that occurs near [Cd]A / lFeS(s)]j = 1. A similar experiment
has been performed for amorphous MnS with comparable results. It is interesting to note
that these displacement reactions among metal sulfides have been observed by other
investigators [21]. The reaction was also postulated by Pankow [22] to explain an
experimental result involving copper and synthetic FeS.
These experiments plainly demonstrate that solid phase amorphous iron and
manganese sulfide can readily be displaced by adding cadmium. As a consequence it is
a source of available sulfide which must be taken into account in evaluating the
relationship between solid phase and aqueous phase cadmium in sediments. A direct
confirmation that the removal of cadmium was via the displacement of iron sulfide is
shown in Figure 4-2. The supernatant from a titration of FeS by Cd2+ was analyzed for
both cadmium and iron. The solid lines are the theoretical expectation based on the
stoichiometry of the reaction (Equation 4-2).
Sediments
A similar titration procedure has been used to evaluate the behavior of sediments
taken from four quite different marine environments: the Long Island Sound and Ninigret
-------
4-6
Pond sediments used in the toxicity tests; and sediments from Black Rock Harbor and the
Hudson River. The binding, capacity for cadmium is estimated by extrapolating a straight
line fit to the dissolved cadmium data. The equation is:
[ZCd(aq)J = max{0,m([Cd]A - fCd]B)}
(4-3)
where [ICd(aq)] is the total dissolved cadmium, [Cd]A is the cadmium added, [CdIB is the
bound cadmium, and m is the slope of the straight line. The sediments exhibit quite
different binding capacities for cadmium, listed in Table 4-1, ranging from approximately
1 /ymol/gm to more than 100 //mol/g. The question is whether this binding capacity is
explained by the solid phase sulfide present in the samples.
TABLE 4-1. CADMIUM BINDING CAPACITY AND AVS OF SEDIMENTS
Final AVS
Sediment
Black Rock Harbor
Hudson River
LI SoundOe(c)
Mixture10'
Ninigret Pond
-------
4-7
is associated with the more soluble iron and manganese 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 associated with the organic matter
in sediments [25].
The possibility that acid volatile sulfide is a direct measure of the solid phase sulfide
that reacts with cadmium is examined in Table 4-1 which lists the sediment binding
capacity for cadmium and the measured AVS for each sediment and in Figure 4-3 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, Table 4-1, suggests that some AVS is lost during titration experiment. In
any case the covariation of sediment binding capacity and AVS is clear in the data in Table
4-1 and Figure 4-3. This suggests that AVS is the proper quantification of the solid phase
sulf ides that can be dissolved by cadmium. The chemical basis for this is examined below.
Solubility Relationships and Displacement Reactions
Iron monosulfide, FeS(s), is in equilibrium with aqueous phase sulfide and iron
concentration via the reaction:
FeS(s) ~ Fe2+ + S2' (4'4)
If cadmium is added to the aqueous phase, the result is:
Cd2* + FeS(s) «• Cd2* + Fe2* + S2' (4'5)
As the cadmium concentration increases, [Cd2 + HS2"] will exceed the solubility product of
cadmium sulfide and CdS(s) will start to form. Since cadmium sulfide 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:
-------
f
4-8
4-3. Cadmium titration of sediments as indicated in figure legend. Cadmium added per
unit dry weight of sediment versus total dissolved cadmium.
-------
0)
z
UJ
Q
UJ
0)
LJL
O
fiC
D
O
<
O
q
d
o
o
c
Q>
E
0>
W
t
q
d
o
q
d
•o
E
\
•o
O
"o
o
UJ
o
o
o
<
o
o oo
• •
i- O
«q
d
CM
•
O
q
d
d
asAiossia
O3
"c
<5
a.
-a
01
•o
•o
CO
E
•o
CO
O
•o
CD
O>
D)
•o
0)
+•"
(O
U
U)
CO
CO
€
0)
E
'o
CD
(0
II
4^ CO
•*= o
C T3
I ^
E o
•o w
ro .52
O T3
• ro
£
3 W
.2» QJ
-------
4-9
Cd2+ + FeS(s) - CdS(s) + Fe2* (4'6)
The iron in FeS(s) is displaced by cadmium to form soluble iron and solid cadmium sulfide,
CdS(s). The consequence of this replacement reaction can be seen using an analysis of
the M(ll)-Fe(ll)-S(-ll) system with both MS(s) and FeS(s) present in Appendix 4A. M(ll)
represents any divalent metal that forms a sulfide 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 products:
{M2+}/[M]A < KMS/KFeS (4-7)
This is a general result that 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 4-6. However, the
ratio is still less than the ratio of the sulfide solubility products.
This is an important finding since the data presented in Chapter 3 indicates that
toxicity is related to metal activity, {M2+}. This inequality guarantees that the metal
3 activity - in contrast to the total dissolved metal concentration - is regulated by the iron
sulfide - metal sulfide system.
The sulfide solubility products and the ratios are listed in Table 4-2. The ratio of
cadmium activity to total cadmium is less than 10 '10-5. For nickel the ratio is less than
, 10"5-6. 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 4-2, Zn, Cd, Pb, and Cu, indicate that metal activity for these
metals will be very small in the presence of excess AVS.
-------
4-10
TABLE 4-2. METAL SULFIDE
Metal Sulfide ' log
*»p.2
FeS -3.64
NiS -9.23
ZnS -9.64
CdS -14.10
PbS -14.67
CuS -22.19
•Solubility products, K_p y, for the reaction M2+ +
(mackinawite), and NiS imnlerite) from ref 27. Solubility
(wurtzite), and pK, = 18.57 for the reaction HS" •» H+
MS(s) is computed from log Ksp 2 and pK2.
SOLUBILITY PRODUCTS"
jOfl
K«P
-22.39
-27.98
-28.39
-32.85
-33.42
-40.94
Log
(KMS/KFaS)
-5.59
-6.00
-10.46
-11.03
-18.55
H§- » MS(s) + H+ for CdS (greenockite). FeS
products for CuS (covellite), PbS (galena), and ZnS
+ S2' from [28). K,p for the reaction M24 + S7' -
Application to Mixtures of Metals
A conjecture based on the sulfide solubility products for the metals listed in Table
4-2 is that the AVS normalized toxicity of metals is additive. Since all these divalent
metals have lower sulfide solubility parameters than FeS, they would all exist as metal
sulfides if their molar sum is less than the AVS. For this case
< 1 (4-8)
[AVS]
no metal toxicity would be expected where [MT]j is the total cold acid extractable ith metal
concentration in the sediment. 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:
> 1 (4-9)
[AVS]
These two equations are precisely the formulas that one would employ 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 one discriminates between non toxic and potentially toxic sediments. The
-------
4-11
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.
The appropriate quantity of metals to use in the metals/AVS ratio is referred to as
"simultaneously extracted metal" or SEM. This is the metal which is extracted in the cold
acid used in the AVS procedure. This is the appropriate quantity to use because some
metals form sulfides which are not labile in the AVS extraction (e.g., nickel). If a more
rigorous extraction were used to increase the fraction of metal extracted which did not
also capture the additional sulf ide extracted, then the sulf ide associated with the additional
metal release would not be quantified. This would result in an erroneously high metal to
AVS ratio [26].
The above discussion is predicated on the assumption that all the metal sulfides
behave similarly to cadmium sulfide. Results of sediment spiking experiments will be
presented in Chapter 5 for cadmium, copper, lead, nickel, zinc, and metals mixtures which
demonstrate the similar behavior of these metals. Further 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 can be tested
directly using sediment toxicity tests. These are discussed in the next chapters.
-------
Appendix 4A
Solubility Relationships for Metal Sulfides. Consider the following situation: a
quantity of FeS is titrated with a metal that forms a more insoluble sulfide. We analyze
the result using an equilibrium model of the M-(ll)-Fe(ll)-S(-ll) system. The mass action
laws for the metal and iron sulfides are
= KMS (A-1)
= KFeS (A"2>
where [M2+], IFe2+], and IS2"] are the molar concentrations; KM**- Kpe2*' ancl Ks2' are the
activity coefficients; and KMS and KFeS are the sulfide solubility products. The mass
balance equations for total MOD, Fe(ll), and S(-ll) are
[MS(s)l - [M]A (A-3)
[FeS(s)l = [FeS(s)]j (A-4)
+ [MS(s)l + [FeS(s)l = [FeSlsHj (A'5>
where
2* = IM2*]/[ZM{aq)] (A-6)
* = [Fe2*]/[ZFe(aq>] (A-7)
aS2- = [S2-]/[ZS(aq)l (A-8)
are the ratios of the divalent species concentrations to the total dissolved M(ll), Fe(ll), and
SMI) concentrations, [ZM(aq)], [ZFe(aq)], and IZS(aq)], respectively. [MS(s)l and [FeS(s)]
are the concentrations of solid-phase metal and iron sulf ides at equilibrium. [FeS(s)]j is the
-------
4A-2
initial iron sulfide concentration in the sediment, and [M]A is the concentration of added
metal.
The solution of these five equations can be obtained as follows, the mass balance
Equations A-3 and A-4 for M(ll) and Fe(ll) can be solved for [MS(s)l and IFeS(s)] and
substituted in the mass balance Equation A-5 for S(ll):
-a"1s2-[S2-] + a~1F.2+[Fe2+]
[M]A
(A-9)
The mass action Equations A-1 and A-2 can be used to substitute for [Fe2 + ] and [M2 + ],
which results in a quadratic equation for [S2"]:
The positive root can be accurately approximated by
Fe2*KFeS j. a M2*K
-1
[M]A
IM1A
(A-10)
(A-11)
which results from ignoring the leading term in Equation A-10. This is legitimate because
the term in parentheses in Equation A-10 is small relative to [M]A due to the presence of
the sulfide solubility products. As a results, IS2"] is also small since it is in the
denominator. Hence, the leading term in Equation A-10 must be small relative to [M]A and
can safely be ignored.
The metal activity can now be found from the solubility equilibrium Equation A-1:
-------
4A-3
KMS
-1 „ -1
a F«2*KFeS j.a r
-1
[M]A
so that
K
MS
IM]A
where
and
Equation A-13 can be expressed as
KMS
[M]A KFeS
KMS
KFeS
(A-13)
(A-14)
(A-15)
(A-16)
The magnitude of the term in parentheses can be estimated as follows. The first term in
the denominator is always greater than or equal to 1 ,/?Fe2+ ^ 1 • because it is the reciprocal
of two terms both of which are less than or equal to 1, Equation A-14. They are aFe2* <
1, which is the ratio of the divalent to total aqueous concentration, and Yf^+ ^ 1, which
is an activity coefficient. The second term in the denominator cannot be negative,
#M2+KMS/KFeS > 0, since all of its terms are positive, thus, the denominator of the
expression in parentheses is always greater than 1, /?Fe2+ + /?M2+KMs/KKeS > 1.
Therefore, the expression in parentheses is always less than 1. Hence, the magnitude of
the ratio of metal activity to total added metal is bounded from above by ratio of the
sulfide solubility products:
-------
4A-4
{Me2+}/[MlA < KMS/KFeS (A'17)
This results applies if [FeS]-, > [M]A so that excess [FeS(s)l is present.
If sufficient metal is added to exhaust the initial quantity of iron sulfide, then
[FeS(s)] = 0. Hence, the iron sulfide mass action equation (A-2) is invalid and the above
equation no longer applies. Instead, the only solid-phase sulfide is metal sulfide and
[MSI = [FeSlj (A-18)
so that, from the metal mass balance equation
{M2+} - KMZ^M^MA ' IFeS(s)j) (A-19)
this completes the derivation of Equations 4-8 and 4-9.
-------
4A-5
[AVS]
[Fe2 + ]
[FeS(s)]
[FeS(s)]j
KFeS
[M2 + ]
[M]A
IMS(s)]
{s2-}
IS2']
[SEM]
[SEMJCd
[SEM]Cu
[SEM]Nj
[SEM]Pb
[SEM]Zn
Glossary
acid volatile sulfide concentration (//mol/g)
activity of Fe2+ (mol/L)
concentration of Fe2+ (mol/L)
concentration of iron sulfide (mol/L)
initial iron sulfide concentration in the sediment (mol/L)
solubility product for FeS(s) [(mol/L)2]
solubility product for MS(s) [(mol/L)2]
divalent metal activity (mol/L)
concentration of M2+ (mol/L)
concentration of added metal (mol/L)
concentration of solid-phase metal sulfide (mol/L)
activity of S2' (mol/L)
concentration of S2" (mol/L)
simultaneously extracted metal concentration (//mol/g)
simultaneously extracted Cd concentration (//mol/g)
simultaneously extracted Cu concentration (//mol/g)
simultaneously extracted Ni concentration (//mol/g)
simultaneously extracted Pb concentration (//mol/g)
simultaneously extracted Zn concentration (//mol/g)
(Fe2-|-}/[ZFe(aq)]
{M2 + }/[Zm(aq)l
{S2-}/[ZS(aq)J
[ZFe(aq)]
activity coefficient of Fe2 +
activity coefficient of M2 +
activity coefficient of S2"
concentration of total dissolved Fe(ll) (mol/L)
-------
4A-6
(ZM(aq)] concentration of total dissolved M(ll) (mol/L)
IZS(aq)] concentration of total dissolved S(ll) (mol/L)
-------
17. Emerson, S., Jacobs, L. and Tebo, B. 1983. The behavior of trace metals in marine
anoxic waters: Solubilities at the oxygen-hydrogen sulfide interface. In: Trace
Metals in Sea Water, pp. 579-608. Editors: C.S. Wong, E. Boyle, K.W. Bruland and
J.D. Burton. Plenum Press, New York.
18. Davies-Colley, R.J., Nelson, P.O. and Williamson, KJ. 1985. Sulfide control of
cadmium and copper concentrations in anaerobic estuarine sediments. Marine
Chemistry 16:173-186.
19. Morse, J.W., Millero, F.J., Cornwell, J.C. and Rickard, D. 1987. The chemistry
of the hydrogen sulfide and iron sulfide systems in natural waters. Earth Science
Reviews 24:1-42.
20. Di Toro, D.M., Mahony, J.J., Hansen, D.J. Scott, K.J., Hinks, M.B., Mayr, S.M. and
Redmond, M.S. 1990. Toxicity of cadmium in sediments: The role of acid volatile
sulfide. Environ. Toxicol. Chem. 9:1487-1502.
\
21. Phillips, H.O. and Kraus, K.A. 1965. Adsorption on inorganic materials VI. Reaction
of insoluble sulfides with metal ions in aqueous media. J. Chromatog 17:549-557.
22. Pankow, J.F. 1979. The dissolution rates and mechanisms of tetragonal ferrous
sulfide (Mackinawite) in anoxic aqueous systems. Ph.D thesis. California Inst. of
Technology, Pasadena, Calif. 1-146.
23. Berner, R.A. 1967. Thermodynamic stability of sedimentary iron sulfides. Am. J.
J Sci. 265:773-785.
24. Goldhaber, M.B. and Kaplan, I.R. 1974. The sulfur cycle. In: The Sea. Vol. 5.
Marine Chemistry, 569-655. Editor: E.D. Goldberg. J. Wiley & Sons, New York.
-------
25. Landers, A.M., David, M.B., Mitchel, M.J. 1983. Analysis of organic and inorganic
sulfur constituents in sediments, soils, and water. Int. J. Environ. Anal. Chem.
14:245-256.
26. Di Toro. D.M., J.D. Mahony, DJ. Hansen, KJ. Scott, A.R. Carlson and G.T.
Ankley. 1992. Acid volatile sulfide predicts the acute toxicity of cadmium and
nickel in sediments. Environ. Sci. and Technol. 26:96-101.
27. Schoonen, M.A.A. and Barnes, H.L. 1988. An approximation of the second
dissociation constant for H2S. Geochim. Cosmochim. Acta 52:649-654.
28. Byrne, R.H., Kump, L.R. and Cantrell, KJ. 1988. The influence of temperature and
pH on trace metal speciation in seawater. Marine Chemistry 25:163-181.
-------
REFERENCES
1. Stumm,W. 1987. Aquatic Surface Chemistry. Chemical Processes at the Particle-
Water Interface. John Wiley & Sons, New York.
2. Dzombak, D.A. and Morel, F.M.M. 1990. Surface Comolexation Modeling.
Hvdrous Ferric Oxide. New York, NY, John Wiley & Sons.
3. Goldberg, S. and Sposito, G. 1984. A chemical model of phosphate adsorption by
soils. H. Noncalcareous soils. Soil Sci. Cos. Am. J. 48:779-783.
4. Barrow, N.J. 1986. Testing a mechanistic model. I. The effects of time and
temperature on the reaction of fluride and molybdate with a soil. J. of Soil Science
37:267-277.
5. Barrow, N.J. 1986. Testing a Mechanistic Model. II. The effects of time and
temperature on the reaction of zinc with a soil. J. of Soil Science 37:287-295.
6. Barrow, N.J. and Ellis, A.S. 1986. Testing a mechanistic model. III. The effects
of pH on fluoride retention by a soil. J. of Soil Science 37:287-295.
7. Barrow, N.J. and Ellis, A.S. 1986. Testing a mechanistic model. IV. Describing
the effects of pH on zinc retention by a soil. J. of Soil Science 37:295-302.
8. Barrow, N.J. and Ellis, A.S. 1986. Testing a mechanistic model. V. The points
of zero salt effect for phosphate retention and for acid/alkali titration of a soil. J.
of Soil Science 37:303-310.
9. Sposito, G, de Wit, J.C.M., and Neal, R.H. 1988. Selenite adsorption on alluvial
soils: III. Chemical modeling. Soil Sci. Soc. Am. J. 52:947-950.
-------
10. Di Toro, D.M., Harrison, F., Jenne, E., Karickhoff, S., and Lick, W. 1987.
Synopsis of discussion session 2: Evironmental fate and compartmentalization. In:
Fate and Effects of Sediment-Bound Chemicals in Aquatic Systems. 136-147.
Editors: K.L Dickson, A.W. Maki and W.A. Brungs. Pergamon Press, New York.
11. Jenne, E.A., Di Toro, D.M., Allen, H.E., and Zarba, C.S. 1986. An activity-based
model for developing sediment criteria for metals: A new approach. Chemicals in
the environment. International Conference, Lisbon, Portugal.
12. Jenne, E.A. 1968. Controls on Mn, Fe, Co, Ni, Cu, and Zn concentrations in soils
and water - the significant role of hydrous Mn and Fe oxides. In: Advances ]n
Chemistry. 337-387. Editor: American Chemical Society, Washington, D.C.
13. Jenne, E.A. 1977. Trace element sorption by sediments and soil -- sites and
processes. In: Symposium on Molybdenum in the Environment. Vol.2, pp. 425-
553. Editors: W. Chappell and K. Petersen. M. Dekker, Inc., New York.
14. Luoma, S.N. and Bryan, W. 1981. A statistical assessment of the form of trace
metals in oxidized sediments employing chemical extractants. The Sci of the Total
Environment 17:165-196.
15. Oakley, S.M., Williamson, K.J., and Nelson, P.O. 1980. The geochemical
partitioning and bioavailability of trace metals in marine sediments. Water Res.
Inst., Oregon State Univ., Corvallis, Oregon. 1-84.
16. Boulegue, J., Lord III, C.J. and Church, T.M. 1982. Sulfur speciation and
associated trace metals (Fe, Cu) in the pore waters of Great Marsh, Delaware.
Geochim.. Cosmochim.. Acta. 46:453-464.
-------
CHAPTER 5
LABORATORY SPIKING EXPERIMENTS
The discussion in the previous chapter has highlighted the importance of metal
sulfides. Sulfides of Cd, Cu, Ni, Pb, and Zn all have lower sulfide solubility product
constants than the sulfides of iron and manganese, which are formed naturally in
sediments as a product of the bacterial oxidation of organic matter [1 ]. As a result, these
metals will displace manganese and iron whenever they are present together with
manganese and iron monosulfides [2]. Because the solubility product constants of these
sulfides are so small, sediments with an excess of AVS will have very low metal activity
in the interstitial water, and no toxicity due to these metals should be observed in the
sediments. If the metals are present in excess of the sulfides (SEM/AVS > 1.0) and there
are no other sediment phases capable of binding the metals (e.g., DOC or TOO metal will
be present in the interstitial water and the sediment may be toxic. The validity of this
theory can be demonstrated through experimentation. Results of acute toxicity testing
with sediments spiked with metals in the laboratory are presented in this chapter.
Predictions of the toxicity of metals-contaminated field sediments using interstitial water
concentration of metals and AVS normalization are presented in Chapter 6. Results of
chronic toxicity tests using sediment colonization experiments are presented in Chapter 7.
In this chapter results from a series of acute toxicity tests using saltwater sediments
spiked with cadmium, copper, lead, nickel, or zinc, and an equimolar mixture of cadmium,
copper, nickel, and zinc will be examined in detail. The methodology for these tests is
presented in Appendix 5A. These tests will be highlighted because they serve as an
example from a single laboratory (Narragansett EPA Research Laboratory) of the methods
used in the sediment spiking experiments with metals and represent a series of tests which
followed a consistent methodology, performed with a relatively sensitive species unable
to avoid the sediment (Amoelisca abdita. the amohipod). A. .abdjta is an estuarine, tube-
building, infaunal amphipod commonly used in sediment toxicity testing. Published results
from tests using polychaetes [4,5] and copepods [6] in saltwater sediments, and
oligochaetes and snails [7] in freshwater sediments, will be combined with the A. abdita
-------
5-2
results. Data from spiked sediment tests in which neither AVS nor interstitial water were
measured [8,9] or which used non-standard methods [10] will not be included.
Results: Saltwater Amphipod Tests
The data handling techniques used in this chapter are discussed below. Detection
limits were calculated for ail chemical analyses based on instrument detection limits and
sample size. In those instances where a mean concentration is a summation of measured
data and data below the limit of detection, 112 the detection limit was used for those
values below the limit of detection. Means for which there were no measured values
above the detection limit are indicated as n.d. in the appropriate tables and graphs. Only
detectable interstitial water metal was included In the calculation of interstitial water toxic
units.
Sediments which caused greater than 24 percent mortality were considered toxic.
Mearns et al. [8] found that sediments which caused greater than 24 percent mortality in
tests with the amphipod Rheooxynius abronius were not consistently classified as
significantly toxic. This criterion is similar to the "80 percent of control survival" criterion
used in the EMAP program [13].
Many of the interstitial and overlying water concentrations discussed herein are
expressed as toxic units. A toxic unit is the measured water concentration divided by the
water-only LC50 concentration for that particular compound for the test organism. For
example, a sediment with an interstitial water concentration equal to the water-only LC50
concentration for the test organism would have one interstitial water toxic unit (IWTU).
When more than one toxic metal is present, IWTUs are calculated as the sum of the toxic
units of the individual metals; e.g., IWTUCd + Nj = (interstitial water cone Cd/LC50Cd) +
(interstitial water cone Ni/LC50Nj). Thus, if interstitial water is the principal source of
metals toxicity, and availability of metals is the same in water of water-only tests and
interstitial water in sediment tests, 50 percent mortality would be expected in sediments
having 1.0 IWTUs. In this document we use < 0.5 IWTUs to indicate sediments unlikely
-------
5-3
to cause significant mortality because on the average water-only LCO and LC50 values
differ by approximately a factor of two [14] and because data in our experiments supports
this value as a break-point between nontoxic and toxic sediments.
Dashed lines are used on all figures to indicate predicted break points in mortality
and chemical concentration. The dashed lines at SEM/AVS = 1.0 indicate the predicted
boundaries between nontoxic sediments and sediments which may be toxic. The dashed
line at 24 percent mortality indicates the demarcation between toxic and nontoxic
sediments, and the dashed lines at 50 percent mortality and IWTU =1.0 indicates the
theoretical mortality at 1.0 IWTU.
Water-Only Tests
Ten-dav static renewal tests were conducted with A. abdita to determine water-only
LC50sfor Cd, Cu, Ni, Pb, and Zn in seawater. The 10-day LC50 values for the water-only
tests were calculated using the trimmed Spearman-Karber method [11]. The LC50 values
from the water-only tests are summarized in Table 5-1. There is no 10-day LC50 value
for cadmium available for Rhepoxvnius hudsoni so the 10-day LC50 value for Ampelisca
abdita (36 //g/L) was used for the calculation of toxic units for this species. This
assumption is reasonable because these amphipods have similar sensitivities; i.e., the 4-
day LC50's for R. hudsoni (640//g/l) and A. abdita (340 jjgtt) differed by less than a factor
of two [12]. We assume that the ten-day LC50s for both species will also be similar.
TABLE 5-1. 10-DAY WATER-ONLY LC50 FOR
AMPELISCA ABDITA
Metal
Cadmium
Copper
Lead
Nickel
Zinc
LC50 U/g/U
36.0
20.5
3,020
2,400
343
95 % Confidence Limits
Not reliable
16.5-25.5
1,980-4,610
2,050 - 2,820
291 - 405
-------
5-4
Spiked sediment tests
Amphipods were exposed to control and metal-spiked sediments in 10-day tests
with continuous renewal of overlying water. In all experiments two sediments of different
AVS concentrations were used: Ninigret Pond {AVS = 1.18-2.25//mol/g) and Long Island
Sound (AVS = 9.72-19.9 //mol/g). In the cadmium test a mixture of these two sediments
was also used (AVS = 4.34//mol/g). The nominal treatments used in most experiments,
expressed as the molar ratio of metal to AVS were 0.0 (control), 0.1, 0.3, 1,3, 10, and
30 (Table 5-2). There were four replicates per treatment in each test: two "biological"
replicates were used to assess mortality, and two "chemical" replicates were used for
metal and AVS analyses of the sediment at test initiation and termination. Twenty (30
in the cadmium test) amphipods were added to each "biological" and the day 10
"chemical" replicate at the start of the test. Interstitial water samples were collected in
diffusion samplers (peepers) from each of these three replicates at the termination of the
experiments.
Sediment Chemistry -
Day 0 Versus Day 10 Chemistry Values -
AVS, SEM, and dry weight sediment chemistry measurements varied somewhat
from day 0 to day 10, but the variation was generally within 20 percent and did not show
a definite time, concentration, or metal-dependent pattern. Therefore, all AVS, SEM, and
dry weight sediment chemistry data will be reported as means of day 0 and day 10 values.
Interstitial Water Metal Versus SEM/AVS
In all the individual and mixed metals experiments the interstitial water metal
concentrations were usually below the limit of detection in sediments with SEM/AVS ratios
below 1.0 (Table 5-2, Figure 5-1). In the cadmium and mixed metals experiments the
values for interstitial water appear high because of the high detection limits in these
-------
TABLE 5-2. SUMMARY OF SEDIMENT CHARACTERISTICS, METAL CONCENTRATIONS, AND AMPHIPOD MORTALITY IN SIX SPIKED SEDIMENT EXPERIMENTS.
« Spiked Nominal Percent Percent Dry Weight of Metal (ug/g) Sum Sum AVS SEM/ Percent II
Metal Sediment Concentration TOC Silt/Clay _, t»h r* »„ Ni Bu|K SEM"" (umol/fl) AVS : IWTU Mortality
Cu Pb Cd Zn U/mol/fl) (umol/ol ||
in
to
0
8
o
en
8
o
en
w
CM
CM
in
*
0
en
00
in
co
00
in
en
00
00
o
O
U
CO
|
•a
a
CJ
in
CO
00
O
o
0
o>
CO
in
8
U)
CM
8
in
CO
in
in
en
00
in
CO
00
in
en
09
m
o
X
d
CO
Zi
|
o
CD'
O
CO
in in in in in in in
-oooin-coo>-cn ^CM
co f en r^
"~r- ^ CMO<0 ^t- *~
incor^oooincoencorven^tcDcocor^oinoocMincMoacoincooocn
incDCMOCMI^CDCD^-CMCnin^-CMCMCaOOO^-OCOCMCMCMi-in^f
cMin'^-dd'-endcodd'-'-coco'dddddcM'uidd'-'t^'oo
CM CM* «-*r~o *- .-CM
«- CO
"~ ^ co *~ *"
CMCMCMCMCMCM CMCMCMCMCM
eo ca O3 . . . . . . CD* t^ en r>-' •- Q* <<> CM CM' CM CM* CM CM' CM in oo* in en •-
CM CD f~ i- O fO *
CM «- CM U)
• « • ooeooooocococoini-u>'~ol)
t- CO i- CO
r- CM
**t*****cncncncncneDcn enenenencn
oooooooooooooooooooooooooooo
*" •• o o d "~ •"ooo'" •~'°d'~ •• " cj d d *"
ZZZXXXXXXtOCOMlOCOMCOZZZZZZZMtOMtOCO
•? •? •?= •? '? '? '? '? 'F Q- D- Q. Q. Q. Q. a D. D. a a a a a S S S m n
in
r-
en *
q t.
in CM
CO •-
CM CO
09 00
co r^
•- o
K 00
t •*
t 00
00 T
z z
in en
t co
00 r-
«- *
CO O
CM en
co en
K *
•» *
en en
en en
en en
o o
•- co
co w
•§ "8
-------
1
x
ui
Q
ID
at
Q
to
X
to
z
£
_j
a
o
tu
a.
O
to
If
-incn i-cnoo
£1
ininininininininin
rxoiixcMixoix'u)ix'|x>u)rx
enooir- »-^-ix
oo
CM
O
CO •" CD CO
O CM ^- CM
QOQOOCMCOQQ
Z Z * Z Z
CO|xcO'T»-CMCD|x«SCDco«-ooO'-eoeM
in in
«- •- co en O
CO |x CO |x
|x en
CM
•'t/J: C^ O) CO O ^f O ^~ U) 0) ^ O) ^™ O t** tO O) 00 f) f) M fO Tj* If) •"• ^ O| r™« f** CO CO O ^™ ^" U) ^™ ^ CO f)
3> oocMcocM«-ooocMincoixcMcMO«-»-inooixco«-cM*eot-in«-O'-cMin«-cMcoco
«.
OOOO'-ftDOOO'-
»- O) O
CO
C^ O) 1^1 *" 00 ^™ ^J" CO O ^ ^ OJ O ^^
f^^CMCOin^COCMCMOcdco'^-CM
en ix
C)dc>C3'-'-COtf»OBU>'*C
>IU§OOO'-rxCOOO<-CMC»COOCO-CM|xCMi-CMinOinoOCMOOO>-CM4-aOCO
CO— i- CM CM fx (O tx CM CMIxm^-
CM in «- T-
:^i co<«-inix^-cMcorxixcocDoooin0>ix^ra>oin>-rom<«-cnixcoco>-ixcocDcooiinincMco
wc
ooor-inininco
in ^*
>-mr«coco
CM in o o
oo CM
cnoocoen^cococncnminoocoiflinocMcDrxixencn CM
QQOQOQOCO<4-'-CMCnOCO<>nCO4>CMCMCM|x4-CMCMOCOCncM|xQQQaOOOCD
ZZZZZZZin^mtfiMiWifiin 'innn^fr^fnenenaimMmrjo^^^^^^^m
cMcaococoincoco'* cMincncor«cncMCMCMcMCMcoco i-
CMCDCOCMU>r- *- - CO * CD CO
i- co co co
«-*
en ix in * CM
•- CM IA •-
Q O
Z Z
eoor-cneo't'-in incoenoorx^cMcM
CMCOCMCMCMCO^CO CMCMCMCMCMCOCOO
ggg§ddddd6ddgg§§g§g§dddo*dnmd§§gg§g
CO O CO t- 00
in ix <• in CM
en ix co ix in
o o
OCO
rxocooooominincococococorx^
(MOinr-^'-COCOCOCOCOCOCOCO
rxcocM^coo^^cDencocoino1
eo'toerirxooi-^oo'i^"-^*-^
«
U
»- co CM * in
«- «- en
corxin^f coixencoco^
QQ»tcpCO'tQCMcO^;|xrxcp
zzoodtMZ^^uiuj^^
in in in w in in
in
^cocoooixcnoin
co O co
sfr
ooooooot**^-*
tt^ooooooot
S OOOOOOO^pt**t***OOOOOOOO*t'»-**^-^'OOOOOOO*
cncncncncncnoi>a> o>AAcncno>cn
-------
e
IU
a.
x
a
IU
M
a.
in
tvs
I?
M
H o
Z o
UJ —
U
O
o
_1
iu
w"
o
(0
S
I-
o
iu
in
u.
O
cc
CM
in
IU
_i
m
c-= in
IS «
CO O
Q-5
£> z
2« co
w1^ o
"S
Wi CO
*
gy «
Q- 1/) CO
C CO
gO CO
• ^^
:
C
o
ll x
ii -
l| o
o
o
c:
o
£ J2
to
c
N
^TB Z
U
in
CM
Q
in
CM
,_
CO
CO
CO
CD
00
?
fv
in
in
00
o
CO
CO
CO
in
CD
CO
CM
^~
Q
z
CM
0
o
CM
^
CO
en
CO
o
X
co_
O
in
-1
c
N
Z
U
u
in
^_
CO
^
co
00
in
«*•
to
CO
in
en
CM
£
CO
rv
CO
O
CM
CO
0
fX
co
CO
CO
CO
fx
CM
^
Q
rv
CO
rv
00
CM
en
CO
CO
O
X
in
-1
c
N
Z
CJ
5
o
o
in
CM
CO
CM
,_
co
rv
-
rv
CO
n
in
in
CO
in
o
00
8
in
S
CM
CO
CO
CO
|v
rx
in
^
»™
Q
CO
rv
rv
CM
CO
CO
CO
CO
o
CO
in
_J
c
IM
z
u
•a
U
8
in
to
in
to
S
CD
co
in
CM
0
rv
co
in
^"
o
CO
J
CM
CM
CM
(0
CO
co
o
rv
CO
o
CO
^>
co
a
CD
co
rv
CO
CM
CO
CO
CO
o
o
in
-1
c
N
Z
u
S
in
o
o
0
o
CM
eS
O
in
0
o
Q
Z
co
CO
CM
CD
CM
O
a
CO
co
o
o
o
in
O
£
o
o
z
z
c
N
Z
3
S
in
£
a
in
0
CO
•-
8
0
o
o
t—
CM
•-
in
CO
CO
CO
^
0
,J.
r—
^_
O
o
in
o
X
6
z
z
c
N
Z
3
S
in
Q
z
'm
i—
0
^f
«-
00
CM
0
CM
o
co
CO
co
m
in
CO
CM
in
a
rv
CM
co
0
O
in
0
X
CO
0
Z
z
c
N
Z
U
•a
U
in
CM
CM
CO
in
in
CM
•_
CO
o
CM
CO
0
CM
CD
O
£
o
CO
n
«™
00
O)
^
«—
a
,j.
co
co
0
O
in
o
*«•
z
z
c
N
Z
o
•o
u
o
CO
co
co
co
CM
CO
00
CM
CO
O
in
CM
in
CM
CO
in
rv
rv
en
CO
CO
O
^~
en
Q
CO
CD
CM
O
O
in
o
CO
z
z
c
N
Z
3
•o
u
8
CM
rv
o
CM
co
en
CM
co
CO
00
CM
CM
CM
U)
CO
|v
00
CM
00
^
CM
00
0
^"
a
00
co
o
CM
0
O
in
0
X
0
z
z
c
N
z
5
S
"S
*—
2
*ff
w
3
(0
C
o
a>
1
CO
e
3
•5
T=
m
c
e
E
B
g
e
|
1
0
o
a
•*
Q
"*•
9
•6 -S
fi ~
P CO
•2 co
W w
3 C
" to
e co
c S
* 5
1 w
C w*
d>
£.J
^ c
=5 „
•I c
« .5
"5 »
ll S
tt
o
O B
CO •*
| ~
C JS
o S
" CO
• -i
• S
1 §
z t
S |
to o
O c-n
"3 1
Isl
>-§ E
• — CO
»lt ^
«» " ^
CO UJ
2 Q to
* * •
-------
5
0
E
3-
JS
o
2
*
1
.g
£
I
£
lUUUUUU"
100000-
10000-
1000-
100-
a
A
A •
•
10-
ll^ •M + A*
O.ir
0.01-
i i i i i
0.001 0.01 0.1 1 10 100 10
SEM/AVS: Cd
5
0
a
o
1
erstitial Wai
IUUUUUU-
100000-
10000-
1000-
100-
1-
0.1-
nm.
C
A
• A
•A
B
A
•
^ AA-IA" A
>*»
-
|
a
€
~®
1
"o
$
0
1
U.UI
0.001 0.01 0.1 1 10 100 10
SEM / AVS: Ni
100000-
10000-
1000-
100-
10-
1
0.1-
nm -
e
A
" A
' A
AA.
M5. . . ,A
IUUUUUU-
100000-
10000-
1000-
100-
10-
1-
0.1-
n m
U.UI
O.C
IfYYYYYI
IUUUUUU
100000-
10000-
1000-
100-
10-
1-
0.1-
n m
U.UI
0.(
b
A
•A
01 0.01 0.1 1 10 100 10
SEM/ AVS: Cu
d
1
I
*$ A. A' *"
301 0.01 0.1 1 10 100 1C
00
00
SEM / AVS: Pb
uuuuu-
10000-
1000-
100-
10-
1-
0.1-
nm.
f
A
A
•
A
^ . -A ;
0.001 0.01
0.1 1 10 100
SEM / AVS: Zn
1000
0.001 0.01 0.1 1 10 100 1000
SEM /AVS: Cd + Cu + Ni + Zn
A US
NIN
Mix
Figure 5-1. Interstitial water metals (IW) concentration (//M/l) as a function of SEM/AVS
ratio in three sediments: Long Island Sound (US), Ninigret Pond (NIN), and a 50/50 mixture
of these two sediments (Mix). Each panel represents data from a separate experiment.
Data from the mixed metals experiment represents the molar sum of cadmium, copper,
nickel, and zinc. Data below the IW detection limit are plotted at one half the detection
limit, indicated by arrows. All IW data in the copper experiment were above the limit of
detection. Data below the SEM detection limit are plotted at SEM/AVS = 0.001.
-------
5-9
experiments. In the cadmium experiment one-half the detection limit of the cadmium
electrode used to measure cadmium ion concentrations was 1.33 //mol/L. For the mixed
metals experiment the sum of one-half of the detection limits of the four metals spiked in
this test was 1.54 ^mol/L. Above an SEM/AVS ratio of 1.0 the interstitial water
concentration increased up to five orders of magnitude with increasing SEM/AVS ratio.
In each experiment there were usually one or more sediments with SEM/AVS ratios of only
slightly greater than 1.0 having interstitial water concentrations below or near detection
limits. This indicates that there are other binding phases in the sediment. In some
sediments spiked with copper, nickel, and a mixture of metals, AVS decreased with
increasing metals concentration (Table 5-2), presumably due to the formation of copper
and nickel sulfides not soluble in the AVS extraction. This underscores the importance of
using SEM rather than total metal in the calculation of metals/AVS ratios.
When the results of all the experiments are plotted together (Figure 5-2} the
relationship between interstitial water concentration and SEM/AVS is confirmed. In most
cases interstitial water concentrations were below the detection limit in sediments with
SEM/AVS < 1.0, increasing with increasing SEM/AVS ratio at SEM/AVS ratios > 1.0.
The relationship between interstitial water concentration and SEM/AVS ratio in the
mixed metals experiment was similar to that in the individual metal experiments when the
molar concentrations of all of the metals are! summed (Figures 5-1f and 5-2). Further
insight into the partitioning of the metals in the interstitial water from the mixed metals
experiment can be gained by plotting the interstitial water concentrations for each
individual metal (Figure 5-3). In the LIS sediment all four metals were below the limit of
detection in treatments with SEM/AVS ratios of 1.25 or lower (Figure 5-3). As the
SEM/AVS ratio of the treatments increased, detectable concentrations of metal began to
appear. The most soluble sulfide (nickel) appeared first and at the highest interstitial
water concentration. As SEM/AVS ratios increased the other metals appeared in order
of their sulfide solubility product constants. The metal with the least soluble sulfide
(copper) appeared last and at the lowest concentration. The relationship between
interstitial water concentration and SEM/AVS ratio in the NIN sediments was similar to that
-------
*
*
. CD
**
o
CD
C/5
UJ
CO
U p
CD
O
O
8
O CD
o o —
8
;c5
n
Q_
z
-K
o
o
J2 > w
2^ ®
CD CD
^
•g 5 I
X O c
E— "53 -c
JQ CD
CO _
m 4-> CO ,.
£ c £ w
® 5 >
!*§~
?5fcl
§ o ro C
o N "° "5.
f- _. -Q
3 2 CD £
1^ c *•• Is
•^ co co ro
CO _, O 4-1
8 III
l = "i
•fe u.* co
S> CD ^
c °£ P *
•B 8^-S
CD C __
Si
111
<"
CD
o
_co
a>
co
CO C Q
C-Q O
1_ f- 4-1
CD C CO C
o
Cl/ioairl)
W CO CD H-
CD fc. CO
fill
V CD C S
in c o ^
£ •= £ «
D SS CD
O)
>
O
.ii' X CD J3
LL CD ~
-------
10000
£ 1000-
V)
O
"o
I
100
10
1
0.1
IW
Metal vs SEM/AVS: SW Mixed Metals - US
LOG
Ksp
Ni -27.98
Zn -28.39
Cd -32.85
Cu -40.94
Ni-*
Cu—>
Zn—*
. Cd—»
0.01
0.01
0.1
1 10
SEM/AVS
100
1000
s_/
tn
o
"o
o
I
.0
1
-2
"c
IW Metal vs SEM/AVS: SW Mixed Metals - NIN
D 10000
o
£ 1000
100
10
1
0.1
0.01
Cd—»
0.01
0.1
1 10
SEM/AVS
100
1000
Figure 5-3. Individual interstitial water metals (/yM/l) concentration in the mixed metals
experiment as a function of SEM/AVS ratio. The top panel represents data from the US
sediment, the bottom panel data from the NIN sediment. Data below the IW detection
limit are plotted at one half the detection limit, indicated by arrows.
-------
5-12
in the LIS sediments (Figure 5-3). In the sediment treatments with SEM/AVS ratios of less
than 1.0 there was no detectable metal in the interstitial water. In one sediment with an
SEM/AVS ratio slightly greater than 1.0 (1.12) there was measurable zinc and cadmium,
but only in low concentrations. In the sediment treatment with the next higher ratio there
was measurable nickel, zinc, and cadmium, with the concentrations decreasing in the that
order. Only in the sediment with the highest SEM/AVS ratio was measurable copper found
in the interstitial water.
Sediment Toxicity
The mortality of amphipods as a function of dry weight metals concentrations
followed a similar pattern in each of the five individual metals and the mixed metals
toxicity tests. Mortality appeared sediment-dependent when plotted on a metals basis
(Figure 5-4). Mortality increased with increasing metals concentration (ug/g dry wt.) for
each sediment, but in each experiment there were treatments in low AVS sediments
(Ninigret Pond) which caused 100 percent mortality at dry weight concentrations which
did not cause appreciable mortality in treatments from the high AVS sediment (Long Island
Sound). Thus, although mortality is concentration dependent for both sediments, the
concentration-response curves do not overlap. Therefore it is not possible to predict
sediment toxicity on the basis of dry weight metals concentration alone (Figure 5-4, Figure
5-5a). Mortality did not appear to be metal-specific when plotted on a molar dry weight
metal basis (Figure 5-5a), indicating that some factor in the sediment was affecting the
toxicity of all five metals similarly. Within one sediment the results from all five metals
were very similar (Figure 5-5b).
Mortality in the individual and mixed metals experiments was sediment independent
when plotted on an SEM/AVS basis (Figure 5-6). Sediments with an SEM/AVS ratio_< 1.0
did not cause mortality significantly different from the control; i.e., greater than 24
percent. In sediments with SEM/AVS > 1.0, mortality increased with increasing
SEM/AVS ratio, although in each experiment there were usually one or two sediments with
SEM/AVS ratios slightly greater than 1.0 (and in one instance 5.8; Table 5-2) which did
-------
£
o
o
100-
80-
60-
40-
20-
a
100-
80-
40-
20-
0.01 0.1 1 10 100 1000
Dry Wt. Metal (umol/g):Cd
0.01 0.1 1 10 100
Dry Wt. Metal (umol/g):Ni
A A
0.01 0.1 1 10 100
Dry Wt. Metal (umol/g): Zn
1000
100-
80-
60-
40-
20-
A A A A
100-
80-
60-
40
20
1000 0.01
100-
80-
60-
40
20
0.01 0.1 1 10 100 1000
Dry Wt. Metal Qjmol/g): Cu
0.1 1 10 100 1000
Dry Wt. Metal (pmol/g): Pb
A A
0.01 0.1 1 10 100 1000
Dry Wt. Metal (pmol/g): Cd + Cu + Ni + Zn
A LIS
NIN
• Mix
Figure 5-4. Percentage mortality of Ampelisca abdita as a function of the sum of the
concentrations of cadmium, copper, lead, nickel, and zinc in//M divalent metal per gram
dry weight sediment in three sediments: Long Island Sound (LIS), Ninigret Pond (NIN), and
a 50/50 mixture of these two sediments (Mix). Each panel represents data from a
separate experiment.
-------
100
80
o 60
o
40
20
0
0.01
A +
A + •
** ''C**
•* if y
'• . ~ jr *i
0.1 1 10
DryWt. Metal Qjmol/g)
Cd A Cu * Ni
Pb
100
1000
Zn
Mix
o
100-
80-
60
40
20
0.01
o
o
-O-O-
coo::-
0.1 1 10
DryWt. Metal(pmol/g)
100
1000
• LJS
o
NIN
Figure 5-5. Percentage mortality of Amoelisca abdita as a function of the sum of the
concentrations of cadmium, copper, lead, nickel, and zinc in //M divalent metal per gram
dry weight sediment. All experiments combined. Upper panel plots data by metal, lower
panel plots data by sediment: Long Island Sound (LIS) and Ninigret Pond (NIN).
-------
Mortality
38
% Mortality
100-
80-
60-
40-
20j
0^
0.0
100-
80-
60-
40-
20-
n.
a m m
*
•
........
i A
01 0.01 0.1 1 10 100 10
SEM / AVS: Cd
c ./A .A A
.... -
100-
80-
60-
40-
20-
0.
i
00 O.C
100-
80-
60-
40-
n^
b -A-A AA
•
M A
01 0.01 0.1 1 10 100 IOC
SEM / AVS: Cu
d A
A
0.001 0.01
0.1 1 10
SEM / AVS: Ni
100
1000
0.001 0.01
0.1 1 10
SEM / AVS: Pb
O
t
O
^—
3?
100-
80-
60-
40-
20-
n-
e B« A
A
P
A ,A A
100 1000
0.001 0.01
0.1 1 10
SEM / AVS: Zn
100
1000
100-
80-
60-
40-
20-
n.
f M A
.- .. .._...........
• A
• "A^ *
0.001 0.01 0.1 1 10 100 1000
SEM / AVS: Cd + Cu + Ni + Zn
LIS
NIN
Mix
Figure 5-6. Percentage mortality of the amphipod, Ampelisca abdita as a function of the
ratio of the sum of the molar concentrations of cadmium, copper, lead, nickel, and zinc
simultaneously extracted (SEM) with acid volatile sulf ide (AVS) to the molar concentration
of AVS (SEM/AVS) in three sediments: Long Island Sound (LIS), Ninigret Pond (NIN), and
a 50/50 mixture of these two sediments (Mix). Each panel represents data from a
separate experiment. Data below the SEM detection limit are plotted at SEM/AVS =
0.001.
-------
5-16
not cause significant mortality. This indicates that there are other binding phases in the
sediment. Thus, it is possible to predict with accuracy which sediments will not be toxic,
and, with less accuracy, which sediments will be toxic. When the results of all of the
experiments are plotted together the mortality of amphipods as a function of SEM/AVS
appears metal and sediment independent for the five individual metals, and for a mixture
of metals (Figure 5-7).
Mortality was not sediment specific when it was plotted against interstitial water
toxic units (IWTU). Sediments with IWTUs of less than 0.5 were not toxic (Figure 5-8).
Sediments with IWTUs of greater than 0.5 were increasingly toxic with increasing IWTU
value. As was the case when mortality is plotted against SEM/AVS and ratios exceeded
1.0, there were usually sediments with IWTU values greater than 0.5 which did not cause
mortality. This was especially true in the range of IWTU values greater than 0.5 but less
than 10.0 (Table 5-2). This indicates that not all of the interstitial water metal is
bioavailable. Thus for both SEM/AVS ratios and IWTUs, sediments likely to be non-toxic
can be predicted with near certainty, but predicting which sediments are likely to be toxic
is less accurate.
As was the case with SEM/AVS ratios, the generality of the relationship between
mortality and IWTU can be seen when the results from all of the experiments are coplotted
(Figure 5-9). This relationship was not metal-specific. Thus the sum of the IWTUs can
be used to make predictions about the toxicity (or lack thereof) of any combination of the
metals tested in these experiments.
When the results of the individual metals and the mixed metals test are taken
together, 98 percent of the 43 sediments with SEM/AVS ratios _< 1.0 were not toxic (i.e.,
caused mortality less than 24 percent). Of the 45 sediments with SEM/AVS ratios > 1.0,
80 percent were toxic. Ninety-four percent of the 52 sediments with IWTU < 0.5 were
not toxic, while 92 percent of the 37 sediments with IWTU >_ 0.5 were toxic. When both
SEM/AVS ratio and IWTU are combined the predictive ability is improved. Ninety-eight
percent of 43 sediments with SEM/AVS ratios <. 1.0 and IWTU < 0.5 were not toxic,
-------
o
*
*
•*
•*
*
^P
^
^^
*^B
*^ **
Iv +
5^2 B *
T*
* • * * *
• "fr ^^ • ^^
' ' * '
^
^ ^^^
• • Tw
v_L
^c*
* • *
*<* ^
-.
•
-
"
O CD O CD CD" 'C
O QO sO *vf f\l
O
O
o
O
O
1
0
CO
"^^
<£
• "*^*»
^
^
^•^B
LU
CO
r—
CD
r
V
O
CD
r—
O
^M^
=8
X
2
^^>
+
c
•
,-J
•
*
D
U
•4
O
*•
03
U
c
U
k_
n
0
E
4.
»^-
o
E
3
in
03
JC
•»-"
*^
o
'+*
CO
(1)
JC
**•
o
•—
U
D
^^
CD
(0
CO
CO
1
CO
CJ
.2
1
"o
>
3
§
to
03
CO
^J
c
o>
U
^J
2 co
w 2
S "Q.
ft) V
iww ^
S co
>»- 4_,
^» C
*= _g
"o o
> V
+^
_. Q}
S 73
0 _
CO 2
LU
§ 0)
^ ^
^ $
in o
co "5
~ JD
•o co
•Is
-d<2
^ CO
S "-
%
U)
o .2
•M
E S ""
Q) '= « O
^ C O Q
._ CO O n
LL O O I'
-------
£
o
100-
80-
40*
2U
rv
0 m m »A
, A
i
'
i . __ __...-...
0.01
0.1
10 100 1000 10000 100000
IWTU: Cd
100-
80-
60-
40-
20-
n.
b A- A* A -A
•
'v.
0.01 0.1
10 100 1000 10000 100000
IWTU: Cu
£
o
tr
0
O^
100-
80-
60-
40-
!i
C A . A- A A
1' .
100-
80-
60-
40-
!i
d AB
• ----.-
i: /
0.01 0.1
10 100 1000 10000 100000
IWTU: Ni
0.01 0.1
10 100 1000 10000 100000
IWTU:Pb
£
o
100-
80-
60-
40-
20^
i
i
e AB • A
A
•
iA AA
0.01 0.1
10 100 1000 10000 100000
IWTU:Zn
100-
80-
60-
*
20 J
n<
f
;
A - A
•
--•
A
0.01 0.1 1 10 100 1000 10000 100000
IWTU: Cd + Cu + Ni + Zn
A
LIS
• NIN
• Mix
Figure 5-8. Percentage mortality of Ampeljsca abdita as a function of interstitial water toxic units in
three sediments: Long Island Sound (LIS), Ninigret Pond (NIN), and a 50/50 mixture of these two
sediments (Mix). In the individual metal experiments IWTU equals the IW concentration of the
individual metal/A, abdita LC50 for that metal. Interstitial water toxic unit (IWTU) for the mixed metals
experiment is the sum of the IW Cd concentration/Cd LC50 for A., abdita + IW Cu concentration/Cu
LC50 + IW Ni concentration/Ni LC50 + IW Zn concentration/Zn LC50. Each panel represents data
from a separate experiment.
-------
CO .ii
+ 03
4
T •
9
$ •
•*
"fee*
^^B
i.
*
"*«
• +
•K "
+
^
* +
*
•4
• •*
^^
*
_ „ i
« ' *
-K
*• *
_
80 O 0 0 C
CO O *sT CM
fWHB
A4-I|D4JOI/\| %
o
8
8
f—
O
0
o
^_ ._
1^
"c
— '
Q °
o x
^ £
k_
CD
.0 |
5
•^—
"t/5
.^- o
•••V
^^
B
O
i —
x
1
+
M
Q_
z
-K
D
0
^
QJ
O
•
=6 .2 ig E
« i 1 s
-- o E
._••«- .5
+2 ™ o «
•F 03 in °-
3 EU 2
O *-* ss
'x _c "^ ^
O += U .
t! o"§S
lili
.5 S "c
~ 21 c o
— O
2 c E +
W -D 3
CO C CO O .
™ 03 03 J*? "~
"8 o .52 2 |,
« C *- ^ "
.8 -B S 5 g
cu 2 .§ £ ^
Eg |§Z
< 0 X 0 ™
^ C 03 C TO
0 8 « 8 "
£" ^ ™ 2 —
W *™ * ^ *i^.
•M 03 E < 03
0 € T3 CO
E JO 03 + ^
0) 3 1 ° 1
2 03 03 O C
S D € Tf .2
CJ 1— 3 o
^i^^§
. +2 5 -2 TJ
? S I- 2 «
10 E ^ c ^
2 'C ~ 03 >
^ • ._j JJ Q
.S5 x "c o "53
U. 03 3 CJ S3
-------
5-20
while 94 percent of 36 sediments with SEM/AVS ratios > 1.0 and IWTU >. 0.5 were
toxic (Table 5-3).
TABLE 5-3. ACCURACY OF PREDICTIONS OF THE TOXICITY OF
SEDIMENTS FROM USING 3. ABDITA SPIKED-SEDIMENT TESTS AND
COMBINED FRESHWATER AND SALTWATER SPIKED-SEDIMENT
TESTS AS A FUNCTION OF SEM/AVS RATIOS, INTERSTITIAL WATER
TOXIC UNITES {IWTUs) AND BOTH SEM/AVS AND IWTUs.
% of Sediment
Study Type
A. abdita
Lab-Spike,
(FW & SW)
Pflrampfpr . Valup •
SEM/AVS <= 1.0
> 1.0
IWTU < 0.5
> 0.5
SEM/AVS,IWTU £ 1.0, < 0.5
> 1.0, s 0.5
SEM/AVS 1.0
IWTU < 0.5
> 0.5
SEM/AVS,IWTU < 1 .0, < 0.5
> 1.0, ^ 0.5
n
43
45
52
37
43
36
92
83
107
77
85
65
Not Toxic
97.7
20.0
94.2
8.1
97.9
5.6
95.7
26.5
93.5
22.1
96.5
12.3
Toxic
2.3
80.0
3.8
91.9
2.1
94.4
4.3
73.5
6.5
77.9
3.5
87.7
Discussion
The results of the amphipod toxicity tests with sediments spiked with metals
indicate that it is not possible to causally predict the toxicity of a sediment using the
concentration of metal on a dry weight basis because the relationship between mortality
and dry weight metals concentration in our tests was sediment specific. In contrast, the
relationships between mortality and SEM/AVS ratio and mortality and interstitial water
concentration or toxicants were demonstrated to be sediment independent. This suggests
that they are the most useful expressions of bioavailable metal for causally predicting
organism response.
-------
5-21
Most of the sediments either caused little or no mortality, or nearly complete
mortality (Figures 5-9 and 5-10). This is a result of the dynamics of metals and AVS in
the sediment, leading to a sharp increase in interstitial water metals concentration when
SEM/AVS ratio exceeds 1.0 and sulfide no longer is a significant binding phase (Figures
5-1 and 5-2). When sufficient sulfide is present in the sediment to bind the metal, little
or no metal is present in the interstitial water [15]. The divalent metals should appear in
the interstitial water in reverse order of the solubilities of their sulfides [12]. Thus we
observed that nickel appeared first in the interstitial water in sediments with SEM/AVS
ratios slightly greater than one, followed by zinc, cadmium, lead, and copper as the
concentration of metals increases relative to that of AVS. When the binding capacity of
the sulfide is exhausted the interstitial water concentrations of metal increase sharply
enough that nearly 100 percent mortality results in most of our test sediments. The effect
is similar to the "throwing of a switch" at SEM/AVS = 1.0. This overwhelming increase
in the interstitial water concentration explains why the chemistry of the anaerobic
sediments controls the toxicity of metals to organisms living in aerobic sediment
microhabitats (e.g., the amphipods living in their burrows in our experiments). It also
explains why the toxicity of different metals in sediments is the same on an SEM/AVS
basis (Figure 5-7) even though their toxicities differs markedly in water-only toxicity tests
(Table 5-1). This sharp increase in interstitial water concentration with increasing
sediment concentration is in contrast to the situation with nonionic organic contaminants,
which are released from the sediment more gradually, primarily as a function, of the KQC
of the compound [16].
When our data are combined with the data available from spiked sediment
experiments in the literature, all data demonstrate that SEM/AVS ratios and interstitial
IWTU's can be used to predict toxicity. Data from freshwater tests using oligochaetes and
snails exposed to sediments spiked with cadmium [7]; and saltwater tests using
polychaetes exposed to sediments spiked with cadmium, copper, lead, nickel, or zinc [4],
using polychaetes exposed to sediments spiked with cadmium or nickel [5], and copepods
exposed to sediments spiked with cadmium [6] all follow the same patterns as our
amphipod results when mortality is plotted against SEM/AVS ratio (Figure 5-10). These
-------
S
o
GO
o
-• »H
T3
u
VI
o
o o
< O
M M
o o
< 0
£ £
Q. Q-
< z
o o
< y
o o
i
-0
U U
o
co o
Q> c
•»-> ._
0) N
CO i.
Q. w
%>-
•~ "3
co c
£ "«
JS —
C E T3
— "O Q)
«0 CO £
.« O o
o.-r a>
i**
fl E
>_ CO O
S€~
CO V
•5'-5
•
•g 2 05
co co •"
^ I CO
*° 7) *"
.li ™
CO (A O
tf <3
in
Q) — «C
0 5 ^
CO < <
S T3 UJ
" 2 w
(U tt
O £ *-
^^ JQ ^?
m » c
-------
5-23
combined data describe tests with six freshwater and saltwater species and sediments
from seven sites having AVS concentrations ranging from 1.9 to 65.7 jjmol/g dry wt and
TOC ranging from 0.15 to 10.6 percent. Mortality in the individual and mixed metals
experiments was sediment independent when plotted on an SEM/AVS basis. Sediments
with an SEM/AVS ratio of less than 1.0 were not toxic. In sediments with SEM/AVS
greater than 1.0, mortality increased with increasing SEM/AVS ratio, but not all sediments
with SEM/AVS ratios of > 1.0 were toxic. This is due, in part, to the presence of other
binding factors.
In addition, organism behavior in a toxicity test can control exposure and limit the
impact of metals in sediments. Many of the sediments which had SEM/AVS ratios > 1.0
but were not toxic were from experiments using the polychaete, Neanthes
araneceodentata. exposed to sediments spiked with cadmium or nickel [5J. This is
especially true of the nontoxic sediments with the highest SEM/AVS ratios. The
polychaetes did not burrow in most of these sediments, and presumably were not fully
exposed to the metals in the sediment (Figure 5-11) and therefore survived in sediments
that would likely otherwise have been toxic {Figure 5-10).
The combined data from all available freshwater and saltwater tests also follow the
same pattern as our saltwater amphipod data when plotted on an IWTU basis (Figure 5-
12). Mortality was not sediment specific when it was plotted against IWTU. Sediments
with IWTUs of less than 0.5 were generally not toxic. Sediments with IWTUs of greater
than 0.5 were increasingly toxic with increasing IWTU value. Here again, as with
SEM/AVS ratios > 1.0 vs mortality, many of the sediments having IWTUs > 0.5 which
were not toxic are likely the result of interstitial water ligands which may reduce the
bioavailability and toxicity of dissolved metals. Polychaete avoidance of otherwise toxic
sediments is also a factor.
With the combined data 96 percent of the 92 of sediments with SEM/AVS ratios
<. 1.0 were not toxic, while 74 percent of the 83 sediments with SEM/AVS ratios > 1.0
were toxic. Ninety-four percent of the 107 sediments with IWTU < 0.5 were not toxic,
-------
1.5 SEM/AVS
0%M
5.7 SEM/AVS
0%M
F
20 SEM/1XAVS
0%M
80 SEM/AVS
90%M
(/)
CONTROL
0%M
0.2 SEM/AVS
0%M
0.4 SEM/AVS
0%M
I
V)
111
CO —.
_ <0
DQ ..
»>
§<
(0 ~~
"5
"S-tiJ
0) (/)
l_ ^_
o u)
W II
c
I!
o>
Q) "'
*?.
w oo
13
II
HI
in
»s 5 o
-------
T> TJ
O O
O
u
O
fi
11
o
0
• ©
o
a s
s g
+-> .=
V N
CO w
"
° 8-1
°.
W CO Q.
.£o
0 .c •-
0) +3 CO
CO (/) CO
" =£ 'c
0 CO =
>. :s o
CO cfl O
s-o r
° C 0)
iro«
® "St >
o> 5 >
£-9-8
. "§..£
CN P •*-
»- co O
-
15.2
o>
il _i
.s>> i
-------
5-26
while 78 percent of the 77 sediments with IWTU >. 0.5 were toxic. When both SEM/AVS
ratio and IWTU are combine.d predictive ability is improved. Ninety-six percent of the 85
sediments with SEM/AVS ratios .< 1.0 and IWTU < 0.5 were not toxic, while 88 percent
of the 65 sediments with SEM/AVS ratios > 1.0 and IWTU_> 0.5 were toxic (Table 5-3).
More IWTU data are available than SEM/AVS data because Green et al. [6] did not
measure AVS in the 10 sediments which they tested. The predictions would be even more
accurate (especially in the elimination of "false positives" (non-toxic sediments with
SEM/AVS > 1.0 and IWTU > 0.5) if data from exposures in which polychaetes avoided
the sediment were not considered [5]. This close relationship between IWTU and sediment
toxicity has been found in an earlier study with cadmium in field sediments [17] as well
as studies with nonionic organic chemicals both in the field [18,19] and in the laboratory
[20,16].
One limitation to the data cited above is that all tests were acute exposures and
these results may not be applicable to chronic exposures. Also metals bioaccumulation
was not measured, except in one case [7]. The applicability of AVS and interstitial water
normalizations to chronic exposures and bioaccumulation in benthic organisms are
discussed elsewhere in this document. Another important limitation to the use of AVS is
that it is only a factor in anaerobic sediments. It did, however, seem to be a controlling
factor in our experiments, with organisms living in aerobic microenvironments. This
anaerobic limitation does not apply to IWTU, of course. The advantages and
disadvantages of each of these prediction methods will be discussed in detail in Chapter
10.
Our results also show that although SEM/AVS and IWTU are useful predictors of
toxicity, there are other important factors as well. The fact that a significant number of
sediments (20 percent) had SEM/AVS ratios of greater than 1.0 but were not toxic
indicates that other binding phases in anaerobic sediments, in addition to AVS, are also
controlling bioavailability. Organic carbon appears to be one of these [21]. While the
SEM/AVS model of bioavailability accurately predicts which sediments will not be toxic,
-------
5-27
a model which utilizes SEM/AVS ratios but incorporates these other variables might more
accurately predict which sediments will be toxic [16].
Similarly, a significant number of sediments with greater than 0.5 IWTUs were not
toxic. Ankley et al. [22] suggested that differences between the hardness of the
interstitial water and that of the water in the water-only tests might effect the accuracy
of prediction of sediment toxicity using IWTUs in fresh water, unless the IWTUs are
hardness-corrected. Further, Green et al. [6] and Ankley et al. [22] hypothesized that
increased DOC in the interstitial water reduced the bioavailability of cadmium in their
sediment exposures, relative to the water-only exposures. Green et al. [6] found that the
LC50 value for cadmium in a water-only exposure was less than 112 that of a pore-water-
only exposure, and less than 1/3 that in pore-water associated with sediments. A
significant improvement in the accuracy of toxicity predictions using IWTUs might be
achieved if DOC binding in the interstitial water is taken into account.
-------
APPENDIX 5A
MATERIALS AND METHODS: AMPHIPOD TESTS
Organism collection and acclimation -
Amoelisca abdita were collected from tidal flats in the Pettaquamscutt (Narrow)
River, a small estuary flowing into Narragansett Bay, Rhode Island. Surface sediment
containing the amphipods was either sieved in the field or transferred to the laboratory
within one half hour and then sieved through a 0.5 mm mesh screen. In the laboratory,
amphipods and amphipod tubes were vigorously sieved in a tub of seawater, then the
sieve was quickly lowered into the water and the amphipods were collected from the
water surface. The amphipods were maintained for three to seven days in the laboratory
in presieved uncontaminated collection site sediment and flowing filtered seawater in 4-
liter glass jars, and acclimated to the test temperature at the rate of 2 to 4°C per day.
During acclimation, amphipods were fed the laboratory-cultured diatom, Phaeodactvlum
tricornutum. ad libitum.
One sediment, Ninigret Pond in the cadmium experiment, was tested using the
amphipod Rhepoxvnius hudsoni. R. hudsoni was collected at Ninigret Pond, Rhode Island,
using collection and acclimation methods similar to those for A. amoelisca. except that R.
hudsoni was washed directly from the sieve into sorting dishes after collection.
Water-only tests -
Ten-day static renewal tests were conducted with A. abdita to determine their
water-only LCBOs for Cd, Cu, Ni, Pb, and Zn in seawater. Animals were exposed, unfed,
to five concentrations of metal and a control, with two replicates per concentration.
Amphipods were exposed in 900 ml_ glass canning jars that contained 800 mL of
water. Acclimated amphipods were sieved from the holding jars, sequentially distributed
to 100 mL plastic cups (10 amphipods per cup), then randomly added to the exposure
chambers. Seventy-five to 100 percent of the water in each replicate was renewed every
-------
5A-2
other day, depending on the experiment. Water was sampled at least once during the test
(usually twice, once near the beginning and once near the end of the test) to determine the
concentration of metal. In some experiments aliquots from the two replicates were pooled
prior to analysis. Exposure chambers were covered with black plastic. The exposure
chambers were checked daily and amphipods which appeared dead were removed and
examined under a dissecting microscope. Live animals were returned to the test and dead
animals were recorded and discarded.
Spiked sediment tests •
Amphipods were exposed to control and metal-spiked sediments in 10-day tests
with continuous renewal of overlying water. In all experiments two sediments of different
AVS concentrations were used: Ninigret Pond (AVS = 1.18 to 2.25 //M/g) and Long
Island Sound (AVS = 9.72 to 19.9//M/g). in the cadmium test a mixture of these two
sediments was also used (AVS = 4.34 //M/g). The nominal treatments used in most
experiments, expressed as the molar ratio of metal to AVS were 0.0 (control), 0.1, 0.3,
1, 3, 10, and 30 (Table 1). There were four replicates per treatment in each test: two
"biological" replicates were used to assess mortality, and two "chemical" replicates were
used for metal and AVS analyses of the sediment at test initiation and termination.
Twenty (30 in the cadmium test) amphipods were added to each "biological" and the day
10 "chemical" replicate at the start of the test. Interstitial water samples were collected
in diffusion samplers (peepers) from each of these three replicates at the termination of the
experiments.
The Long Island Sound (LIS) sediment was collected from an uncontaminated site
in central Long Island Sound (40°7.95'N and 72°52.7'W) with a Smith-Mclntyre grab
sampler, returned to the laboratory, press sieved wet through a 2 mm mesh stainless steel
screen, homogenized and stored at4°C. There were two separate collections of LIS (LIS1
and LIS2) sediment. The percent total organic carbon for LIS1 was 0.88, for LIS2 it was
0.99. The grain size composition of LIS1 was 5 percent sand, 71 percent silt, and 24
percent clay. Grain size data are not available for LIS2, but it was of similar composition.
-------
5A-3
Sediment was also collected in Ninigret Pond (NIN), Charlestown, Rhode Island.
The upper few inches of sediment were collected with a shovel, returned'to the laboratory,
sieved wet through a 2 mm stainless steel screen, rinsed several times to remove high-
organic fine particles, homogenized and stored at 4°C. There were two collections made
in Ninigret Pond, from two different sites (NIN1 and NIN2). Both sediments had TOC
values of 0.15 percent and were 100 percent sand (Table 1). NIN2 was made up of
slightly finer sand, most of which would pass through a 0.5 mm sieve. Most of the NIN1
sediment was retained on a 0.5 mm sieve.
Sediments were spiked with metal chloride or nitrate salts in glass, 1 gallon jars.
Methods differed slightly from experiment to experiment but typically we added 1200 mL
of wet sediment to 2000 ml of 20°C seawater that contained the desired weight of metal
chloride. The spiked sediments were stirred with a nylon stirrer attached to an electric drill
until homogeneous, then the overlying air in each jar was replaced with nitrogen and the
jars were capped and rolled for 1 hour. Jars of sediment were held at 20° C for 8 to 10
days before the start of the test. We siphoned the water and any precipitate off the
sediment surface and rehomogenized the sediment before adding it to the exposure
chambers.
The exposure chambers were 900 mL glass canning jars, each with a 1.3 cm
diameter overflow hole (covered with 400-micron Nitex® mesh) 11.7 cm from the bottom
of the jar. Each jar contained 200 mL of sediment and held about 600 mL of sea water
over the sediment. Each jar was covered with a 8 cm diameter glass Carolina dish with
a 17 mm diameter hole for the seawater delivery tube and air line consisting of a 2 mL
glass pipette. We positioned the water delivery and air lines so that the sediment was not
disturbed.
Diffusion samplers [23,24]" peepers", were constructed from polyethylene vials (21
mm high, 20 mm diameter, 5 mL capacity). A 1.6 cm hole was cut in the cap,
polycarbonate membrane (1 micron) was placed over the open end of the vial and the cap
replaced under water so that the sampler was filled with 20°C, 30 ppt salinity water at
-------
5A-4
the start of the test. A 21.5 cm long nylon strap was attached to the vial to serve as a
handle to facilitate handling and removal of the diffusion sampler from the exposure
chamber. For the mixed metals experiment each diffusion sampler consisted of two vials
attached back-to-back to double the volume of interstitial water collected. (The peeper
design used in the cadmium experiment was different and is described in Di Toro et al.,
[121.
Sand-filtered Narragansett Bay water, heated to 20°C ± 1 °C, with a mean salinity
of 30 ppt (28 to 34 ppt) was used in the experiments. Seawater flowed into each
exposure chamber from a distribution system consisting of chambers with self-priming
siphons and splitter chambers. Flow rate for each exposure chamber was approximately
28 to 35 volume additions per day (except in the cadmium experiment in which it was
approximately 10 volume additions per day). Exposure chambers were placed in 20°C
water baths to maintain temperature. The exposure chambers were kept under constant
light to help keep the amphipods burrowed into the sediment.
The test was started by placing a diffusion sampler in each exposure chamber and
adding 200 mL of sediment, to just cover the diffusion sampler. Seawater was allowed
to flow through the chambers for 1 day. Amphipods were removed from the holding
containers as described above, distributed sequentially to 100 mL plastic cups until there
were 20 amphipods per cup (30 in the cadmium experiment), then one cup of amphipods
was added randomly to each exposure chamber. The seawater delivery system was
turned off for 1 hour and any amphipods that had not burrowed into the sediment in that
time were replaced, except in those replicates where there was an obvious dose response
(i.e. where there were a greater than average number of unburrowed individuals in both
replicates). Samples of sediment were taken from the day zero chemistry replicates for
metals and AVS analyses. All but about 1 cm of overlying water was removed from each
day zero chemistry replicate with a vacuum pump and pipette tip. The sediment and a
small amount of remaining seawater was homogenized with a stainless steel spatula.
Approximately half the sediment was placed in an acid-stripped polyethylene jar for acid-
extractable metals analysis, while the remainder was poured into a 100 mL polyethylene
-------
5A-5
specimen cup for AVS and SEM analysis. Each jar was capped and the samples held in
the dark at 4°C until analysis.
The experimental chambers were checked daily and amphipods which appeared
dead were removed and examined under a dissecting microscope. Live animals were
returned to the test, dead animals were recorded and discarded. The volume of water
delivered to each exposure container was measured before each test and the total flow
rate to the system was measured and adjusted daily. Temperature of the water bath and
salinity of the incoming seawater was measured daily. The overlying water in each
biological replicate was sampled for metal concentration at least once near the beginning
and near the end of each test. In some tests the samples from the two replicates were
pooled. Each overlying water sample was placed in an acid-stripped 7 ml_ polyethylene
vial and acidified with 50^1 of concentrated nitric acid (pH<_1).
At the end of the test the diffusion samplers were carefully removed from each
replicate. Any sediment remaining on the cap or membrane portion of the sampler was
rinsed off using clean seawater. The membrane was then punctured with an acid-stripped
5 mL disposable pipette tip and the contents of the sampler removed by pipette. The
interstitial water collected from each diffusion sampler was added to an acid-stripped 7 mL
polyethylene vial, acidified with 50 p\ of concentrated nitric acid (pH <_1) and stored for
metals analysis. The sediment from the "chemical" replicates was sampled for metals and
AVS content as described for the day zero chemistry replicates. The contents of each
amphipod "biological" replicate were sieved through a 0.5 mm screen. Material retained
on the sieve was examined immediately or preserved with Rose Bengal stain for later
sorting. Amphipods were counted and any missing animals were assumed to have died
and decomposed. Any replicates in which 10 percent or more amphipods were not found
were recounted by another investigator as a QA check. '*
-------
5A-6
Chemical analyses -
Sediment samples were analyzed for AVS by a cold-acid purge- and-trap technique
described by Di Toro et al. [12,2]. SEM analyses for the copper, nickel and zinc
experiments were performed using the graphite furnace AA. SEM analyses for the lead
and the mixed metals experiments were performed using inductively coupled plasma
emission spectrometry (ICP). SEM was not measured in the cadmium experiment because
the importance of SEM vs total metal was not understood at that time. However,
cadmium does not form sulfides which are insoluble in the AVS procedure [2] so acid-
extractable and SEM cadmium concentrations are interchangeable. SEM for only the metal
under study was measured in the individual chemical experiments. However, the sum of
the SEM for all of the cationic metals is only 3.2 //M/g for LIS, and 0.081 //M/g for NIN
and thus of little importance in the SEM/AVS ratio for the level of metal spiking used.
To allow for comparisons with other metals toxicity studies, acid-extractable metals
analyses were also performed. For this analysis, metals were extracted from freeze-dried
sediments by ultrasonic agitation with 2 M cold nitric acid (50 mL to 5 g wet sediment)
and the extracted metals separated from the sediment residue by centrifugation. The
resultant solution was analyzed by ICP.
The acidified interstitial and overlying waters were analyzed for trace metals by ICP.
The interstitial water samples from the mixed metals experiment were diluted fivefold with
2 M HNO3 in order to provide sufficient solution for analysis. The interstitial water
samples from the copper experiment were also analyzed using the graphite furnace. The
interstitial waters from the cadmium experiment were analyzed using a cadmium ion-
specific electrode. Total cadmium was estimated by multiplying the Cd2+ measured by
the electrode by 20 (Di Toro et al., 1990), which is the ratio of total Cd to Cd2+ in
seawater.
-------
REFERENCES
1. Pesch, C.E., Hansen, D.J., Boothman, W.S., Berry, W.J., and Mahony, J.D. The
role of acid-volatile sulfide and interstitial water metal concentrations in determining
bioavailability of cadmium and nickel from contaminated sediments to the
polychaete, Neanthes arenaceodentata. Environ. Toxicol. Chem. In Press.
2. Ankley, G.T., Phipps, G.L., Leonard, E.N., Benoit, D.A., Mattson, V.R., Kosian,
P.A., Cotter, A.M., Dierkes, J.R. Hansen, D.J., and Mahony, J.D. 1991. Acid
volatile sulfide as a factor mediating cadmium and nickel bioavailability in
contaminated sediments. Environ. Toxicol. Chem. 10:1299-1307.
3. Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Carlson, A.R., and Ankley,
G.T. 1992. Acid volatile sulfide predicts the acute toxicity of cadmium and nickel
in sediments. Environ. Sci. Techno). 26:96-101.
4. Berry, W.J., Robson, D., Hansen, D.J., Boothman, W.S., Mahony, J.D. 1994.
Predicting the toxicity of metals-spiked laboratory sediments using acid volatile
sulfide and interstitial water normalizations. Manuscript.
5. Ankley, G.T., Mattson, V.R., Leonard, E.N., West, C.W., and Bennett, J.L. 1993.
Predicting the acute toxicity of copper in freshwater sediments: Evaluation of the
role of acid volatile sulfide. Environ. Toxicol. Chem. 12:315-320.
6. Ma, D. and Zhang, F. 1988. Lead, zinc and cadmium distributions in different
geochemical phases of surface sediments from Jinzhou Bay. Acta Scientiae
Circumstantiae. 8:49-55.
7. Samant, H.S., Doe, K.G., and Vaidya, O.C. 1990. An integrated chemical and
biological study of the bioavailability of metals in sediments from two contaminated
harbors in New Brunswick. Canada Sci. Total Environ. 96:253-268.
-------
8. Uthe, J.F., Chou, C.L., and Robinson, D. 1980. Cadmium pollution of Belledune
Harbor, New Brunswick, Canada. Canadian Technical Report of Fisheries and
Aquatic Sciences. No. 963. J.F. Uthe and V. Zitko, Eds. pp. 65-71.
9. Ray, S., McLeese, D.W., Metcalfe, C.D., Burridge, L.E., and Waiwood, B.A. 1980.
Cadmium Pollution of Belledune Harbor, New Brunswick, Canada. Canadian
Technical Report of Fisheries and Aquatic Sciences. No. 963. J.F. Uthe and V.
Zitko Eds. pp. 11-34.
10. Bower, P.M., Simpson, H.C., Williams, H.C., and Li, Y.H.. 1978. Heavy metals in
the sediments of Foundry Cove, Cold Spring, NY. Environ. Sci. Technol. 12:683-
687.
11. Bieri, R., Bricker, 0., Byrne, R., Diaz, R., Helz, G., Hill, J., Huggett, R., Kerkin, R.,
Nichols, M., Reinharz, E., Shaffner, L, Wilding, D., and Strobel, C.. 1982. Toxic
substances. In: Chesapeake Bay Program Technical Studies: A Synthesis. E.G.
Macalaster, D.A. Barker and M. Kasper (Eds.). U.S. EPA, Washington D.C. pp.
263-375.
12. Pinkney, A.E., Gowda, G., and Rzemien, E. 1991. Sediment toxicity testing of the
Baltimore Harbor and C & D Canal approach channels with the amphipod,
Leotocheirus olumulosus. Versar Inc. report to Dr. Peter Dunbar, Maryland
Department of Natural Resources, MD. October 1991. 17 pp.
13. Green, A.S., Chandler, G.T., and Blood, E.R. 1993. Aqueous-, pore-water- and
sediment-phase cadmium: Toxicity relationships for a meiobenthic copepod.
Environ. Toxicol. Chem. 12:1497-1506.
14. Allen, H.E., Gongmin, F., and Deng, B. 1993. Analysis of acid-volatile sulfide
(AVS) and simultaneously extracted metals (SEM) for estimation of potential
toxicity in aquatic sediments. Environ. Toxicol. Chem. 12:1441-1453.
-------
CHAPTER 6
FIELD COLLECTED SAMPLES
The objective of this chapter is to further demonstrate the utility of interstitial water
concentrations of metals and sediment concentrations normalized based on SEM/AVS
ratios to explain the bioavailability of sediment-associated metals to benthic organisms.
The sediments examined in this chapter are all field collected from locations with high
metals concentrations. The first part of this chapter presents previously unpublished data
on the relationship between total metal concentrations, interstitial metal concentrations
and SEM/AVS ratios, and toxicity to the saltwater amphipod (Ampelisca abdita) exposed
to sediments from five marine sites located in Maryland; Massachusetts; New York; New
Brunswick, Canada and Liaoning Province, China together with previously published results
using New York sediments with a saltwater polychaete (Neanthes arenaceodentata) [1].
Next, data are presented on these relationships with a freshwater amphipod (Hvalella
azteca) and an oligochete (Lumbriculus varieaatus) exposed to sediments from four field
locations. Data from locations in New York, Michigan, and Washington have been
published previously [2,3] while those from Missouri are new. All are herein analyzed
collectively. Finally, this chapter combines results from all experiments using field-
collected saltwater and freshwater sediments with those from all available laboratory
spiked-sediment tests using a variety of saltwater and freshwater species [4].
Methods for sediment collection, storage and handling, chemical analyses, and
toxicity testing for the saltwater amphopod (A. abdita) exposed to five saltwater sediments
can be found in Appendix. Methods used to collect, store, and handle freshwater
sediments and test sediments with the amphipod, H.. azteca. have been described for
samples from Steilacoom Lake, Washington, and Keweenaw Waterway, Michigan, by
Ankley et al. [5] and for H. azteca and the oligochaete worm, .L varieaatus. with
sediments from Foundry Cove, New York by Ankley et al. [2]. These same procedures
were also used with sediments from Turkey Creek, Missouri. General biological and
chemical procedures, as well as the conceptual experimental design, were essentially the
same for saltwater and freshwater tests, except that for freshwater tests bulk metals
-------
6-2
analyses were not performed and interstitial water was extracted by centrif ugation instead
of diffusional samplers.
Results Saltwater Field Sites
Description of Field Sites and Toxicity Test Results
Jinzhou Bay is located in the northeastern quadrant of the Bohai Sea, China (Figure
6-1). It has an area of about 150 km2, including 62 km2 of tideflats, with an average
depth of 3.5m [6]. A zinc smelter located near the mouth of the Wuli River is the largest
source of metals to the bay, although other industrial discharges are contributors.
Sediments for this study were collected from seven locations along a 30 km transect from
the river to the northeastern portion of the bay. Total concentrations of divalent metals
in sediments collected ranged from 261 to 36,200 //g/g dry weight (Table 6-1). Zinc
constituted between 78.5 and 86.5 percent of the total. Sediments also contained low
concentrations of PAHs {< 12//g/g for individual PAHs), PCBs (< 0.0'3 //g/g for individual
congeners) and chlorinated pesticides « 0.03 //g/g for any individual pesticide).
Concentrations of TOC ranged from 0.11 to 11.5 percent, AVS 3.0 to 126 //mol/g, SEM
2.9 to 374 //mol/g and SEM/AVS ratios from 0.51 to 8.36. The sum of the interstitial
water toxic units (IWTU) for the five divalent metals ranged from no metal detected
«0.01) to 0.58. The four sediments with the highest metals concentrations were toxic
(> 24 percent mortality) to A., abdita. However, only the most contaminated sediment
contained greater than 0.5 IWTU and had an SEM/AVS ratio > 1.0, which suggests that
metals may not be principal cause of the toxicity observed in the other three sediments
(Figures 6-2 and 6-3).
Belledune Harbor, which receives outfalls from a lead smelter and fertilizer plant, is
located in the southwestern portion of Chaleur Bay, New Brunswick, Canada (Figure 6-1).
Harbor sediments are particularly enriched, relative to adjacent areas, in concentrations of
cadmium, lead, and zinc; other metals are somewhat elevated [7]. The closure of the
lobster fishery due to the elevation of cadmium concentrations in algae, snails, mussels,
-------
Om 500m tOOOm
\ -14
\ Patapsco River
Chaleur Bay
Belledune, N.B
Belledune Harbour, N.B.
Fertilizer
Outfall
0 250 500 750
Meters
Foundry Cove, NY
Battery
Plant
1 2 Source
16
Foundry
Brook
100m
MA
Figure 6-1. Location of field sites and
stations sampled in Jinzhou Bay, China;
Belledune Harbor, New Brunswick,
Canada; Bear Creek, Maryland; Foundry
Cove, New York and a salt marsh in
Massachusetts.
-------
TABLE 6-1.
Station
Percent Percent
TOC Sat/day Cd
Bulk Divalent Metals fcig/0)
Oi N!
Pb Zn
Sum
fontA ;
SEM
imol/g
AVS
fimo\/g
SEM/
AVS
IWTU
Percent
Mortality
. : Jinzhou Bay. China : . . :
NJAM'
1
2
3
4
5
6
7
*
11.5
2.00
0.37
o.eo
0.26
0.11
0.17
• *•
• * •
»»*
• • •
*«*
• ••
• • •
*••
0.10
303
151
9.13
41.4
2.60
8.40
4.80
13.0 2.30
2010 31.0
295 24.2
39.4 10.2
46.8 17.4
6.40 8.60
11.8 9.40
13.0 7.40
32.0 67.0
5410 28400
608 4440
113 737
142 1320
22.2 221
17.0 300
17.6 239
1.40
496
80.5
81.1
84.2
84.7
86.5
84.8
1.42
374
63.5
10.8
18.5
2.99
3.85
2.92
12.2
44.7
126
17.8
36.6
3.02
5.42
3.56
0.11
8.36
0.51
0.61
0.51
0.99
0.71
0.82
ND
0.58
NO
ND
0.17
0.03
ND
ND
2.0
100
90.0
37.5
30.0
2.5
7.5
2.5
: / Beltedune Harbor, N.B. . ','".'
US'
1
2
3
4
5
6
7
e
9
10
US'
1
2
3
4
5
6
7
8
9
10
11
12
13
14
. IS
16
US'
1
2
3
4
5
6
7
8
9
10
11
12
13
14
US'
US'
REF'
REF'
0.99
0.98
.29
.08
.62
.20
.12
.10
0.73
0.87
0.92
0.88
10.2
5.20
13.1
8.13
9.37
5.03
0.79
13.6
5.82
10.9
0.55
16.4
14.6
7.18
4.76
1.45
3.89
7.10
7.38
5.75
5.47
6.15
3.32
0.13
5.19
4.40
0.17
4.61
0.16
4.19
3.14
0.99
0.99
1.24
1.24
94
• •
»•
• •
• •
• *
• •
**
• •
• •
• *
94
• *«
• *•
***
*«•
• ••
* • •
• » •
• ••
• • •
* • •
• »•
• • •
• • •
*• •
*• •
• • •
87
99
97
97
96
80
58
7
97
93
6
98
4
94
91
94
94
17
17
0.90
9.68
11.5
15.0
12.7
8.01
6.76
7.37
1.22
1.94
2.04
0.40
38900
5920
5950
9520
13100
5500
66.1
522
6320
18.6
35.2
163
88.5
363
20.6
1O.1
0.77
8.79
10.0
5.36
4.84
3.45
4.82
0.00
5.82
4.19
0.17
2.61
0.17
1.71
1.34
0.00
0.00
0.42
0.34
45.8 20.9
50.8 33.2
101 38.0
104 39.7
85.5 38.8
57.6 37.3
49.3 31.7
68.5 31.2
15.5 28.6
19.5 27.7
21.9 33.3
56.3 26.5
143 31500.0
87.3 5180.0
81.4 4160.0
106 3700.0
116 7670.0
101 2340.0
28.4 60.6
66.9 386
74.3 3500.0
31.3 38.5
18.1 45.4
57.6 137
44.5 92.1
104 227
92.7 28.9
26.2 17.8
52.6 27.6
206 63.7
228 62.6
266 60.0
207 55.7
151 38.5
191 49.9
3.18 2.01
254 46.8
241 52.8
9.40 4.87
140 50.7
41.4 2.84
139 46.7
96.9 39.0
55.7 25.2
61.5 27.5
9.93 4.49
6.62 3.11
31.7 122
498 401
955 784
1140.0 897
906 874
539 681
463 503
689 843
94.5 137
131 178
149 192
Foundry Cove, NY
35.1 160
194 403
157 297
92.9 278
135 313
156 356
357 303
10.0 79.5
98.3 219
113 246
27.2 101
6.2 65.4
87.7 231
47.9 142
127 317
177 234
48.2 124
Bear Creek, MD
12.9 141
195 158O
212 17OO
209 1140
175 958
162 567
173 1000
3.0 35.6
250 1110
274 978
9.4 69.3
162 617
7.4 42.7
128 459
88.5 346
Salt Marsh, MA
35.1 142
38.8 164
9.5 28.1
6.8 22.8
3.09
90.6
92.1
92.8
92.8
91.6
91.7
93.5
83.7
86.3
74.7
3.95
885
146
129
154
254
96.3
3.32
16.0
120
2.99
2.39
8.62
5.44
14.1
6.57
2.93
3.52
29.3
31.7
23.7
19.7
12.5
20.0
0.64
23.1
21.0
1.34
13.3
1.39
10.6
7.90
3.63
4.12
0.71
0.54
2.70
6.65
17.3
18.2
16.3
11.1
10.2
16.8
1.96
3.16
1.87
0.24
778
93.4
105
136
168
52.2
0.67
8.64
86.4
1.27
0.42
1.91
1.94
5.36
0.56
0.20
2.82
28.3
30.6
20.2
17.4
17.1
16.5
0.63
22.9
19.1
1.25
11.8
0.74
9.84
6.71
X
2.82
2.52
0.41
0.38
16.8
27.2
80.3
102
96.6
47.4
38.6
56.1
5.54
16.7
11.3
12.0
5.60
18.0
12.2
26.8
64.6
12.5
0.44
20.2
24.7
2.62
0.41
0.40
0.69
37.1
13.1
1.38
9.75
269
304
76.1
70.1
45.3
46.6
0.40
147
89.3
0.45
50.0
0.40
7.20
0,40
15.5
16.1
13.8
11.0
0.16
0.24
0.22
0.18
0.17
0.23
0.26
0.30
0.35
0.19
0.17
0.02
139
5.17
8.65
5.10
2.59
4.19
1.55
0.43
3.50
0.48
1.03
4.83
2.8O
0.14
0.04
0.15
0.29
0.11
0.10
0.27
0.25
0.38
0.35
1.58
0.16
0.21
2.78
0.24
1.85
1.37
16.8
0.18
0.16
O.O3
0.03
ND
O.O4
0.06
0.18
0.07
0.02
O.O5
0.62
ND
0.06
ND
ND
43.5
9.21
3.33
2.16
1.81
1.07
0.42
0.45
1.48
0.30
0.30
1.42
0.44
1.11
1.60
ND
0.03
0.03
0.03
0.03
0.03
0.03
0.03
0.04
0.02
0.03
0.03
0.03
0.02
0.03
0.03
ND
ND
ND
0.06
10.0
7.5
12.5
12.5
12.5
15.0
12.5
12.5
7.5
17.5
5.0
10.0
82.5
32.5
52.5
20.0
80.0
35.0
17.5
17.5
15.0
7.5
17.5
12.5
15.0
12.5
17.5
20.0
0.0
82.5
95.0
85.0
85.0
67.5
40.0
5.0
92.5
12.5
5.0
2.5
5.0
17.5
2.5
5.0
0.0
12.5
7.5
-------
TABLE 6.
(continued)
Station
REF-
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Percent
TOC
1.47
3.22
1.56
0.36
0.48
2.00
2.28
2.49
2.55
0.61
1.98
3.54
4.39
1.13
2.74
2.45
0.51
1.18
3.13
0.63
2.03
0.13
Percent
Silt/Clay
11
17
4
2
3
62
21
24
49
4
8
61
2
16
26
11
33
12-
5
13
28
5
17
2
- Bulk Divalent Metals
Cd
0.46
2.07
0.56
0.00
0.00
2.02
5.77
1.60
1.03
0.06
0.44
7.74
0.00
1.67
2.34
' 0.68
0.00
2.58
0.60
0.81
0.37
0.47
0.54
0.26
Cu
15.2
147
144
128
32.4
300
155
823
1420
152
359
851
111
523
572
236
639
348
184
179
47.5
93.2
88.7
29.8
Nl Pb Zn
3.99 8.3 28.4
25.0 41O 517
48.6 66.9 398
9.18 87.6 65.6
5.94 28.1 32.9
36.5 95.7 463
108 61.0 1100
113 143 930
72.1 304 1480
12.0 63.4 236
29.0 137 456
31.1 128 2310
14.8 65.6 239
43.2 151 545
42.9 192 629
20.0 62.6 272
28.0 226 685
28.1 97.5 410
11.6 38.7 172
12.5 52.4 188
11.5 28.4 86.0
7.79 19.9 103
9.51 29.4 152
4.26 6.2 41.9
• Reference sediment! from Long Island Sound (LISI, lower Narragansett Bay
•• NO = No detectable metal.
Sum
fnnol
0.79
12.6
9.50
3.59
1.25
12.9
21.4
29.8
47.8
6.51
13.7
49.9
5.97
18.0
20.3
8.52
22.1
12.7
5.91
6.16
2.39
3.27
4.03
1.21
: SEM
: fimollQ
0.50
8.39
6.36
1.30
0.95
9.60
17.8
25.0
31.7
3.00
7.05
23.0
1.41
11.4
13.9
6.08
16.2
7.20
4.57
3.72
1.71
2.08
2.52
0.73
AVS
fimoVy
4.00
86.4
1.42
0.44
0.59
419
12.4
16.6
69.1
0.50
11.5
14.1
1.32
85.5
19.1
18.6
2.35
21.8
5.86
38.9
11.7
18.0
18.0
3.24
(NJAM) or a clean site nearby
SEM/
AVS
0.13
0.1O
4.49
3.00
1.60
0.02
1.44
1.51
0.46
6.00
0.61
1.63
1.07
0.13
0.73
0.33
6.90
0.33
0.8
0.10
0.15
0.12
0.14
0.23
(REF).
IWTU
ND
0.30
1.00
0.14
0.22
ND
0.19
ND
ND
0.12
ND
0.64
0.10
O.O9
0.25
0.11
0.11
0.10
0.07
ND
ND
ND
ND
0.07
Percent
Mortality
2.5
12.5
40.0
12.5
7.5
17.5
0.0
10.0
22.5
5.0
15.0
7.5
5.0
17.5
10.0
5.0
7.5
5.0
12.5
10.0
7.5
5.0
2.5
2.5
-------
JINZHOUGkA) BELLEDUNEfAA)
100-r
80-
£» 60-
co
g 40
5
3? 20
o-
A
•
A
A
4
A Ji
100-r
80-
60-
40
20
0
jL
ff*
0.01 0.1 1 10 100 1000 0.01 0.1 1 10 100 1000
FOUNDRY COVE (AjaJ FOUNDRY COVE (hLaJ
100-
80-
S» «~
*•• fill ~
i 40-
a*
* 20-
o-
A A
... ,. . . A
4
A ^ A AA M^
A A A ^^
100 1
80
60-
40-
20
0^
A
A A ^A ^
A A Jl4
0.01 0.1 1 10 100 1000 6.01 0.1 1 10 100 1000
BEAR CREEK (A&) SALT MARSH (A^
100-
80-
.
-i" 60-
o 40-
^ 20-
0
A^
A A
A
A
A *
V AA A
^»
100
80
60
40
20
0
• -
A
A *i \4A
4BU^ ^? •
0.01 0.1 1 10 100 1000 0.01 0.1 1 10 100 100C
SEM/AVS SEM/AVS
Figure 6-2. - Percent mortality of the amphipod (Ampelisca abdita. A.a.) and polychaete
(Neanthes arenaceodentata. N.a.) as a function of SEM/AVS ratios in sediments from
Jinzhou Bay, China; Belledune Harbor, New Brunswick, Canada; Foundry Cove, New York;
Bear Creek, Maryland and a salt marsh in Massachusetts.
-------
JINZHOU (A.a.)
BELLEDUNE
'ffi
I
^
cfc
^
"3
o
s
s*
•
^
•-=
i
Figure
luu-
80
60
40;
20
i
i
A
^ A
IUU
80-
60-
40-
20 :A" A A A
A l|
0.01 0.1 1 10 100 0.01 0.1 1 10 100
FOUNDRY COVE (A^aJ FOUNDRY COVE (fcLaJ
100-
80-
60-
40-
20 ^
i
o-
A A
A
A A
i A (
100 -i
80-
60-
40-
20-
oi
>;;.':"".
I— A — A11 A
0.01 0.1 1 10 100 0.01 0.1 1 10 100
BEAR CREEK (A&) SALT MARSH (A.a.)
100-
80-
60-
40-
20-
0
1
. 41
A
• A
f
4^
^*
100 -]
80
60
40
A
20r">^>".
ol
E *V A F ^ '.
& A
0.01 0.1 1 10 100 0.01 0.1 1 10 100
Interstitial Water Toxic Units Interstital Water Toxic Units
6-3. - Percent mortality of the amphipod (Ampelisca abdita. A.a.) and polychaete
(Neanthes arenaceodentata. N.a.) as a function of interstitial water toxic units (IWTU) of
cadmium, copper, nickel, and zinc in sediments from Jinzhou Bay, China; Belledune
Harbor, New Brunswick, Canada; Foundry Cove, New York; Bear Creek, Maryland and a
salt marsh in Massachusetts.
-------
6-8
scallops, barnacles, crabs, and lobsters has been of particular concern [8,9]. Sediments
for our study were collected by Ponar grab from 10 stations; 7 inside and 3 outside the
harbor. Total concentrations of divalent metal in these sediments ranged from 277 to
2,200 yt/g/g dry weight, with 74.7 to 93.5 percent of the total consisting of lead and zinc
(Table 6-1). Concentrations of TOC ranged from 0.73 to 1.62 percent, AVS 5.5 to 102
//mol/g, and SEM 1.9 to 18.2/;mol/g with SEM/AVS ratios of 0.17 to 0.35. The sum of
the interstitial water toxic units ranged from <0.01 to 0.62. None of the sediments were
toxic (<_ 24 percent mortality) to A., abdita as would be predicted based upon SEM/AVS
ratios and IWTUs (Figures 6-2 and 6-3).
Foundry Cove, New York is located on the upper tidal reach (salinities 0 to 6 mg/kg)
of the Hudson River immediately south of Cold Spring, New York (Figure 6-1). A battery
plant was the principal source of the approximately equimolar concentrations of cadmium
and nickel in the sediments; smaller amounts of cobalt were also discharged [101.
Sediments for our study were collected by shovel or Ponar grab from 16 stations in East
Foundry Cove. Total concentrations of divalent metal in our sediments ranged from 170
to 71,200//g/g dry weight with cadmium plus nickel accounting for up to 99.0 percent
of the metal measured in the most contaminated sediments (Table 6-1). Concentrations
of TOC ranged from 0.55 to 16.4 percent, with many sediments consisting principally of
partially decayed marsh vegetation. Concentrations of AVS ranged from 0.40 to 64.6
//mol/g, SEM 0.20 to 778 ^mol/g and SEM/AVS ratios 0.04 to 139. The sum of the
interstitial water toxic units (IWTU) for cadmium and nickel ranged from <0.01 to 43.4.
Molar concentrations of cadmium and nickel in the interstitial water were similar. However,
cadmium contributed over 95 percent to the sum of the toxic units because the 10-day
LC50 for nickel to /V, abdita (2400 pg/U is 67 times that of cadmium (36.0 pg/U.
Sediments with the highest dry weight metals concentrations (8,600 to 71,200//g/g) were
generally toxic (> 24 percent mortality) to A., abdita. In contrast, others with similar
concentrations (<_ 13,800//g/g) were not toxic (^.24 percent mortality). Sediments with
SEM/AVS ratios <. 1.0 were always nontoxic, whereas only 5 of 11 sediments with
SEM/AVS ratios >1.0 were toxic (Figure 6-2). Sediments with .<. 0.50 IWTUs were
always nontoxic, those with > 2.2 IWTUs were always toxic and two of seven sediments
-------
6-9
with intermediate IWTUs (>_ 0.5 to .<. 2.2) were toxic (Figure 6-3). Data on chemical
concentrations and polychaete (N. arenaceodentata) mortality in tests with Foundry Cove
sediments are not included in Table 6-1 because they have been presented elsewhere by
Pesch et al. 11]. Six of 17 sediments tested with this polychaete had SEM/AVS ratios
< 1.0, 16 of 17 sediments had interstitial water toxic units < 1.0 and none of the 17
sediments were toxic (Figures 6-2 and 6-3).
Bear Creek is a tributary of the Patapsco River just east of Baltimore, Maryland
(Figure 6-1). Sediments from this portion of Baltimore Harbor are known to be toxic and
contain high concentrations of metals, PAHs, PCBs, and other substances [11,12] from
many municipal and industrial sources. Sediments used in our study were collected from
14 stations using a modified Van Veen grab. Total concentrations of divalent metals in
sediments ranged from 43.8 to 2210 //g/g dry weight, with zinc accounting for
approximately 75 percent of the total concentration (Table 6-1). Concentrations of TOC
ranged from 0.13 to 7.38 percent, silt and clay 4 to 99 percent, AVS 0.40 to 304//mol/g,
SEM 0.63 to 30.6 //mol/g, and SEM/AVS ratios 0.10 to 16.8. Seven of the 14 sediments
from Bear Creek were toxic to A., abdita these included 7 of the 9 sediments with the
highest dry weight metals concentrations (11.8 to 30.6 //mol/g). Sediments that were
nontoxic contained metals concentrations from 0.6 to 21.0 //mol/g. Both toxic and
nontoxic sediments had ^.0.03 interstitial water toxic units of metal (Figure 6-3). Given
the absence of interstitial water metal, it is not surprising that SEM/AVS ratios for
sediments from Bear Creek were not related to sediment toxicity; ie., five sediments
having SEM/AVS ratios > 1.0 were not toxic and seven of the sediments having
SEM/AVS ratios < 1.0 were toxic (Figure 6-2). Most toxic sediments released visible oil
sheens when stirred suggesting that PAHs may ultimately prove to be a source of the
observed sediment toxicity. These observations support the conclusion that toxicity
observed in Bear Creek sediments was not metal-associated.
The salt marsh containing a small tidal creek less than 500 m long (Figure 6-1) is
near Fairhaven, Massachusetts on the western side of Buzzards Bay. The creek is divided
by a hurricane barrier into an upper section of low salinity and a lower section with higher
-------
6-10
salinity. A metal products manufacturer was the principal source for metals in the
sediments. Sediments were collected by plastic scoops from 23 locations; 10 from the
upper side of the hurricane barrier and 13 from the lower section. Total concentrations
of divalent metals in these sediments ranged from 82.6 to 3320 //g/g dry weight (Table
6-1). Zinc and copper were the principal metals on a dry weight basis in these sediments.
Concentrations of TOC ranged from 0.13 to 4.39 percent, silt and clay 1.5 to 61.5
percent, AVS 0.44 to 419 //mol/g, SEM 0.73 to 31.7 //mol/g, and SEM/AVS ratios 0.10
to 6.90. Only 1 of 23 sediments from the salt marsh was toxic to A. abdita (Figures 6-2
and 6-3). The SEM/AVS ratio for the toxic sediment was 4.49 and IWTUs were 1.00. All
other sediments were nontoxic, had SEM/AVS ratios from 0.03 to 6.90 and IWTUs from
0.03 to 0.64 (21 of 22 contained < 0.3 IWTUs).
Freshwater Field Sites
Description of Field Sites and Toxicity Test Results
High concentrations of copper in sediments from Steilacoom Lake, Washington
originated principally from attempts to control aquatic vegetation using copper sulfate.
Copper SEM concentrations in sediments from eleven locations tested ranged from 0.60
to 3.91 //mol/g (38 to 248 //g/g), AVS from < 0.02 to 5.65 //mol/g, and SEM/AVS ratios
from 0.23 to 67.5 (Table 6-2) 15]. Eight of the 11 sediments tested had SEM/AVS ratios
> 1.0. No copper was detected in interstitial water (IWTU < 0.22) and no sediments
were toxic to HL azteca (Figures 6-4 and 6-5). Absence of toxicity in sediments having
SEM/AVS ratios > 1.0 (Figure 6-5) and the lack of detectable copper in the interstitial
water is likely a consequence of the presence of other sediment binding phases [51.
In contrast, 10 of 11 sediments from Keweenaw Watershed, Michigan, were lethal
to H. azteca 15]. Mining-derived copper concentrations in sediments ranged from 0.36 to
174 //mol/g (22.9 to 11,000 //g/g), AVS < 0.006 to 11.6 //mol/g, and SEM/AVS ratios
0.4 to > 17,500 (Table 6-2). The one sediment not toxic to amphipods had 0.41 toxic
units of copper in interstitial water and an SEM/AVS ratio of 0.40 (Figures 6-4 and 6-5).
-------
TABLE 6-2. SUMMARY OF SEDIMENT CHARACTERISTICS, METAL CONCENTRATIONS AND AMPHIPOD
(HYALELLA AZTECA = H.A.) OR OLIGOCHAETE (LUMBRICULUS VARIEGATUS = L.V.) MORTALITY IN
IESHWATER SEDIMENTS FROM STEILACOOM LAKE, WASHINGTON;
JRKEY CREEK, MISSOURI AND FOUNDRY COVE, NEW YORK.
Station
SEM, umol/g(T)
Day 0/10
AVS, umol/g
Day 0/10
Average
! SEM/AVS \
KEWEEN A WATERSHED, MICHIGA
IWTU(2>
tl-A. (L..V.
Percent Mortality
H.v. (L,y.)
Steilacoom Lake. Washington
1
2
3
4
5
6
7
8
9
10
11
*
1
2
3
4
5
6
7
8
9
10
11
*
1
2
3
4
5
6
7
*
0.89/1.36
3.05/2.00
1.93/1.85
2.84/3.78
1 .25/3.56
0.60/2.02
0.66/1.11
1 .33/2.68
1.95/3.91
1 .27/2.01
2.90/2.05
-
-/4.68
-/26.4
-/62.6
-/15.1
-/19.6
-/5.65
-/8.49
-/0.74
-/28.1
-/0.36
-/10.8
-
-/67.2
-/51.5
-/85.4
-/47.6
-/50.1
-/82.1
-/94.5
-
4.01/5.65
2.89/1.02
0.30/<0.02
1 .94/0.92
2.06/0.29
4.16/0.81
1 .48/1 .44
1.60/<0.05
0.39/<0.03
2.17/1.60
0.65/0.41
-
Keweenaw
-/1 1 .6
-/< 0.006
-/0.03
-/0.08
-/0.06
-/0.01
-/0.12
-/0.01
-/0.46
-/0.09
-/0.02
-
Turkey
-/28.1
-/52.2
-/30.1
-/48.4
-/44.2
-/38.2
-/78.2
-
0.23
1.51
>49.5
2.79
6.45
1.32
0.61
>27.2
>67.5
0.93
4.11
-
Watershed. Michigan
0.40
> 4,440
2,090
189
332
565
70.7
17,500
61.1
4.0
674
-
Creek. Missouri
2.39
0.99
2.84
0.98
1.13
2.15
1.21
.
<0.22
<0.22
<0.22
<0.22
<0.22
<0.22
<0.22
<0.22
<0.22
<0.22
<0.22
-
0.41
1.19
9.97
4.52
2.81
1.71
5.45
19.6
3.19
0.52
3.10
-
0.73
0.44
1.11
0.38
1.83
1.31
0.49
.
0
5
0
10
5
15
0
0
10
5
0
0
20
55
100
100
85
75
95
100
95
35
90
10
20
15
45
0
5
20
45
5
-------
TABLE 6-2. SUMMARY OF SEDIMENT CHARACTERISTICS, METAL CONCENTRATIONS AND
AMPHIPOD (HYALELLA AZTECA = H.A.) OR OLIGOCHAETE (LUMBRICULUS VARIEGATUS =
L.V.) MORTALITY IN FRESHWATER SEDIMENTS FROM STEILACOOM LAKE, WASHINGTON;
KEWEENA WATERSHED, MICHIGAN, TURKEY CREEK, MISSOURI AND FOUNDRY COVE, NEW
YORK.
(continued)
SEM, umol/g Percent Mortality
Station pay 0/10 : ;. ;'. :Day.0/10 '\ SEM/AVS IH.A. (Lv, ;i H.y, (Ly.)
Foundry Cove. New York
1 789/703 3.12/5.65 189 18.8(0.50) 100(87)
2 66.4/115 . 9.39/13.8 7.69 11.5(1.54) 100(0)
3 43.8/915 15.3/13.6 4.78 94.6(2.44) 100(0)
4 92.2/106 10.4/19.0 7.24 7.29(0.61) 100(0)
5 50.1/74.8 7.59/9.83 7.11 4.58(0.18) 80(0)
6 176/210 46.9/31.2 5.25 11.93(0.55) 100(0)
7 0.29/0.50 0.09/0.10 4.11 -(3.291) 100(0)
8 9.23/14.0 5.12/5.20 2.25 77.3 (6.49) 40 (0)
9 92.9/58.5 6.73/14.7 8.90 3.16(1.53) 100(24)
10 0.31/0.31 0.92/1.17 0.32 2.43(0.27) 0(0)
11 0.38/0.52 0.39/0.16 2.11 -(0.54) 100(0)
12 3.02/2,20 1.92/6.15 0.97 32.7(1.35) 60 (-)
13 2.02/1.79 1.18/0.64 2.25 110(1.13) 80(0)
14 3.34/7.86 20.0/10.0 0.31 2.03(3.02) 20(0)
15 1.78/0.44 9.07/9.92 0.12 0.40(0.32) 0(0)
16 0.00/0.05 0.94/0.49 0.05 -(0.38) 0(0)
* Reference sediments were from uncontaminated West Bearskin Lake, Minnesota
(1'Simultaneously extracted metal (SEM) is SEM copper for Steilecoom Lake and Keweenaw
Watershed, SEM zinc for Turkey Creek and SEM cadmium plus nickel for Foundry Cove.
(2)lnterstitial Water Toxic Units (IWTU) are calculated using 10-day water only LC50s for Hvallella
azteca of 2.8 fjg/L for cadmium, 31 //g/L for copper, 780 //g/L for nickel and 436 //g/L (hardness
330 mg/L) for zinc and for Lumbriculus varieaatus 158 //g/L for cadmium and 12,200 //g/L for
nickel.
-------
100
80
"S
40-
*
A
0
o A
0 0
o
e
D D
o
D
o
A
D
JL
AD I
A
—.—&-#&hSrr—fiWr-iWr—*-i
0.1 1 10
Interstitial Water Toxic Units
100
1000
D Foundry, hia, O Steilacoom, J±a, A Turkey, tLsu
* Foundry, Ly, ° Keweenaw,
Figure 6-4. Percent mortality of the amphipod Hvalella azteca (H.a.). and the oligochaete,
Lumbriculus varieqatus (L.v.). as a function of interstitial water toxic units (IWTU) of
metals in sediments from four freshwater field locations. The lower horizontal dashed line
at 24 percent indicates the boundary between toxic and nontoxic sediments. The higher
horizontal dashed line at 50 percent mortality and the vertical dashed line at 1.0 IWTU
indicate the hypothetical boundary between sediments expected to be toxic to less than
50 percent of the organisms {IWTU < 1.0) and those expected to be toxic to greater than
50 percent (IWTU > 1.0). Interstitial water concentrations with nondetectable metal are
plotted at 0.01 IWTU.
-------
80-
60-
CO
S 40
20-
-D-DDCD-
-ffl-
a*
D D
A A
D
AO
0.01
O O
<&o o
-#—C^T/O^-0-A-irWr-Cr-**—0-0—r—
0.1 1 10 100
SEM/AVS
1000 10000 100000
D Foundry, hLa. O Steilacoom. tLa* A Turkey, i±a.
-tr Foundry, LY. ° Keweenaw,
Figure 6-5. Percent mortality of the amohipod. Hvalella azteca (H.a.l. and the oligochaete,
Lumbriculus varieoatus (L.v.l. in sediments from four freshwater locations as a function
of the SEM/AVS ratio. The horizontal dashed line at 24 percent mortality indicates the
boundary between toxic and nontoxic sediments. The vertical dashed line at SEM/AVS
= 1.0 indicates the boundary between sulfide bound unavailable metal and potentially
available metal.
-------
6-15
Toxic sediments had 0.52 to 19.6 IWTUs of copper and SEM/AVS ratios >_ 4.0. AVS
concentrations in the 10 toxic sediments were extremely low {< 0.01 to 0.46 //mol/g)
with comparatively high copper concentrations (0.36 to 1.74^mol/g); nine SEM/AVS
ratios were >_ 61. Amphipod mortality in response to copper concentrations in water-only
tests (Figure 6-6) was almost identical to amphipod mortality as a function of interstitial
water copper concentration in sediment tests (Figure 6-3} [5]. The 10 day LC50 for
amphipods exposed to copper in water-only tests did not differ from the LC50 based on
interstitial dissolved copper concentrations and amphipod mortality from tests with
Keweenaw sediments, 31(28 to 35)//g/L versus 28(21 to 38)//g/L.
Sediments from Turkey Creek, Missouri contained high and relatively uniform
concentrations of zinc (47.6 to 94.5//mol/g; 3,110 to 6,180/;g/g) and AVS (28.1 to 78.2
yt/mol/g; Table 6-2) originating from strip mine tailings. Therefore, SEM/AVS ratios (0.98
to 2.84) and IWTUs (0.44 to 1.83) varied little in the seven sediments tested. The two
sediments having SEM/AVS ratios <. 1.0 were nontoxic and had <, 0.44 interstitial water
toxic units of zinc (Figures 6-4 and 6-5) SEM/AVS ratios of the five remaining sediments
ranged from 1.13 to 2.84, IWTUs from 0.49 to 1.83. Two of these sediments were toxic.
Sediments from Foundry Cove, New York tested with saltwater A. abdita and N..
arenaceodentata were also tested using the freshwater amphipod H. azteca and the
oligochaete L. variegatus by Ankley et al. [2]. Sediments contained approximately
equimolar concentrations of cadmium and nickel with the sum of the SEM concentrations
of these metals from freshwater tests ranging from < 0.01 to 789//mol/g, AVS 0.39 to
31.2 ywmol/g, and SEM/AVS ratios from 0.05 to 189 (Table 6-2). Four of five sediments
with SEM/AVS ratios _<. 1.0 were not toxic to amphipods while all sediments having
SEM/AVS ratios > 1.0 were toxic (Figures 6-4 and 6-5). Only the 2 sediments with the
highest SEM/AVS ratios (8.90 and 189) were toxic to the oligochaete; 14 of 16 sediments
were not toxic. Sediments with interstitial water toxic units .>. 3.16 were toxic to
amphipods; when 0.40 to 2.43 toxic units were present, no toxicity was observed.
Interstitial molar concentrations of nickel almost always exceeded those of cadmium by
one to three orders of magnitude [5]. However, cadmium was most likely the cause of
-------
100
80
OJ
60
§ 40
20
0
o o
10 100
Copper (ug/L)
1,000
Figure 6-6. Toxicity of copper to Hvalella azteca versus copper concentrations in a water-
only exposure (open symbols) and interstitial water in sediment exposures using
Keweenaw Waterway sediments (closed symbols) (Ankley et al., 1993).
-------
6-17
both amphipod and oligochaete mortalities because cadmium is over 250 times more toxic
than nickel to H. azteca: 10 day water-only LC50 for cadmium 2.8 //g/L, and nickel 780
//g/L. Similarly, cadmium is about 80 times more toxic to L varieoatus than nickel; 10 day
water-only LC50 158 jjg/L for cadmium and 12,200//g/L for nickel. Cadmium contributed
from 88.6 to 99.9 percent of the total interstitial toxic units of metals.
Discussion
Saltwater Field Sites
Bulk metals concentrations in saltwater sediments can not be used to causally relate
metal concentrations to the acute response of amphipods and polychaetes (Figure 6-7).
Mortality of amphipods in 70 sediments from five saltwater locations, or polychaetes in
16 sediments from Foundry Cove, was not related to the sum of the molar concentrations
of cadmium, copper, lead, nickel, and zinc on a dry weight of sediment basis. Sediments
having dry weight metals concentrations from 9.50 to 885 //mol/g from 17 stations in
Jinzhou Bay, Bear Creek, Foundry Cove, and the marsh in Massachusetts were toxic
(mortality > 24 percent). In contrast dry weight metals concentrations from 0.20 to 885
//mol/g were nontoxic (mortality <, 24 percent); an overlap of 2 to 3 orders of magnitude
in metals concentration.
Normalizing metals concentrations in these sediments using SEM/AVS ratios,
without insight into mortality caused by co-occurring toxic substances, also does not
permit accurate causal predictions of metal toxicity in sediments from the field (Figure 6-
8). Of the 59 sediments with SEM/AVS ratios _< 1.0 (Table 6-1), 49 (83 percent) were
not toxic and 10 (17 percent) were toxic. These 10 toxic sediments were from Jinzhou
Bay and Bear Creek. Of the 37 sediments with SEM/AVS ratios > 1.0, only 7 were toxic.
Absence of toxicity when SEM/AVS ratios are > 1.0 has commonly been observed.
However, when SEM/AVS ratios are <, 1.0, toxicity has been observed in only 4 of 92
sediments spiked with metals [4] and 1 of 15 sediments from freshwater field sites [2,5]
(Table 6-3). For all five of these sediments, the true SEM/AVS ratios may have been >
-------
100
80-
60-
40
20 -I
"••
• •
10 100
Bulk Metal (|L/mol/g dry wt)
1000
A Salt Marsh,
• Bear.
* Belledune.A^ • Foundry.
+ Jinzhou.A^ * Foundry,
Figure 6-7. Percent mortality of the amphipod, Amoelisca abdita (A.a.). and the
polvchaete. Neanthes arenaceodentata (N.a.K in sediments from five saltwater locations
in the United States, Canada, and China as a function of the sum of the concentrations of
cadmium, copper, lead, nickel, and zinc in //moles divalent metal per gram dry weight
sediment. The dashed horizontal line at 24 percent mortality indicates the boundary
between toxic and nontoxic sediments.
-------
100
80-
60-
CO
20-
_ A * A W*t» *AAJ
^^ A ^^ ^r ^^ ^^k ^1 ^v ^
^L i ^^U ^* "f^j ^^ !A
4' MA^^^ ^ ^Jf^
v&i jV^. TT
0.01
0.1
1 10
SEM/AVS
100
1000
A Salt Marsh, Aa * Belledune.Aa • Foundry. Aa
• Bear Creek, Aa + Jinzhou.Aa * Foundry, Na
Figure 6-8. Percent mortality of the amphipod, Amoelisca abdita (A.a.), and the
polychaete, Neanthes arenaceodentata (IM.a.). in sediments from five saltwater field
locations as a function of the ratio of the sum of the molar concentrations of cadmium,
copper, lead, nickel, and zinc simultaneously extracted (SEM) with acid volatile sulfide
(AVS) to the molar concentration of AVS {SEM/AVS ratio). The horizontal dashed line at
24 percent mortality indicates the boundary between toxic and nontoxic sediments. The
vertical dashed line at SEM/AVS = 1.0 indicates the boundary between sulfide-bound
unavailable metal and potentially available metal.
-------
TABLE 6-3. ACCURACY OF PREC
FROM SUING SALTWATER (SW) A
SPIKED-SEDIMENT TESTS AND (
TESTS AS A FUNCTION OF SEM//
UNITS (IWTUs) AND
••-.:-. rercer
Study Type Parameter
SW Field SEM/AVS*
IWTU
SEM/AVS, IWTU
FW Field SEM/AVS
IWTU
SEM/AVS, IWTU
Lab-Spike, SEM/AVS
(FW & SW)
IWTU
SEM,AVS, IWTU
All SEM/AVS*
IWTU
SEM,AVS, IWTU
1Nontoxic sediments <24 percent
mortality.
)ICTION OF THE TOXICITY OF SEDIMENTS
ND FRESHWATER (FW) FIELD LOCATIONS,
:OMBINED FIELD AND SPIKED-SEDIMENT
WS RATIOS, INTERSTITIAL WATER TOXIC
BOTH SEM/AVS AND IWTUs.
rt of Sediments
; Value : n
1.0 31
<0.5 59
>0.5 15
<1.0, <0.5 39
>1.0, >0.5 11
<1.0 15
> 1 .0 48
<0.5 20
>0.5 38
<1.0, <0.5, 10
>1.0, >0.5 34
^1.0 92
> 1 .0 83
<0.5 107
>0.5 77
<1.0, <0.5, 85
>1.0, >0.5 65
<1.0 149
>1.0 162
<0.5 187
StO.5 129
2=1.0, < 0.5, 134
>1.0, >0.5 110
Nontoxic1
100.0
80.6
100.0
53.3
100.0
45.5
93.3
47.9
95.0
42.1
100.0
29.4
95.7
26.5
93.5
22.1
96.5
12.3
96.6
43.2
95.7
31.0
97.8
20.9
mortality. Toxic sediments > 24
Toxic1
0.0
19.4
0.0
46.7
0.0
54.5
6.7
52.1
5.0
57.9
0.0
70.6
4.3
73.5
6.5
77.9
3.5
87.7
3.4
56.8
4.3
69.0
2.2
79.1
percent
-------
6-21
1.0 as concentrations were within the precision expected in AVS and SEM analyses and
three had > 0.5 IWTU of metal. Given the fact that field sediments from highly
industrialized locations contain many substances other than metals and are often toxic,
non-metals associated toxicity should always be suspected. If toxic sediments have
SEM/AVS ratios <. 1.0, we might suspect the cause to not be metals; with SEM/AVS >
1.0, toxicity may be related to metals.
Metals concentrations, when expressed on a sum of the interstitial water toxic unit
(IWTU) basis (Figure 6-9), can provide insight that in part may explain apparent anomalies
between SEM/AVS ratios and the observed toxicity of these sediments. In spite of the
presence of very high dry weight metals concentrations, 56 of 70 sediments had < 0.5
iwtu of metal. Of the 10 toxic sediments having SEM/AVS ratios < 1.0, none had > 0.5
IWTU of metal. This suggests that metals are unlikely the cause of the toxicity. Three of
these sediments were from Jinzhou Bay and seven from Bear Creek (most of which
released oil when agitated). The absence of toxicity in many sediments having SEM/AVS
ratios > 1.0 is understandable because most (66.7 percent; 12 of 18) of these nontoxic
sediments had < 0.5 IWTUs of metal. Of the seven toxic sediments having SEM/AVS
ratios > 1.0 (one each from Jinzhou Bay and the salt marsh and five from Foundry Cove)
all had > 0.5 IWTU of metals. Further, interstitial metal concentrations are likely to
overestimate the concentration of available metal because of differences in metal form,
greater binding to dissolved organic carbon or ligands in interstitial water [13], release of
bound metal in sampling or analytical procedures [141 or organism avoidance of metal
exposure [1].
We believe that is inappropriate to further include in this chapter data from locations
having sediments whose toxicity is almost certainly not due to metals. This decision is
additionally justified because in experiments with metal-spiked sediments [4], only 3 of 85
sediments having IWTU < 0.5 and SEM/AVS ratios < 1.0 were toxic. Therefore, data
from Bear Creek, Maryland and Jinzhou Bay, China are not included in the text, figures,
and tables that follow. These data were included above to demonstrate the value of both
SEM/AVS ratios and IWTUs to discriminate between metals-associated and nonmetals-
-------
iuu-
•
80-
•
.-""S.
^ 60-
>,
"(5
5 40;
^
20 j
n 1
T
•
**
•
• J
+
1 * • ^t*A*A* I1*
f -a- "IT 4-^ ^^ u-
• B
.. „.. •„
L
• .
•
•
*
0.01
0.1 1 10
Interstitial Water Toxic Unit
100
A Salt Marsh,
* Belledune, &,& • Foundry,
+ Jinzhou.Aa. * Foundry,
Figure 6-9. Percent mortality of amohipod. Amoelisca abdita (A.a.l. and the polychaete,
Neanthes arenaceodentata (N.a.l. as a function of interstitial water toxic units (IWTUs) of
metals in sediments from five saltwater field locations. IWTUs are the sum of metal-
specific interstitial water concentrations/10-day LC50 for cadmium, copper, lead, nickel,
and zinc. The lower horizontal dashed line at 24 percent indicates the boundary between
toxic and nontoxic sediments, The higher horizontal dashed line at 50 percent mortality
and the vertical dashed line at 1.0 IWTU indicate the hypothetical boundary between
sediments expected to be toxic to less than 50 percent of the organisms {IWTU < 1.0)
and those expected to be toxic to greater than 50 percent (IWTU > 1.0). Interstitial water
concentrations with nondetectable metal are plotted at 0.01 IWTU.
-------
6-23
associated toxicity in sediments. For the data from saltwater field locations in Belledune
Harbor, the salt marsh and Foundry Cove, all 42 sediments with SEM/AVS ratios <. 1.0
were not toxic (Figure 6-10; Table 6-3). Of the 31 sediments that had SEM/AVS ratios
> 1.0, only six sediments (from Foundry Cove and the salt marsh) were toxic and all had
> 0.50 IWTUs (Table 6-3). Of the 25 nontoxic sediments with SEM/AVS ratios > 1.0,
71.4 percent (10 of 14) of the sediments tested with amphipods and 90.9 percent (10 of
11) of the sediments tested with polychaetes had < 0.5 IWTUs, thus in part explaining
the absence of toxicity.
Saltwater and Freshwater Field Sites Combined
Metals concentrations in sediment interstitial water from all freshwater sites
suggests that metals contributed to the observed mortalities of amphipods and
oligochaetes (Figure 6-4). Therefore, all available freshwater data are included in Figure
6-5 and 6-11 to 6-15. For sediments with IWTUs .>. 0.5, 57.9 percent of 38 sediments
were toxic; 20 of 26 for amphipods and 2 of 12 for oligochaetes (Table 6-3).
The pattern of organism response to metals normalized on an SEM/AVS basis is
similar for saltwater (Figure 6-10) and freshwater (Figure 6-5) sediments. Therefore, data
in both figures were pooled (Figure 6-11) to illustrate the overall utility of the SEM/AVS
normalization to explain metals availability in field sediments. The absence of toxicity in
all but one sediment having SEM/AVS ratios <. 1.0 from all field sediments is important
given that total divalent metals concentrations (or SEM) for these sediments ranged from
43.8 to 13,800 /yg/g for tests with saltwater or freshwater amphipods, 170 to 71,200
//g/g for polychaetes, and 170 to 76,800 for oligochaetes (Tables 6-1 and 6-2). Fifty-six
of 57 (98.2 percent) of these freshwater and saltwater sediments having SEM/AVS ratios
<. 1.0 were not toxic to sensitive organisms. In the toxic sediment with SEM/AVS < 1.0,
the SEM/AVS ratio was 0.97 (Table 6-3; Figure 6-11). For field sediments having
SEM/AVS ratios > 1.0, 31 of 79 (39.2 percent) were toxic (Table 6-3). Therefore, we
believe that SEM/AVS ratios of <, 1.0, can accurately predict field sediments likely to not
be acutely toxic due to metals. Use of an SEM/AVS ratio of > 1.0 alone to predict
-------
100
80-
g 60-
co
20-
0.01
1 10
SEM/AVS
100
tooo
A Salt Marsh, &a. • Belledune,
if Foundry. N.a.
Foundry, A.a.
Figure 6-10. Percent mortality of the amphipod, Ampelisca abdita (A.a.). and the
polychaete, Neanthes arenaceodentata (N.a.). in sediments from three saltwater field
locations as a function of the SEM/AVS ratio. The horizontal dashed line at 24 percent
mortality indicates the boundary between toxic and nontoxic sediments. The vertical
dashed line at SEM/AVS = 1.0 indicates the boundary between sulf ide bound unavailable
metal and potentially available metal.
-------
100
80-
g 60-
£
oj
fc 40-
20-
d D
A
D A
—^—
0.01
T*
-------
• • • •
9
•
w
» •
1 •* .
I • • u • •
! • • f. •
Is •
1 . • • • • ••
V
•*** • * * ••••
<• • , % • *•
• • •••*•*•
1 W tt^^
•*•*••• •£
** • • * •.•*/
• Jl •
1
• •.•
<• 8
•
o o o o o <
O CO ID TT CM
^^
0
o
o
0
0
0
T™
,_
0
T™
' ^m
i O
t
o
Do
*3>
"•••«.
0
^•H
HI
CO
O
"5
«ik
fli
^»
^
S
o
i—
V)
a
3
_o
'^
£>
E
3
—1
u
03
£
A
.0
•••
o
a
J
c
c
concentratio
W
CO
4-»
0}
E
-t-J
a
'5
^
^
•o
5
M—
0
c
'*J
u
c
a
«^-
cc
M
re
^
K
u.
C
c
a
^
03
V)
D
O
OT
CO
JZ
Q
E
<
T3
O
«=
CJ
. (Q
» a
w.
CO
^>
i
w
re
renaceodent
re
CO
CD
+-
)2
-o
cc
' ^.^
,;
w
(J
""••n"
CC
«
•(-1
1
c
co
=:
4-> •"-•
and sedimer
boundary b(
fm
11
i=L
IIP
w ro
V £
m °
IB
m Q}
^ "
**" o5
0)
C '*
.2 ^
"•P 4_i
<0 m
U
O O
— c
~o iz
."£ 2
*• c
i_ 0
0) .ti
•S fc
> £
'±: -o
CO
E4S
o/0
co
LL. >
0)
C "
i!
7 -^
CO
•
-------
100
80-
~ 60-
).01
0.1
1 10 100
Interstitial Water Toxic Unit
1000
10000
A
•
•
*
a
*
o
o
Salt Marsh, A^.
Belledune, A.a.
Foundry, &SL
Foundry, N-a.
Foundry, tf.a.
Foundry, L.V.
Steilacoom, H.a.
Keweenaw, H.a.
A
O
0
0
0
0
0
e
Turkey, H.a.
Cd,A,£L
Cu. A.a.
Ni. A.a.
Pb.AA
Zn. A.a.
Mix. A.a.
Cu.CJL.
0 Pb.QjL
0 Zn.fijL
© Cd.JbLa.
O Ni. N.a.
* Cd, AJ.
® Cd. H.s.
® Cd.Ly.
Figure 6-13. Percent mortality as a function of interstitial water toxic units of metals for
saltwater and freshwater benthic species including oligochaetes Lumbriculus varieaatus
(L.v.). polychaetes Caoitella caoitata (C.c.). and Neanthes arenaceodentata (NLa.),
harpacticoids Amphiascus tenuiremis (A.t.). amphipods Amoelisca abdita (A.a.) and
Hvalella azteca (H.a.). and snails Helisomasp. (H.so.) exposed to sediments from saltwater
field locations (solid symbols), freshwater field locations (open symbols) and sediments
spiked with individual metals or mixtures (closed circles with internal symbol = saltwater,
or open circles with internal number = freshwater). Field locations, metal spiked and
species tested are indicated. The lower horizontal dashed line at 24 percent indicates the
boundary between toxic and nontoxic sediments. The higher horizontal dashed line at 50
percent mortality and the vertical dashed line at 1.0 IWTU indicate the hypothetical
boundary between sediments expected to be toxic to less than 50 percent of the
organisms (IWTU < 1.0) and those expected to be toxic to greater than 50 percent (IWTU
> 1.0). Interstitial water concentrations with nondetectable metal are plotted at 0.01
IWTU.
-------
0.1
1 10
SEM/AVS
100
1000
10000
A
4
•
*
D
*
O
0
Salt Marsh, A.a.
Belledune, A.a.
Foundry, A.a.
Foundry, N.a.
Foundry, H.a.
Foundry, L.v.
Steilacoom, H.a.
Keweenaw, tLa*
A Turkey, JdLa^
O Cd, A.a.
e CU.AA
®nll A Sk
I "if ^^^^^4,
O Pb, A.a.
0 Zn. A.a.
© Mix, A.a.
O Cu. C.c.
o Pb.C.a,
O Zn. C.c.
© Cd. N.a.
O Ni. N.a.
(D Cd. H.s.
® Cd.LiL
Figure 6-14. Percent mortality as a function of SEM/AVS ratio for saltwater and
freshwater benthic species including oligochaetes Lumbriculus varieaatus (L.v.).
polychaetes Capitella caoitata (C.c.). and Neanthes arenaceodentata (N.a.). amphipods
Ampelisca abdita (A.a.) and Hvalelja azteca (H.a.). and snails Helisoma SD. (H.s.) exposed
to sediments from saltwater field locations (solid symbols), freshwater field locations (open
symbols) and sediments spiked with individual metals or mixtures (closed circles with
internal symbols = saltwater, or open circles with internal symbol = freshwater). Field
locations, metal spiked and species tested are indicated. The horizontal dashed line at 24
percent mortality indicates the boundary between toxic and nontoxic sediments. The
vertical dashed line at SEM/AVS = 1.0 indicates the boundary between sulfide-bound
unavailable metal and potentially available metal.
-------
100-
.
80-
.
g 60-
^
75
£ 40-
^S«
5
20-
0'
-1C
D
o
.™..x.™^..™. .„. ™.....
^^tf A\ ^A T "A
AW ^^ 1»M^>/ *c!ii»^* • '"W
)0 --10 --1 k)
o CSDhO-
0 0
0
o
D D •
e
e
A
AD
IB
•
foPm
O v A A •*•-*•
-A-^lfSk— -!*- i i i -fr-f-t— V
1 10 100
D
'
^
100
SEM-AVSOL/mol/gdrywt)
A Salt Marsh, AA * Foundry,
4 Belledune, AA D Foundry,
• Foundry, AA -fr Foundry. L.v.
O Steilacoom, H.a.
^ Keweenaw, H.a.
o Turkey, H.a.
Figure 6-15. -Percent mortality of amphipods Ampelisca abdita (A.a.) and Hvalella azteca
(H.a.). oligochaetes Lumbriculus varieaatus (L.v.) and polychaetes Caoitella capitata (C.c.)
and Neanthes arenaceodentata (N.a.) exposed to sediments from three saltwater and four
freshwater field locations as a function of the sum of the molar concentratons of SEM
minus the molar concentration of AVS (SEM-AVS). The dashed horizontal line at 24
percent mortality indicates the boundary between toxic and nontoxic sediments. The
vertical dashed line at SEM-AVS = 0.0 indicates the boundary between sulfide-bound
unavailable metal and potentially available metal.
-------
6-30
sediment toxicity is useful but less accurate than predicting absence of toxicity (Table 6-3)
as would be expected based on partitioning theory and organism-sediment interactions.
In both oxic and anoxic transition zones occupied by organisms, other sediment binding
phases, metal form, and avoidance behavior of organisms can limit metal availability,
exposure, and toxicity.
Field Sites and Spiked Sediments Combined
The utility of metals concentrations normalized by dry weight, interstitial water toxic
units (IWTUs), or SEM/AVS ratios to explain the bioavailability of divalent metals and
permit prediction of sediment toxicity is summarized in Figures 6-12, 6-13, 6-14, and
Table 6-3. The figures and table are compilations of all available data from 10-day lethality
tests where mortality, IWTUs and SEM/AVS ratios are known from experiments with
sediments most certainly toxic only because of metals. They include sediments from
saltwater field sites, freshwater field sites, or sediments spiked with individual metals or
metal mixtures. The relationship between benthic organism mortality in 10-day sediment
lethality tests and bulk metals concentrations in spiked and field sediments is not useful
to causally relate metal concentrations to organism response (Figure 6-12). The overlap
among bulk metals concentrations which cause no toxicity and those which are 100
percent lethal is almost four orders of magnitude. Sediments having less than 0.01 //mol
of metal/g dry weight are all reference or control sediments.
The toxicities observed when sediment concentrations are normalized on an IWTU
basis are typically consistent with the toxic unit concept; that is if IWTUs are _<_ 1.0
sediments should be lethal to <. 50 percent of the organisms exposed; significant mortality
probably should be absent at < 0.5 IWTU (Figure 6-13). The exceptions to the
expectation that sediments with IWTUs < 0.5 should not be toxic are the two cadmium-
spiked freshwater sediments where pore water sampling procedures were a likely problem.
Of the spiked and field sediments evaluated which had IWTUs < 0.5, 95.7 percent of 187
sediments were nontoxic (Table 6-3). For all sediments having IWTUs > 0.5, 69.0
percent of 129 sediments were toxic (Table 6-3). Given the effect on toxicity or
-------
6-31
bioavailability of the presence of DOC or ligand-associated metal in interstitial water, water
quality (hardness or salinity), and organism behavior, it is not surprising that many
sediments having IWTUs > 0.5 are not toxic.
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 Di Toro et al. [15], and recommendations for assessing the bioavailability of
metals proposed by Ankley et al. [16]. Sediments spiked with metals and field sediments
from saltwater and freshwater locations with SEM/AVS ratios .<. 1.0 were uniformly (96.6
percent of 149 sediments) nontoxic (Figure 6-14; Table 6-3). The majority (56.8 percent)
of 162 sediments having SEM/AVS ratios > 1.0 were toxic. Use of both IWTUs and
SEM/AVS ratios did not improve the accuracy of predictions of sediment that were
nontoxic (97.8 percent; Table 6-3). However, it is noteworthy that toxic sediments were
predicted with 79.1 percent accuracy in 110 sediments when both SEM/AVS > 1.0 and
IWTUs >_ 0.5 were used jointly as decision parameters (Table 6-3). This approach is,
therefore, very useful in identifying sediments of concern.
Because AVS can bind divalent metals and presumably some other metals on a mole
to mole basis, normalizing metals concentrations in sediments from the field as the
difference of SEM-AVS, instead of the conventional SEM/AVS ratio, can provide important
insight into the extent of available additional sulfide binding capacity, or the extent to
which AVS binding has been exceeded (Figure 6-15). Further, absence of organism
response when AVS binding is exceeded can indicate the potential magnitude of
importance that other binding phases may have in controlling bioavailability. This insight
into additional binding capacity of AVS and other sediment phases and the magnitude of
exceedance of binding are important advantages for normalization of the concentration of
metals in sediments on an AVS basis over that of interstitial water concentration. For
most nontoxic saltwater and freshwater field sediments we have tested, 1 to 100//moles
of additional metal would be required to exceed the sulfide binding capacity; i.e., SEM-AVS
= -1 to -100//mol/g. In contrast, most toxic field sediments contained 1.0 to 1,000
A/moles of metal beyond the binding capacity of sulfide alone. Data on nontoxic field
-------
6-32
sediments whose sulfide binding capacity is exceeded (SEM-AVS is > 0.0 i/moles/g)
provides the best indication of magnitude and importance of non-sulfidic binding phases.
This is particularly true for some sediments from locations such as Steilacoom Lake and
the Keweenaw Watershed where AVS concentrations were low resulting in high SEM/AVS
ratios with little difference between SEM concentrations and sulfide binding potentials
(SEM-AVS is numerically low, whereas SEM/AVS ratios are high). The field sediments we
tested frequently contain 1.0 to 1,000//moles of metal over that bound by sulfide yet they
remain nontoxic. This indicates that the role of other sediment phases in metal
bioavailability has great significance. Therefore, further refinement on the prediction of
sediments likely to cause toxicity will require estimates of partition coefficients and binding
strengths of these sediment phases.
Summary
We believe that results from tests using sediments spiked with metals and
sediments from the field in locations where toxicity is metals-associated demonstrate the
value of normalizing sediment concentrations by SEM/AVS ratio and IWTUs, instead of dry
weight metals concentrations, in expressing biological availability of metals. Importantly,
data from spiked sediment tests strongly indicate that metals are not the cause of the
toxicity observed in field sediments when both SEM/AVS ratios are < 1.0 and IWTU are
< 0.5. Concentrations of metals in sediments on an SEM-AVS basis provides important
insight into available additional binding capacity by sulfides and other phases of sediments
and the extent to which sulfide binding has been exceeded. Predictions of sediments not
likely to be toxic because of metals based on SEM/AVS ratios and IWTUs for all data from
spiked and field sediment tests are extremely accurate (>_ 95.7 to 97.8 percent) using
either or both parameters. While predictions of sediments likely to be toxic are less
accurate (56.8 to 79.1 percent), this approach is extremely useful in identifying sediments
of potential concern. Several sources of uncertainty related to sediment geochemistry, the
kinetics of binding and release, and organism-sediment interaction need further research.
-------
APPENDIX 6A
METHODS
Saltwater Field Sites
Sediment Collection, Storage and Handling -
Sediments were collected by plastic scoop, shovel, Ponar grab, or modified Van
Veen grab from Jinzhou Bay, China (September, 1992); Belledune Harbor, New Brunswick,
Canada (August, 1990); Bear Creek, Maryland (February, 1992); a tidal marsh near
Fairhaven, Massachusetts (March 1991) and Foundry Cover New York (August 1989)
(Figure 6-1). Samples consisting of approximately 5 to 10 cm of surficial sediment were
homogenized and aliquots removed for total metal, total organic carbon and grain size
analyses. Sediments were transported under ice and stored at 4°C in sealed glass jars
with limited headspace containing nitrogen until use. Prior to conducting toxicity tests,
sediments were rehomogenized, taking care to limit oxidation of metal sulfides.
At all stations at the salt marsh in Massachusetts, interstitial water diffusion
samplers (peepers) were placed immediately below the sediment surface 13 days prior to
sediment collection to permit comparisons between in situ interstitial metals concentrations
and interstitial metals concentrations quantified during toxicity tests. Peepers consisted
of 5 ml polyethylene vials (21 mm high, 20 mm diameter), covered with a 1 micron
polycarbonate membrane and filled with 30 mg/kg salinity water [4]. A plastic strap
around the peeper extended above the sediment to facilitate recovery. Immediately prior
to sediment sampling, peepers were removed and rinsed to remove sediments. The water
contained within the peepers was removed by pipette and placed in a 7 ml polyethylene
vial and acidified with 50 //I of concentrated (pH < 1.0) nitric acid.
Toxicity Tests -
The 10-day lethality tests with the amphipod, Amoelisca abdita. generally followed
methodologies described by ASTM 117], Di Toro et al. [15] and Berry et al. 14]. Those
-------
6A-2
with the polychaete Neanthes arenaceodentata are described by Pesch et al. 11].
Amphipod exposure chambers consisted of 900 ml glass canning jars, with a 1.3 cm
diameter overflow hole covered with 400 micron Nitex® mesh. Each chamber contained
200 ml of sediment and 600 ml of seawater. Polychaete chambers consisted of 600 ml
beakers containing 200 ml of sediment. One day before the start of the test, sediment
from each station was placed into each of four [two chemistry (day 0 and 10), and two
biology] replicate exposure chambers. For each experiment with sediments from the five
saltwater locations, one or more treatments consisted of four replicate chambers
containing sediment from an uncontaminated reference station in central Long Island
Sound, Narragansett Bay or an uncontaminated sediment from a location near the study
site. Sediments from all stations at Foundry Cove and stations 1 to 10 at the
Massachusetts salt marsh site had interstitial salinities less than those tolerated by
Ampelisca or Neanthes. therefore, sediments were mixed with brine to obtain 26 to 32
mg/kg interstitial salinities prior to testing. Peepers were placed in both biology replicates
and the day 10 chemistry replicate. To provide continuous renewal of overlying water,
filtered seawater (20°C; 28 to 34 mg/kg salinity) flowed through each replicate chamber
at approximately 30 volume additions per day.
Each exposure began with random placement of 20 amphipods or 15 polychaetes
in the day 10 chemistry replicate and in the two biology replicates for each treatment.
Sediment from the day 0 chemistry replicate was homogenized, and aliquots removed and
frozen for AVS, SEM, and bulk metal analyses. Experimental chambers were checked daily
for dead animals and water flow. Overlying water was sampled at the beginning of every
test and at least once thereafter, with samples acidified and stored in vials as described
above. On day 10, peepers were removed from each sediment and the water sample
acidified and stored. Sediment from the day 10 chemistry replicate was homogenized, and
aliquots removed for AVS and SEM analyses. Sediments from the biology replicates were
sieved through a 0.5 mm mesh screen to quantify dead and surviving organisms. Samples
with more than 10 percent of the amphipods missing were recounted by a second person.
Missing amphipods were assumed to be dead. For illustrative purposes, sediments were
classified as toxic if mortality was greater than 24 percent as proposed by Mearns et al.
-------
6A-3
[18] from results of sediment tests with the amphipod Rhepoxvnius abronius. Sediments
having less than or equal to 24 percent mortality were considered as nontoxic.
Chemical Analyses -
Sediment samples were analyzed for AVS by the cold-acid purge and trap technique
described by Allen et al. [14], Cornwell and Morse [19] and Boothman and Helmstetter
[20]. SEM and bulk metals analyses were performed using inductively coupled plasma
emission spectrometry (ICP). For analyses of bulk (dry weight) metals, the metals were
extracted from freeze-dried sediments by ultrasonic agitation with 2 M cold nitric acid (50
ml/5 g wet sediment) at 60°C overnight followed by centrifugation. Results of sample
blanks and recoveries of known metal additions demonstrated 85 to 100 percent
recoveries from sediments, 85 to 115 percent recoveries from sample extracts and an
absence of contamination in our analytical procedures. The SEM concentration reported
is the sum of cadmium, copper, lead, nickel, and zinc on a micromole per gram dry
sediment basis. Concentrations of all metals in sediments exceeded analytical detection
limits.
Interstitial water from peepers and overlying water were analyzed using ICP or
graphite furnace atomic absorption spectroscopy. Detection limits varied as a function of
sample size and methods of analysts. Concentrations in water are reported as the sum of
the interstitial water toxic units (IWTU) of detectable metal. IWTUs are the sum of //g
metal/L in interstitial water •*• 10 day LC50 in water-only tests in fjg/L for all five metals,
where the 10 day LC50 for A. abdita is 36.0 ^g Cd/L, 20.5 jjg Cu/L, 3020//g Pb/L, 2400
//g Nj/L, and 343 //g Zn/L [4] and for N. arenaceodentata is 3,670//g Cd/L and 16,090//g
Ni/L [1J. Thus, if interstitial water is the principal source of metals toxicity, and availability
of metals is the same in water of water-only tests and interstitial water in sediment tests,
50 percent mortality would be expected with sediments having 1.0 IWTUs. In this
section, we use 0.5 IWTUs to indicate sediments unlikely to cause significant mortality
because on the average water-only LCO and LC50 values differ approximately by a factor
of two. This factor is reasonable because mortality was always absent in spiked-sediment
-------
6A-4
saltwater tests at 0.5 IWTU [4]. For illustration, a concentration of 0.01 IWTU is used to
indicate interstitial water samples that contain no detectable metal.
-------
REFERENCES
1. Goldhaber, M. B. and Kaplan, I.R. 1974. The sulfur cycle. In E.D. Goldberg, ed.,
The Sea, Vol. 5- Marine Chemistry. John Wiley and Sons, New York, NY, pp. 569-
655.
2. Di Toro, D.M., Mahony, J.D., Hansen, D J., Scott, K.J., Carlson, A.R., and Ankley,
G.T. 1992. Acid volatile sulfide predicts the acute toxicity of cadmium and nickel
in sediments. Environ. Sci. and Technol. 26:96-101.
3. Scott, K.J. and Redmond, M.S. 1989. The effects of a contaminated dredged
material on laboratory populations of the tubiculous amphipod, Ampelisca abdita.
In U.M. Cowgill and L.R. Williams, eds., Aquatic Toxicology and Hazard
Assessment : Twelfth Symposium. American Society of Testing and Materials,
Philadelphia, PA, pp. 289-303.
4. Casas, A.M., and Crecelius, E.A. 1994. Relationship between acid volatile sulfide
and the toxicity of zinc, lead and copper in marine sediments. Environ. Toxicol.
Chem. 13:529-536.
5. Pesch, C.E., Hansen, D.J., Boothman, W.S., Berry, W.J., and Mahony, J.D. The
role of acid-volatile sulfide and interstitial water metal concentrations in determining
bioavailability of cadmium and nickel from contaminated sediments to the marine
polvchaete. Neanthes arenaceodentata. Environ. Toxicol. Chem. In press.
6. Green, A.S., Chandler, G.T., and Blood, G.T. 1993. Aqueous-, pore-water-, and
sediment-phase cadmium: toxicity relationships for a meiobenthic copepod. Environ.
Toxicol. Chem. 12:1497-1506.
7. Carlson, A.R., Phipps, G.L, Mattson, V.R., Kosian, P.A., and Cotter, A.M. 1991.
The role of acid-volatile sulfide in determining cadmium bioavailabilty and toxicity
in freshwater sediments. Environ. Toxicol. Chem. 10:1309-1319.
-------
8. Mearns, A.J., Swartz, A.M., Cummins, J.M., Dinnel, P.A., Plesha, P. and Chapman
P.M. 1986. Inter-laboratory comparison of a sediment toxicity test using the
marine amphipod, Rhepoxvnius abronius. Mar. Environ. Res. 19:13-37.
9. Mickey, C.W. and Roper, D.S. 1992. Acute toxicity of cadmium to two species
of infaunal marine amphipods (tube-dwelling and burrowing) from New Zealand.
Bull. Environ. Contam. Toxicol. 49:165-170.
10. Kemp, P.P., and Swartz, R.C. 1988. Acute toxicity of interstitial and particle-
bound cadmium to a marine infaunal amphipod. Mar. Environ. Res. 26:135-153.
11. Hamilton, M.A., Russo, R.C., and Thurston R.V. 1977. Trimmed Spearman-Karber
method for estimating median lethal concentrations in toxicity bioassays. Environ.
Sci. Technol. 11:714-719; correction 1978, 12:414.
12. Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Hicks, M.B., Mays, S.M.,
and Redmond, M.S. 1990. Toxicity of cadmium in sediments: The role of acid
volatile sulfide. Environ. Toxicol. Chem. 9:1489-1504.
13. U.S. Environmental Protection Agency. 1994. Statistical summary, EMAP-
Estuaries, Virginian Province-1991. Office of Research and Development,
Washington, DC. EPA/620/R-94/005. March, 1994. 77pp.
14. Stephan, C.E., Mount, D.I., Hansen, D.J., Gentile, J.H., Chapman, G.A., and
Brungs, W.A. 1985. Guidelines for deriving numerical national water quality
criteria for the protection of aquatic organisms and their uses. PB85-227049.
National Technical Information Service, Springfield, VA. 98pp.
15. Gonzales, A.M., Mahony, J.D., and Di Toro, D.M. 1992 The role of organic carbon
in the toxicity of anoxic sediments contaminated with copper and other metals: An
experimental study. Abstract. 13th Annual Meeting, Soc. Environ. Toxicol. Chem.
p. 162.
-------
16. Di Toro, D.M., Zarba, C.S., Hansen, D.J., Berry, W.J., Swartz, R.C., Cowen, C.E.,
Pavlou, S.P., Allen, H.E., Thomas, N.A., and Paquin, P.R. 1991. Technical basis
for establishing sediment quality criteria for non-ionic organic chemicals using
equilibrium partitioning. Environ. Toxicol. Chem. 10: 1541-1583.
17. Swartz, B.C., Dittsworth, G.R., Schults, D.W., and Lamberson, J.O. 1985.
Sediment toxicity to a marine infaunal amphipod: Cadmium and its interaction with
sewage sludge. Mar. Environ. Res. 18:133-153.
18. Hoke, R.A., Ankley, G.T., Cotter, A.M., Goldstein, T., Kosian, P.A., Phipps, G.L.,
and VanderMeiden, F. 1994. Evaluation of equilibrium partitioning theory for
predicting acute toxicity of field-collected sediments contaminated with DDT, DDE,
and ODD to the amphipod. Hvalella azteca. Environ. Toxicol. Chem. 13:157-166.
19. Swartz, R.C., Cole, F.A., Lamberson, J.O., Ferraro, S.P., Schults, D.W., DeBen,
W.A., Henry Lee II, and Ozretich, R. J. 1994. Sediment toxicity, contamination and
amphipod abundance at a DDT- and dieldrin-contaminated site in San Francisco Bay.
Environ. Toxicol. Chem. 13:949-962.
20. Adams, W.J., Kimerle, R.A., and Mosher, R.G. 1985. Aquatic safety assessment
of chemicals sorbed to sediments. In R.D. Cardwell, R. Purdy and R.C. Bahner, eds.,
Aquatic Toxicology and Hazard Assessment: Seventh Symposium. STP 854.
American Society for Testing and Materials, Philadelphia. PA, pp. 429-453.
21. Mahony, J.D., Di Toro, D.M., Gonzalez, A.M., Berry, W.J., and Ankley, G.T. 1991.
A sediment component in addition to acid volatile sulfide that may further control
toxicity of metals. 12th Annual Meeting of the Society of Environmental
Toxicology and Chemistry, p. 91.
-------
22. Ankley, G.T., Phipps, G.L, Leonard, E.N., Benoit, D.A., Mattson, V.R., Kosian,
P.A., Cotter, A.M., Dierkes, J.R., Hansen, D. J. and J.D. Mahony. 1991. Acid-
volatile sulfide as a factor mediating cadmium and nickel bioavailability in
contaminated sediments. Environ. Toxicol. Chem. 10:1299-1307.
23. Carignan, R., Rapin, F., and Tessier, A. 1985. Sediment porewater sampling for
metal analysis: A comparison of techniques. Geochim. Cosmochim. Acta 19:2493-
2497.
24. Hesslein, R.H. 1976. An in situ sampler for close interval pore water studies.
Limnol. Oceanogr. 21:912-914.
-------
CHAPTER 7
COLONIZATION EXPERIMENTS
All experiments presented thus far demonstrating that AVS is important in
controlling the toxicity of sediment-associated metal have relied on 10-day laboratory
lethality tests using individual benthic species exposed to homogenized sediments from
field sites or sediments spiked with divalent metals [1,2]. Chronic exposures of individual
species or benthic communities have not only recently been completed using sediments
whose metal and AVS concentrations have been measured and have varied with depth as
is normal in the field [3,4,5].
The benthic colonization test [6] is particularly useful in evaluating the effects of
substances on developing benthic communities in the laboratory and field. In this test, the
most sensitive early life stages of benthic organisms found in unfiltered seawater are
chronically exposed to chemicals as they settle and grow in replicated control and treated
sediment-filled aquaria. Resultant communities are diverse, consisting of thousands of
individuals, represented by 40 or more species and several phyla. Results from early
colonization experiments where chemicals were continuously added to incoming seawater
revealed test sensitivities predicted by water quality criteria (WQC); i.e., generally,
observed effect concentrations were greater than WQC and no observed effect
concentrations were less than WQC [71. More recently, organic chemicals have been
spiked into sediments [8,9] and comparisons between results of these tests and chemical-
specific sediment quality criteria indicate SQC are protective [71.
This chapter presents the results of a 118 day benthic colonization experiment in
which sediments were spiked with cadmium to obtain nominal SEM/AVS ratios of 0.0
(control), 0.1, 0.8, and 3.0. Numbers and kinds of organisms that colonized the sediment
are compared to total cadmium, interstitial water cadmium, and SEM/AVS ratios as
exposure conditions changed temporally and spatially as a function of sediment depth to
evaluate if these measurements can be used to determine sediments which are not
chronically toxic to benthic organisms. The methods used are presented in Appendix 7A.
-------
7-2
Results
Exposure
At the beginning of the experiment the average AVS concentration of the sediments
for the four treatments was 17.2//mol/g dry weight and measured SEM/AVS ratios, 0.00,
0.10, 0.60, and 2.63, approximated nominal values; control, 0.1, 0.8, and 3.0,
respectively (Table 7-1). The average AVS concentration in all samples of homogenized
sediment analyzed throughout the experiment was 17.2 //mol/g (range 12.2 to 22.7
//mol/g), a concentration the same as that at test initiation. The average SEM/AVS ratios
in cadmium-spiked sediments over the length of the experiment also differed little from
that at test initiation for nominal SEM/AVS = 0.1 (0.09 for day 0 versus 0.10 overall) and
SEM/AVS = 0.08 (0.66 for day 0 versus 0.60 overall). The apparent 25.3 percent
decrease in SEM/AVS ratio in the 3.0 SEM/AVS treatment (3.52 for day 0 versus 2.63
overall) probably results from an unusually low analysis of AVS of 13.0 //mol/g at test
initiation. If the mean AVS for all treatments on day 0 is more representative of the true
value in this treatment, then a more accurate SEM/AVS ratio would be 2.66 for day 0 and
2.57 overall. Therefore, chemical analyses of homogenized sediments indicate that the
exposure was constant throughout the experiment.
However, chemical analysis of horizons from sediment cores demonstrate that SEM
and AVS concentrations and SEM/AVS ratios varied with sediment depth. Data from two
cores sampled on day 14 from a control treatment show AVS concentrations < 1.0//mol/g
in the surface 1.0 cm, from about 1.0 to 8.0//mol/g at 1.0 to 3.0 cm depth, and greater
than 15.0//mol/g below 3.Ocm depth (Figure 7-1). At day 28, AVS concentrations in the
surface 3.0 cm of sediment in the control, 0.1 and 0.8 SEM/AVS treatments were lower,
with those in the surficial 0.6 cm being 20 percent of those on day 0 (Figures 7-1 and 7-
2). By contrast, AVS concentrations below 3.0 cm in depth remained similar to those at
-------
.3
• Control day 14 (7/17/91)
D Control day 28 (7/31/91)
A 3.0 SEWAVS day 14
(7/17/91)
A Pettaquamscutt River
(7A8/91)
10 15 20
AVS jimol/g
Figure 7-1. Comparison between the AVS concentrations by sediment depth in the
cadmium colonization experiment on day 14 (July 17,1991), day 28 (July 31, 1991), and
in Pettaquamscutt River, Rhode Island (July 18, 1991).
-------
T - 28 days
10 20 30 40 50
T = 57 days
T - 117 days
0 10 20 30 40 SO
0 10 20 30 40 SO
10.0
10.0
0 10 20 30 40 SO
10.0
10 20 30 40 SO
10 20 30 40 SO
10.0
10.0
I0°0 10 20 30 40 SO
0°
10.0
10 20 30 40 SO
10.0
0 10 20 30 40 SO
10.0
10 20 30 40 SO
0 10 20 30 40 SO
0°
10.0
10 20 30 40 SO
Cd & AVS (vmol/g) Cd & AVS (wmol/g) Cd & AVS (vmol/g)
Figure 7-2. Vertical distribution of SEM cadmium AVS and SEM/AVS ratios on days of 28,
57, and 117 of the cadmium colonization experiment. The four treatments are control and
nominal SEM/AVS ratios of 0.1, 0.8, and 3.0. Measured SEM cadmium concentrations
in //mol/g are plotted as a thin line, AVS concentrations in jt/mol/g as a cross hatched area
and SEM/AVS ratios as a thick line. The striped area indicates where SEM exceeds AVS.
-------
,5
test initiation. Vertical distribution of AVS in the 0.1 and 0.8 SEM/AVS treatments were ^P
generally similar to the control on days 24, 56, and 117. Loss of AVS in surficial
sediments in these treatments was due to oxidation, most likely of principally iron sulf ide.
TABLE 7-1 . MEASURED CONCENTRATIONS OF ACID VOLATILE SULFIDE (AVS,
//MOL/G), CADMIUM SIMULTANEOUSLY EXTRACTED WITH AVS (SEM,
fjMOLIG) AND SEM/AVS RATIOS IN HOMOGENIZED SEDIMENTS FROM DAY 0
TO 1 17 OF THE BENTHIC COLONIZATION TEST.
, Day of Measurement
Nominal
Measure
Control
0.1
.8
3.0
Mean
0
SEM
AVS
SEM/AVS
SEM
AVS
SEM/AVS
SEM
AVS
SEM/AVS
SEM
AVS
SEM/AVS
AVS
• 14
0.0
20.4
0.0
1.5
16.1
0.09
12.9
19.5
0.66
45.8
13.0
3.52
17.2
28
0
13
0
1
13
0
12
22
0
38
13
2
15
.0
.7
.0
.6
.6
.12
.9
.3
.58
.3
.5
.84
.8
56
0
15
0
1
12
0
11
17
0
49
19
2
16
.0
.8
.0
.5
.4
.12
.5
.2
.67
.1
.6
.50
.2
117
0.0
12.2
o.o
1.4
19.0
0.07
11.6
21.1
0.55
47.2
22.7
2.08
18.8
Mean
0
21
0
1
13
0
11
20
0
40
.0
.6
.0
.3
.6
.10
.8
.5
:58
.5
15.4
2
17
.63
.8
SEM/AVS
0
16
0
1
14
0
12
20
0
44
16
2
17
.0
.7
.0
.5
.9
.10
.1
.1
.60
.2
.8
.63
.2
However, in the 3.0 SEM/AVS treatment from day 14 throughout the experiment,
profiles of AVS concentration with depth were different from profiles in lower treatments.
On day 14, concentrations of AVS in sediments from the 3.0 SEM/AVS treatment were
essentially constant at all depths (Figure 7-1). In this sediment, concentration of AVS
ranged from 13.8 to 16.2 //mol/g (mean = 15.1 //mol/g) in individual horizons in the
surface 2.4 cm, and from 10.7 to 21.7 //mol/g (mean =17.5 //mol/g) below 2.4 cm. At
later sampling days, AVS was only slightly oxidized in the 3.0 SEM/AVS treatment
compared to the other treatments (Figure 7-2). Lesser oxidation in this treatment was
likely due to lower oxidation rates of cadmium sulf ide [10]. Little oxidation of AVS below
-------
7-6
2.4 cm is evident during the experiment in the control, 0.1 and 0.8 SEM/AVS treatments
and below 1.2 cm in the 3.0 SEM/AVS treatment.
Vertical profiles of AVS in sediments from the cadmium colonization experiment on
day 14 (July 17, 1991) and day 28 (July 31, 1991) were qualitatively and, to a lesser
extent, quantitatively similar to those observed in sediment cores from the nearby
Pettaquamscutt River, Rhode Island (July 18, 1991) 111] (Figure 7-1). Relative to AVS
concentrations with depth in the experiment, AVS concentrations in Pettaquamscutt River
sediment were only slightly higher from the surf ace to about 5.0 cm, proportional increase
in concentrations from the surface to about 2.5 cm were similar and AVS concentrations
were stable in both the field and our laboratory sediments from about 2.5 to 5.0 cm.
Sediments from Pettaquamscutt River consisted of approximately 39.5 percent sand, 58.0
percent silt, 2.5 percent clay, and 1.3 percent TOC. The granulometry of sediments used
in our experiment was markedly different; 5.6 percent sand, 70.7 percent silt, 23.7
percent clay, and 1.0 percent TOC.
Loss of cadmium from spiked sediments was essentially confined to the surface 1.2
cm (Figure 7-2). Decreases of cadmium in surface sediments were less than decreases in
AVS in the control, 0.1 and 0.8 SEM/AVS treatments. Decreases in cadmium with depth
were similar for all three cadmium-spiked treatments. In the surface 0.6 cm, cadmium
concentrations for all treatments averaged 65.4, 67.0, and 22.7 percent of those below
3.0 cm on day 28, 56, and 117, respectively. Cadmium concentrations in the 0.6 to 1.2
cm horizon were 95.4, 86.8, and 79.0 percent of those below 3.0 cm on day 28, 56, and
117, respectively. Cadmium concentrations below 1.2 cm remained unchanged
throughout the experiment.
Vertical profiles of measured SEM/AVS ratios were consistent with the observed
oxidation of AVS in surficial sediments as tempered by losses of cadmium (Figure 7-2).
At sediment depths greater than about 2.4 cm, there was little or no oxidation of AVS, no
loss of cadmium and SEM/AVS ratios remained stable near the nominal values. In surficial
sediments, losses of AVS exceeded those of cadmium causing SEM/AVS ratios to increase
-------
7-7
dramatically, which could potentially release previously bound metal to affect benthic
organisms. While SEM/AVS ratios increased to as much as 0.75 in individual 0.6 cm
horizons in the nominal 0.1 SEM/AVS treatment, the molar concentration of AVS was
always in excess of that for cadmium, therefore, metals toxicity was unlikely. In the
nominal 0.8 SEM/AVS treatment, loss of AVS exceeded that of cadmium and measured
SEM/AVS ratios exceeded 1.0 in from one to all five (22 of the 40 samples) 0.6-cm
sediment horizons down to 3.0 cm in all cores sampled on days 28, 56, and 117.
SEM/AVS ratios in the surface 1.8 cm of sediment averaged 1.45 on day 28, 1.01 on day
56, and 1.94 on day 117. Measured SEM/AVS ratios in individual horizons from the
nominal 3.0 SEM/AVS treatment always exceeded a ratio of 1.0, hence, were always
potentially toxic.
Concentrations of cadmium measured in interstitial water collected by peepers were
consistent with sulfide binding (Table 7-2). Cadmium concentrations in interstitial water
were below the limit of analytical detection (< 3 //g/U in 70.8 percent of 24 samples from
control replicates and 50 percent of 24 sample from the 0.1 SEM/AVS treatment. In the
0.1 SEM/AVS treatment, average concentrations in surface (6.4/yg/L) and bottom (3.6
//g/L) peepers were less than the saltwater acute (42 //g/L) and chronic (9.3 //g/U water
quality criteria [12]. Therefore, neither acute lethality nor chronic effects would be
expected in sediments where the nominal SEM/AVS ratio was 0.1. Interstitial water
concentrations in the nominal 0.8 SEM/AVS treatment are elevated over that in the control
as might be expected given that measured SEM/AVS ratios frequently exceeded 1.0.
Cadmium concentrations in interstitial water in the 0.8 SEM/AVS treatment averaged 58
//g/L in surface peepers and 48 //g/L in bottom peepers. These average concentrations
were sufficiently high to be acutely toxic to the most sensitive saltwater species and
chronically toxic to many sensitive arthropods and polychaete species 112]. Interstitial
concentrations in the 3.0 SEM/AVS treatment always exceeded acute and chronic water
quality criteria concentrations and were sufficiently great to result in acute lethality for
most saltwater species tested in water-only tests [12]. Therefore, both SEM/AVS ratios
and interstitial water concentrations indicate effects on benthic taxa should occur in this
-------
7-8
treatment. Given the above exposure conditions and speculation on organism response,
the following section describes observed organism responses.
TABLE 7-2. MEAN CADMIUM CONCENTRATION fc/G/L) IN INTERSTITIAL
WATER CONCENTRATION COLLECTED IN THREE DIFFUSION SAMPLERS
(PEEPERS) PLACED IMMEDIATELY BELOW THE SEDIMENT SURFACE AND
THREE PEEPERS 2.0 CM ABOVE THE BOTTOM OF TEST AQUARIA.
TREATMENTS WERE CONTROL AND NOMINAL MOLAR RATIOS OF CADMIUM
SIMULTANEOUSLY EXTRACTED METAL (SEM) AND ACID VOLATILE SULFIDE
(AVS).
Nominal Peeper Day of Measurement Treatment
SEM/AVS
Control
0.1
0.8
3.0
NDa = Less
one-half the
Location
>
Surface
Bottom
Surface
Bottom
Surface
Bottom
Surface
Bottom
: 14
NDa
4
8
ND
28
38
138,000
174,000
than detection limit 3.0
detection
76,
135,
//g/L
28
ND
7
8
ND
48
85
000
000 1
56
ND
ND
8
10
157
48
66,000 28,
54,000 88,
Mean concentrations
limit for samples below the
detection limit
117
ND
ND
ND
ND
ND
20
000
000
derived
•
Mean
ND
3.5
6.4
3.6
58
48
77,000
138,000
using
Effects
Sediment-associated cadmium had an effect on both the timing of organism
appearance on and in the sediment and the abundance and species composition of
organisms found in each treatment. In the control, 0.1 and 0.8 SEM/AVS treatment, a
reddish brown microfloral layer appeared on the sediment surface after about week three;
however, this layer did not appear until after about week 9 in the 3.0 SEM/AVS treatment.
Animals including snails, polychaetes and tonicities began to appear in the control, 0.1 and
0.8 SEM/AVS treatments shortly after the appearance of diatoms at about week three,
whereas in the 3.0 SEM/AVS treatment they did not appear until after about 9 to 11
weeks.
-------
7-9
The microfloral layer, sampled on day 80 contained the pennate diatoms
Entomonies. Nitzschia. Plaaiotropis. Bacillaria. Amphora. Thalassiosira. Navicula.
Rhzicoshoenia. Licmophora. Lithodesmium. Cvclotella. Rhabdonema. Skeletonema.
Diploneis. and Chaetoceros. in approximate order of decreasing abundance. The density
of pennate diatoms measured on day 80 corroborated visual observations of periphyton
abundance on the sediment surface (Table 7-3). The mean diatom density in control
aquaria, 1.92 x 106 cells/cm2 of sediment surface, was not different from the 1.32 x 106
cells/cm2 in the 0.1 SEM/AVS aquaria or the 1.00 x 106 cells/cm2 in the 0.8 SEM/AVS
aquaria. Diatom densities were significantly lower (0.16 x 106 cells/cm2 of sediment
surface) in the 3.0 SEM/AVS aquaria. The density of diatoms in five of eight aquaria for
this treatment were less than the lowest density (0.17 x 106 cells/cm2) in any aquarium
from other treatments.
TABLE 7-3. NUMBER OF DIATOMS x106 PER SQUARE
CENTIMETER OF SEDIMENT SURFACE IN REPLICATE AQUARIA
Treatment: Nominal CdSEM/AVS Ratio
= Significantly different from control; a - 0.05
A total of 14,347 individual macrobenthic benthic organisms, representing 54
species from 8 phyla, was collected from sediments sieved from all aquaria: eight control
aquaria and eight aquaria for each of the three treatments that contained sediments spiked
-------
7-10
at 0.1, 0.8, and 3.0 SEM/AVS (Tables 7-4 and 7-5). Annelids, arthropods, and chordates
were most abundant; over half (27 of 52) of the species were polychaetes. Most
individuals of all species were subadult; only Nereis succinea. Polvdora socialis and
Molgula manhattensis had significant numbers of adults. Individuals of all feeding types,
suspension feeders, deposit feeders, selective deposit feeders, omnivores, and carnivores,
were collected.
TABLE 7-4. TOTAL NUMBER OF SPECIES AND
INDIVIDUALS {IN PARENTHESES).
Treatment: ;NominaI
CdSEM/AVS
PHYLA
Annelida
Mollusca 2(4)
gastropoda
Bivalvia 4(6)
^rthropoda 7(3640)
Mematoda ?(563)
Sipuncula 1(1)
Dnidaria 0(0)
^hynocoela 1(1)
2hordata 4(197)
CONTROL 0.1
18(347) 19(330)
1(4)
2(3)
5(2078)
?(432)
1(2)
1(1)
0(0)
4(59)
0.8 3.0
2K2653) 8a(159a)
5(7)
1(5)
3(4)
0(0)
8(4228) 6(1294a)
?(254a) ?(128a)
1(1)
0(0)
0(0)
3(102)
0(0)
1(4)
0(0)
3(469a)
TOTAL 37(4196) 33(2909) 39(4862) 19a(2058)
Significantly different from controls; a = 0.05
Species richness and abundance decreased with increasing concentration of
cadmium in the sediment (Tables 7-4 and 7-5). In the nominal 0.1 SEM/AVS treatment,
no effect on either number of species or individuals, relative to the control, was detected
in the colonized benthic communities. In the nominal 0.8 SEM/AVS treatment, no
significant effects were detected in the overall total number of individuals or species.
However, there were significantly fewer polychaetes (Mediomastus ambiseta. Streblosoio
-------
TABLE 7-6. NUMBER OF INDIVIDUALS OF BENTHIC SPECIES THAT COLONIZED SEDIMENTS IN EIGHT CONTROL
AQUARIA AND EIGHT AQUARIA CONTAINING SEDIMENTS WITH CADMIUM SPIKED AT NOMINAL MOLAR RATIOS OF
CADMIUM TO ACID VOLATILE SULFIDE OF 0.1, 0.8. OR 3.0. NUMBER OF AQUARIA CONTAINING THAT SPECIES IN
PARENTHESES.
SEM/AVS
i .. Taxeo ;
Maereinvartabrataa:
Elida
Ntritt uccintt
Palydon toemUt
Mfdiomtstut tmbltttt
folvdon Kotl
Poh/don ctuStfyl
Cnnodrilui ttmna
Srr»o/ojoto teneoVctf
Podtrkt obtain
fwurfa aanoumaa
StbtH* microethtlmt
Btcctrdit ».
Thtmtte
Potycimis tf.
SthtnoMltat
NtettiY* oka
Hvdroidn ditrahut
Paetinana oouW/i
CaorfaMa eaortata
Amphtntt trctict
Dodtctcefit conchtrwn
Awbna rwnota
ffao/ta htttmoodt
Potomillm ntolfctt
Saiettaat
Stbtltrit vutotrit
Ntrinidu tridtntttt
tambahd
Molluaea
Gaatrepoda
Actiodnt etnt^culttt
CaAerMa arvzn
Antnhft Itfntnty
Cmidult famfcam
Bitthmto.
Odostomi* to.
Nudibmnchit
Rivarvia
Pflricolm pholtdHormit
Nucult trmultu
Mteamtltntt
bjnMttp
Arthrepada
Cruataeaa Harpacticoda
harpacticeid bb
htrotaicoM tl
Aajgacljcojo^ B£i
Cumaeaa
Dimmtis jcutefa
Amphipoda
J»«j»
-------
7-12
benedicti and Podarke obscura) and unidentified meiofaunal nematodes in this treatment
(Table 7-4). Sediments containing a nominal SEM/AVS ratio of 3.0 were colonized by
fewer macrobenthic species (19) than controls (37), but the total number of individuals
was not significantly affected (Table 7-4 and 7-5). This was principally because the
number of polychaete species and their total abundance was significantly diminished.
There were significantly fewer individuals of the polychaete species M- ambiseta. S_.
benedicti. P. obscura and the unidentified harpacticoid copepod species. Bivalve molluscs
were absent in this highest treatment. The number of chordates, principally the tunicate
M- manhattensis. was significantly greater in the 3.0 SEM/AVS treatment (Table 7-5).
Numbers of these tunicates would have been even higher if myriads of small « 1 mm
diameter) tunicates, attached to large individuals but having no direct contact with
sediment as well as those contacting sediments, had been counted.
The length-frequency distributions for the polychaete Nereis succinea in control, 0.1
and 0.8 SEM/AVS treatments indicates recruitment was continuous with no effect on
abundance or growth (Figure 7-3a). In the 3.0 SEM/AVS treatment, there were no worms
over 45mm in length; whereas 8.5 percent of the worms in other treatments exceeded this
length (Figure 7-3b). Only 15.3 percent of the worms in the 3.0 SEM/AVS treatment were
over 15mm in length, whereas 45.8 percent exceeded this length in other treatments.
Extremely small (0 to 5mm) worms predominated (37.8 percent) in the 3.0 SEM/AVS
treatment compared to 11.6 percent for other treatments. The absence of large (>
45mm) worms, markedly reduced occurrence of worms of intermediate lengths,
preponderance of very small worms and absence of visible colonizers of any macrobenthic
species for the first half of this study suggests that the sediments in the highest treatment
were initially lethal but could later be tolerated by resistant species. Therefore, polychaete
length-frequency distributions may indicate delayed recruitment, not decreased growth.
r
The response of these communities to cadmium-spiked sediments can be assessed
by evaluating changes in abundance of individual species and phyla as was done above,
or by summarizing organism abundance in the entire assemblage using an index. Cluster
analyses that compare the species present or absent in each replicate with those in other
-------
CONTROL
0.1x
0.8x
5 10 15 20 25 30 35 40 45 50 55 110
LENGTH (MM)
3x
M*«n oJ Control, O.tx.
-------
7-14
replicates from the same or different treatments, indicate that three clusters may exist.
Although all treatments have many species in common, the kinds of species in the control
treatment are most similar to those in the 0.1 SEM/AVS treatment. Species present or
absent from replicate aquaria in these treatments are different from those in the 0.8 and
3.0 SEM/AVS treatments (Figure 7-4). Further, the clusters of presence-absence data for
the species in aquaria in the 0.8 and 3.0 SEM/AVS treatments indicate differences in
species composition between these highest treatments. Numbers of species in the control,
0.1, 0.8, and 3.0 SEM/AVS treatments were 37, 33, 39, and 19 respectively, based on
raw counts (Table 7-4) and 51,46, 57, and 34 respectively, based on Jackknife estimates
of maximum species richness [13]. Because numbers of species in the control, and 0.1
and 0.8 SEM/AVS treatments are similar, shifts in community structure detected in the 0.8
SEM/AVS treatment compared to these two lower treatments probably occurred because
of changing species composition and not the number of species present. In the 3.0
SEM/AVS treatment, there was a change in both species abundance and composition.
Discussion
Concentrations of AVS in marine sediments vary with depth as a function of
seasonal processes [14]. Iron sulfide is formed by the anaerobic diagenesis of organic
matter. As a result of bacterial sulfate reduction, concentrations of AVS increase in
surficial sediments during warm months of greatest productivity and sediment oxygen
demand. AVS is readily oxidized in cold months when productivity is minor and oxidizing
conditions greatest. Therefore, in winter AVS concentrations in the surficial sediments
decrease. For example, in the Pettaquamscutt River, Rhode Island AVS concentrations in
the upper 3 cm of sediment may vary 15 to 25 fold between summer and winter [11].
Maximum concentrations occur in surface sediments between 2 and 5 cm depth and
concentrations below these depths down to 15 cm decrease only marginally with season.
Vertical profiles of AVS in the cadmium colonization experiment in week 2 (July 17,
1991) and week 4 (July 31,1991) were qualitatively and, to a lesser extent, quantitatively
similar to those measured in the Pettaquamscutt River, Rhode Island by Boothman and
-------
0.48
0.64
0.80
0.96
O = CONTROL
• = 0.1x
D = 0.8x
• = 3x
'.O
-O
-O
-O
-O
-O
-4
•D
-D
n
D
n
a
Figure 7-4. Cluster analysis of presence-absence data for species from sediments from
control and nominal 0.1, 0.8, and 3.0 SEM/AVS treatments.
-------
7-16
Helmstetter [11] at the same time as our experiment. Oxidation of AVS in surficial
sediments in our experiment proceeded rapidly, with most occurring within two to four
weeks of test initiation. The oxidation of sediments observed initially must be associated
with passive oxygen diffusion and not be biotically driven as organisms large enough to
bioturbate the sediments were not present during the first few weeks of the experiment.
Their presence later apparently did little to change existing AVS profiles. Lesser oxidation
of AVS early and throughout the experiment in the 3.0 SEM/AVS treatment, and the loss
of cadmium from only the surficial 1.2 cm compared to 2.4 cm for AVS, agrees with
laboratory sediment suspension experiments that demonstrate rapid (100 percent in 60 to
90 minutes) oxidation of iron sulfide versus slow (10 percent in 300 hours) oxidation rates
for cadmium sulfide [10]. Further oxidation of sediments should result in release of only
a portion of sulfide-associated cadmium to interstitial water because sedimentary Fe and
Mn are transformed into their oxyhydroxides which, along with organic carbon, can bind
released cadmium in oxic sediments [15]. In this and other laboratory experiments [16],
AVS gradients and the complex microhabits associated with sediment geochemical
processes and organism burrowing developed rapidly. These observations indicate that
the opinion that bioassays are simplistic because they do not have the vertical gradients
and microhabits that occur as part of naturally occurring geochemistry and biological
processes [17] may not apply to all laboratory exposures.
The bioavailability and toxicity of divalent metals in field sediments can not be
predicted using metals concentrations on a sediment dry weight basis [18]. However, Di
Toro et al. [19] in laboratory studies using spiked sediments, observed that acute toxicity
and interstitial water concentrations of cadmium were related to sediment concentrations
on a //mol cadmium per //mol acid volatile sulfide basis. They hypothesized that this
normalization should apply to other divalent metals in both freshwater and marine anoxic
sediments, where the metals concentrations are expressed as the ratio of the sum of the
molar concentrations of all divalent metals to the molar concentration of AVS, with the
metal-sulfide solubility products providing insights into the likely metal of concern in a
mixture. These observations were further extended to apply to copper, lead, nickel, and
zinc, as well as metal mixtures, in acute lethality tests with spiked sediments and
-------
7-17
saltwater amphipods and polychaetes [1,16,20,21 ] and freshwateroligochaetes and snails
[22,23]. Studies revealed the requirement that metals concentrations be expressed as the
molar concentration of metal simultaneously extracted with AVS, not total metal. Acute
lethality tests with homogenized sediments from freshwater and marine locations
consistently demonstrated an absence of toxicity when SEM/AVS ratios were ^ 1.0 and
that sediments having a ratio > 1.0 were sometimes toxic, but frequently were nontoxic
[2,20,22,24,25]. Absence of lexicologically significant concentrations of metal in
interstitial water in nontoxic sediments and the presence of interstitial metals
concentrations of concern in toxic sediments highlighted the utility of the toxic unit
concept as applied to interstitial metal in predicting sediment toxicity when SEM/AVS is
> 1.0. Absence of toxicity when SEM/AVS is > 1.0 suggested the presence of other
binding phases in these sediments. Ankley et al. [26] summarized much of the above
information and proposed methodologies for assessing the potential bioavailability of
metals in sediments. The studies above have involved exposures of 10 days or less and
homogenized sediments, therefore, extrapolations using SEM/AVS ratios or interstitial
water concentrations to chronic responses of benthic organisms in laboratory or field
sediments with vertical gradients and microenvironments should be done with caution.
Because of these limitations, this colonization experiment and those of Hare et al. [3] and
Liber et al. [4] were conducted.
Biological effects of cadmium-spiked sediments observed in our colonization
experiment are consistent with the previous interpretation of SEM/AVS ratios, when
vertical gradients are considered, and with interstitial water cadmium concentrations and
the relative sensitivities of benthic taxa in water-only aquatic toxicity tests (Table 7-6;
Figure 7-5). In the nominal 0.1 SEM/AVS treatment, molar concentrations of AVS in
vertical profiles of sediment were always in excess of those of cadmium; interstitial
cadmium concentrations (ND to 10//g/L) were either below detection limits, or less than
known concentrations of toxicological significance in water-only tests [12]; and no
significant effects were observed on sediment colonization by benthic species. In the
nominal 0.8 SEM/AVS treatment, measured SEM/AVS ratios averaged 0.60 in
homogenized sediments. However in surf icial sediments, molar concentrations of cadmium
-------
)TENTIAL EFFECTS GIVEN
Y TESTS.
a- Z
09
t- cc
— <
H >
UJ 2
K en
i!
ui 2
o- 5E
x 0
UJ (J
: CADMIUM COLONIZATIOI
Y SENSITIVITIES OF BENTh
i 5
i-o
z z
li
UJ (/)
« 0
C
sjl
Bfz
U. UJ
UJ U
|i
v» 2
F Q
5 <
z o
O -J
u <
ui H
CO S?
o *
O- S-
X Z
UJ —
u. 0
OUJ
«£
<( (/)
^» CQ
1°
i
UJ X «-
0 0. TO
Z ro 01
3
3
o
Q
Z
1
00
U)
o
01
S
00
n
d
-^
^.
9
o
0
o
6
,.
o
Acute lethality unlikely.
Chronic lethality or
•o
O n
is £
i£
9- '£
W C
II
o
Fewer polychaetes
(Mediomastus. Streblospi
r?
in
5
Ol
1
eo
-
0)
•_
q
r-
^
-'
•^
CO
i
in
0
o
Species presence/absenci
altered.
en
Acute lethelity for most
macroinvertebrate specie
'a
c
E
O 01
o E
•n «
0 *O
Q
•o «~
o:S
S 0
Fewer species, polychaet
(Mediomastus. Streblospi
**.
5
1
8
=d *
5* *^
II
_
in
^j
00
00
A
01
oi
-^
in
n
§
co
-------
o
o
o
CO
o
CD
c
CO
a:
42 "CD
w .£
S E
£i
« T>
X C
o m
£1
9 o
« ii
^-» c
if
II
gi
« 0
2 co
S.E
M CO
0 C
3 O
« a
> z
o ^
'E 0
P u
w o
c ._
8 Si
m
£.28
0 « 2
|-E€
S ° E
8*5
°1
88
el
" 0
CO
0
CO
£«^
w 2. •
CO 0-C3J
5s°
I if
>l|
.ll|
i»!
fc- 4-t . .
*• ® c
2i
u E
•s «
o 0
ajj
co W
•o >
II
CD til
w
Ss
*3 **
CD I
«-> >
g.£
SB
8 S
§S
II
11-
O *- >
+rf
•
0
CO
0
6
C
I E
u> 3
^ E
«•§
>• o
= « .2 .2
m K «^ .t!
« 2 SS S-o-g
"S-SggS
• tV 4^
£-53
*+. CO
3«J
S M -o « - J
£ /•» C TS 0 r»
CD
'§ S
•c'S
o '>
TJS
^
«=JS
§l
£*
o •= •=
.2*- «* "
•^ o g
c "c 3
O O -.s
C '^
0 co
+* w
&|
a 1
2 «.
«s
^ c
w> m
0 ™
EW *" — __ TO 0) ^**j
> -1-I..CU00C
^ w_*-0-gocSw-
S !i°:I|i8|fl
f5 S. *i O Q.T3 = c .. t ~
0
a.
o
CM
1/Bn
uimiupeo
-------
7-20
frequently exceeded those for sulfide and cadmium concentrations in interstitial water (ND
to 157 fjgIL) often exceeded those that in water-only tests are known to be acutely toxic
to sensitive species. Chronic effects may be projected to occur to many sensitive
arthropods and polychaetes if acute toxicity and acute-chronic ratios are used to estimate
chronically toxic concentrations (Figure 7-5). Significant reductions in the abundance of
polychaetes and nematodes, but not arthropods, and alterations in species composition
were observed. In the nominal 3.0 SEM/AVS treatment, molar concentrations of cadmium
always exceeded those of sulf ide in both homogenized sediments and in sediment profiles,
and interstitial water concentrations of cadmium were of sufficient magnitude (28,000 to
174,000 fjg/L) to pose significant acute and chronic risks to almost all algal and
macrobenthic saltwater organisms for which data are reported in the water quality criteria
document [12]. Biological effects observed in our study were severe, with numbers of
species reduced by about one-half. Significant reductions occurred in total polychaete
species, abundance and size of certain polychaete species and abundance of nematodes
and harpacticoid copepods. Bivalve molluscs were absent and diatom density was reduced
10-fold. Interestingly, the total number of individuals of benthic organisms was similar to
that of controls. This suggests that tolerant benthic species that could avoid exposure by
their epibenthic habits replaced more sensitive species. For example, the tunicate Moloula.
known to be resistant to many substances in this test [6], rests on sediments utilizing
water from above the sediment surface and was particularly abundant in 3.0 SEM/AVS.
Alternatively, organisms may seek micro-environments known to occur in the surface 1
to 2 cm of sediments [14], as was observed in the oolvchaete Neanthes arenaceodentata
exposed to cadmium or nickel-spiked sediment [16].
We conclude that the present theories used to predict the acute biological
consequences of divalent metals in sediments may be applicable to chronically exposed
benthic organisms. Predictions of the toxicoiogical significance of metals in laboratory and
field sediments must consider vertical profiles of SEM and AVS relative to biologically
active sediment strata, interstitial water metal concentrations and the potential for release
of nonavailable metal as a result of oxidation of AVS (including both iron and cadmium and
-------
7-21
other toxic metal sulfides) as a part of the normal seasonal sulfide cycles and sediment
bioturbation.
In addition to the marine colonization study described in this chapter, there have
been two colonization experiments with freshwater sediments that have examined the role
of AVS in determining metal bioavailability [3,4]. Because these experiments only recently
have been completed, they are not described in detail herein; however, the studies will be
presented in full at the SAB meeting. For the sake of completeness, below we briefly
describe the freshwater colonization experiments, including major results and conclusions.
Hare et al. [3] conducted an experiment in which cadmium was spiked into clean
field-collected sediments, which were then placed in trays and put in the Precambrian
shield lake from which the sediments were initially collected. Nominal SEM:AVS ratios in
the samples were 0.05 (control), 0.1, 0.5, 2, and 10. SEM:AVS ratios and pore water
cadmium concentration were measured in 3 cm horizons of the sediment over the course
of slightly greater than one year, after which macroinvertebrate samples were collected
to evaluate benthic community structure and cadmium bioaccumulation by the benthos.
The results of the bioaccumulation study are addressed in Chapter 8. Except for the
sample with a nominal SEM-.AVS ratio of 10, water overlying the sediments (collected with
"peepers") contained non-detectable levels of cadmium. At all the test concentrations,
oxidation of AVS in surficial sediments resulted in greater SEM:AVS ratios in the
shallowest horizon than in deeper portions of the core. Pore water cadmium
concentrations in the control and 0.1 treatment were consistently low. However, in the
0.5 treatment, pore water cadmium concentration were elevated, particularly in the
shallowest sediment horizons, presumably due to surficial oxidation of AVS. In the
samples with nominal SEM:AVS ratios of 2 and 10, pore water cadmium concentrations
were consistently elevated; however, there appeared to be little if any impact of the
cadmium on organism abundance (Figure 7-6). One of the explanations suggested by the
authors for the seeming lack of impact was that SEM:AVS ratios actually can be quite
misleading. Although samples with very small AVS concentrations (such as those in their
study, ca., 0.2 umol AVS/g) might exceed an SEM:AVS ratio of 1, the actual amount of
-------
Abundance Accumulatio
I
600
400
200
0
75
50
25
0
400
V) 200
20°
10°
C
— o
2000
1000
0
75
50
25
Chaoborus
Polycer,tropus
H*H
Procladius
III M-H M-H 1-
._j .... , .. P ,
+H—M-H—H-H—I-H
4-
Proclaclius sp.c
Serc'entia
Chirnomus
n
0.6
0.4
0.2
0.0
1.5
1.0
0.5
0.0
10
o
Q_
I
10 .
(Q
1
0.1
10
1
10
1
0.1
0.01 0.1 1 10 0.01 0.1 1 10
Cd/AVS
Figure 7-6. Comparison of SEM (cadmium) :AVS to abundance of, and bioaccumulation
of cadmium by six taxa of benthic macroinvertebrates from a field study.
-------
7-23
total free metal could be quite small. This led the authors to conclude that it may be more
appropriate to use the absolute difference between SEM and AVS to predict the presence
of bioavailable metal [31.
Liber et al. [4] conducted a colonization experiment with zinc spiked into clean
sediments from a small mesotrophic pond near Duluth, Minnesota. The spiked sediments
were replaced in the pond in trays, and sampled periodically over the course of about 14
months for determination of SEM:AVS ratios and zinc pore water concentrations, and
benthic community structure. Five zinc concentrations, ranging from 0.8 to 12 umol/g
were tested in order to cover the expected seasonal range in AVS concentrations.
However, the zinc-sulfide complex proved to be exceptionally stable to oxidation relative
to iron monsulfide, (see Chapter 9), in that there was a concentration-dependent increase
in sediment AVS content with increasing zinc concentration (Figure 7-7). The net result
of this was that SEM:AVS ratios at the four lowest treatments never exceeded one, and
only slightly exceeded one in the highest zinc spiking regime. This exceedence occurred
only in surficial (0 to 2 cm) sediments; similar to the study of Hare et al. [3], AVS
concentrations in the shallowest horizons, irrespective of the zinc treatment, were smaller
than those in deeper sediments (Figure 7-7). Regardless of the measured SEM:AVS ratio,
zinc was rarely detected in the pore water, and never at biologically significant
concentrations. Sediment cores, collected during each sampling period, were not toxic to
Chironomus tentans or Hvalella azteca in laboratory bioassays, nor was there any
discernable impact on diversity of abundance of benthic communities in the zinc-spiked
samples (Figure 7-8). Results form the study confirmed that when molar AVS
concentrations exceed those of SEM, little or no free metal is present in pore water, and
toxicity to benthic organisms does not occur.
-------
CO
>
05
.c
in CM o> CD co o
in CM en CD co o
in CM en co co o
in. CM en CD co o
o
«^
o
CO
J*
'a
(0
T3
CO
JO
o
*->
o
V)
s
a>
E
T3
CO
(0
CO
+••
_c
(0
to
•4-f
0)
O
c
o
o
0)
O)
(0
CO
< to
CO
• *-"
h» CO
£a
II
NOIlVaiNHONOO SAV NV3H
-------
(Q
CD
VI
I
00
CD
CD
r-t
o'
Q)
O
O
5'
<
CD
•^
i-f
CD
ST
Q)
f»
CD
Q)
cr
c
a.
03
3
O
CD
H-
V)
O
ca
CD
a.
3'
CD
3
Q>
<
CO
O
3
c*
O
O
3"
CD
CD
CA
Q)
5
(O
a.
0)
r+
CD
S 8 S I 8
NUMBER OF ORGANISMS PER TRAY
X § o 8 § S § 8
33=
m
>»
-r
o
- -. - M M ^^^B
. 'I § § § I ~
n
MM O
O
5
S
•— —
(AUGUST 23, 1993)
MACROINVERTEBRATE
NUMBER OF ORGANISMS PER TRAY
o S 8 S 8 8 § S I o S 8 S § i S o § I §
>•»
«t
3D
m
w
n
^~ I
^
s
i
M
8
8
8
i
o
o
&
_*— .
(OCTOBER 1, 1993)
W
2
O
m
•D
m
30
30
>
-------
APPENDIX 7A
METHODS
Effects of cadmium spiked into sediments on macrobenthic organisms that colonized
sediments was tested in a 118-day (July 4 to October 21, 1991) experiment using a
control and three cadmium-spiked sediment treatments, with nominal SEM/AVS ratios of
0.1, 0.8, and 3.0. There were 12 replicate, 8 biological and 4 chemical, aquaria (13.3 x
30 x 15 cm high) for each treatment (Figure 7A-1). A stratified random placement
strategy was used for location of treatments and replicates among the four test
apparatuses was used. Approximately four liters of sediment from central Long Island
Sound south of Milford, Connecticut (5.6 percent sand, 70.7 percent silt, 23.7 percent
clay and 1.0 percent TOO, defaunated by freezing, were added to a depth of 8 cm to each
aquarium. A splitter box delivered unfiltered saltwater, (29 to 32ppt salinity; 16 to
23.5°C) from the West Passage of Narragansett Bay to one end of each aquarium at 200
ml/min (about 300 volume additions per day). This unfiltered saltwater contained
planktonic larvae and other lifestages of benthic organisms. A drain hole at the opposite
end maintained water depth over the sediment at 2 cm.
Sediment from central Long Island Sound (mean 17.2//mol AVS/g dry weight) was
spiked with cadmium chloride dissolved in seawater, homogenized, and stored at 20°C for
26 days prior to test initiation. Treatments were nominal //mol SEM cadmium///mol AVS
ratios of 0.0 (control), 0.1, 0.8, and 3.0. The Long Island Sound sediment has low levels
of metals (about 0.4 jjglg cadmium, 40//g/g copper, 60//g/g lead, 15//g/g nickel, and 130
/jg/g zinc) that contribute 3.17 //mol/g to the total SEM and 0.18 to the total divalent
metal SEM/AVS ratio. Hereafter, SEM/AVS ratio used will be //mol SEM cadmium///mol
AVS. Sediments from this location have proven biologically acceptable as control or
reference sediment in a great number of tests conducted previously at U.S. EPA,
Narragansett, Rhode Island. The experiment began after addition of sediments to aquaria
and initiation of water flow.
Replicate chemical aquaria were sampled on days 14, 28, 56, and 117. Each
chemical replicate contained six interstitial water diffusion samplers (peepers), three
-------
•o
CO
CD
c .c .c
— oo
ECO CO
ui uj
0}
a
CD
•= • . CD
E « » a
CO
w a.
T3
c
o -~
ISS5
fl) O < t
^^ ^^ ^™
CD tl ^ 3
to o
c
o
CO T3
^ *—
>
v)
co
CO 0) 00
3 CO •
_ _, O
co
0) 3 v. Q)
> -O C S
0) CD O S
51. ° co
" -
x c3 ^
O H- CO
'i g s •§
|1|1
o ® >"°
co £- i: CD
to I— Jo £
co • 'c 2
co en c s
CO f
2 £ CD
co
CO
k^ i_ W (U r~
°r ° -a c «
40 o c -S |
— •= CO CD -S
S £ ^ » T3
CD
CD
°>£ CO
o
CD
£
3 C O 5
00
™ O
n5!
_c •»-• o
C — CD
5 Q.£
CO
- co >E
IS 3 w O
U. CO CO U
-------
7A-3
immediately below the sediment surface and three 2 cm above the bottom of the test
aquaria (Figure 7A-1). Peepers were polyethylene vials containing 6.3 ml of filtered
seawater with 1 -micron pore size polycarbonate mesh across the top. On each sampling
day, three core tubes, 6.5-cm ID, were first inserted into the sediment in the front, middle
and end of the chemistry replicate for each treatment and the top capped. Next, peepers
were withdrawn, rinsed with seawater to remove adhering sediment, the membrane
punctured with a disposable pipette tip, the water withdrawn, acidified, and stored for
analysis. Cores were removed, the bottom capped, and frozen. Remaining sediments
were homogenized and an aliquot frozen. Before chemical analysis, sediment cores from
days 28, 56, and 117 were extruded from core tubes and sliced into five 0.6 cm horizons
in the first 3.0 cm of sediment and in 2.0 cm horizons for the remainder of the core.
Cores from day 14 were sliced at a variety of depth horizons. Each horizon from each core
was homogenized before chemical analyses. Homogenized remaining sediments and cores
were analyzed for SEM cadmium, AVS, and interstitial water metal by AA or ICP using the
methods of Boothman and Helmsletter [1], Allen et al. [27], and U.S. EPA [28].
All biological replicates were sampled on day 80 to obtain periphyton samples and
on day 118 to remove the macrobenthos. Periphyton were pipetted from two 8-mm
circles of surfical sediment from each aquarium at the midline and one-third the distance
from each end. The volume of each these two samples was increased to 3.0 ml and the
samples refrigerated. Diatoms in three aliquots containing 0.0636 //I from each sample
were counted within 24 hours using a Palmer-Maloney cell. Counts were adjusted to
cells/cm2. Sediments were sieved through stacked 2.5, 0.5, and 0.3-mm sieves to remove
macrobenthic organisms. The organisms were relaxed in magnesium sulfate and preserved
in 10 percent buffered formaldehyde with rose bengal stain. Organisms were identified to
species except for a few small or damaged specimens. Relaxed polychaetes, Nereis
succinea. were measured for length prior to preservation. For tunicates, only individuals
immediately in contact with sediment were counted; myriads of extremely small tunicates
attached to larger tunicates-hence not in contact with sediment and present only in one
control and five 3.0 SEM/AVS replicates-were not counted.
-------
7A-4
Analysis of variance was used to detect differences in the abundance of periphyton
and individual macrobenthic animal species and phyla. Cluster analyses were performed
to compare the kinds of macrobenthic species present or absent in each replicate or
treatment using the simple matching coefficient [29].
Sjk = a + d
a + b + c + d
where, a = number of species present in both replicates,
b = number of species present in replicate k only,
c = number of species present in replicate ] only,
d = number of species absent in both the replicate j and replicate k but
that occur in at least one other replicate in the experiment, and
(a + b + c + d) = the total number of species found in the experiment.
Numbers calculated by this measure range from 0 (dissimilar) to 1 (similar).
-------
REFERENCES
1. Berry, W.J., Hansen, D.J., Mahony, J.D., Di Toro, D.M., Robson, D.L., and
Boothman, W.S. 1994. Predicting the toxicity of metals-spiked laboratory
sediments using acid volatile sulfide and interstitial water normalizations.
Manuscript.
2. Hansen, D.J., Berry, W.J., Mahony, J.D., Boothman, W.S., Robson, D.L., Ankley,
G.T., Ma, D., Van, Q., and Pesch, C.E. 1994. Predicting the toxicity of metals-
contaminated field sediments using interstitial water and acid volatile sulfide
normalizations. Manuscript.
3. Hare, L., Carignan, R., and Huerta-Diaz, M.A. 1994. A field experimental test of
the hypothesis that acid volatile sulfide (AVS) concentrations improve the prediction
of metal toxicity and accumulation by benthic inverebrates. Limnol. Ocean. Vol.
39.
4. Liber, K., Ankley, G., Call, D., Markee, T., and Schmude, K. 1994. Seasonal
relationships between acid volatile sulfide concentrations and toxicity of zinc to
benthic macroinvertebrates. Manuscript In Preparation.
5. DeWitt, T., and Swartz, R. 1994. Personal communication.
6. Hansen, D.J., and Tagatz, M.E. 1980. A laboratory test for assessing impacts of
substances on developing communities of benthic estuarine organisms. American
Society of Testing and Materials. Special Technical Publication 707. pp. 40-57.
7. Hansen, D. J. and Flemer, D.A. 1993. Extrapolation of single species tests to
water and sediment quality criteria: A comparison with field and laboratory
saltwater benthic colonization experiments. Abstract. 14th Annual Meeting. Soc.
Environ. Toxicol. Chem. p.49
-------
8. Tagatz, M.E., Plaia, G.R., and Deans, C.H. 1985. Effects of 1,2,4-trichlorobenzene
on estuarine macrobenthic communities exposed via water and sediment.
Ecotoxicol. Environ. Safety. 10:351-360.
9. Flemer, D.A., Stanley, R.S., Ruth, B.F., Bundrick, C.M., Moody, P.M., and Moore,
J.C. 1994. Recolonization of estuarine organisms: Effects of microcosm size and
pesticides. Hydrobiologia. In press.
10. Mahohy, J.D., Di Toro, D.M., Koch, R., Berry, W. and 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. Soc.
Environ. Toxicol. Chem. p.46.
11. Boothman, W.S. and Helmstetter, A. Vertical and seasonal variability of acid
volatile sulfides in marine sediments. Manuscript.
12. U.S. Environmental Protection Agency. 1985. Ambient Water Quality Criteria for
Cadmium. Office of Water Regulations and Standards. EPA 440/5-84-032.
127pp.
13. Heltshe, J. and Forrester, N. 1983. Estimating species richness using the jacknife
procedure. Biometrics. 39:1-11.
14. Jorgensen, B.B. 1977. The sulfur cycle of a coastal marine sediment (Limfjorden,
Denmark). Limnology and Oceanography. 22:814-832.
15. Zhuang, Y., Allen, H.E., and Fu, G. 1994. Effect of aeration of sediment on
cadmium binding. Environ. Toxicol. Chem. 13:717-724.
-------
16. Pesch, C.E., Hansen, D.J., Boothman, W.S., Berry, W.J., and Mahony, J.D. The
role of acid-volatile sulfide and interstitial water metal concentrations in determining
the bioavailability of cadmium and nickel from contaminated sediments to the
marine polvchaete Neanthes arenaceodentata. Environ. Toxicol. Chem. In press.
17. Luoma, S.N. and Carter, J.L. 1993. Understanding the toxicity of contaminants in
sediments: Beyond the bioassay-based paradigm. Environ. Toxicol. Chem. 12:793-
796.
18. Luoma, S.N. 1989. Can we determine the biological availability of sediment-bound
trace elements? Hydrobiologia 176/177:379-396.
19. Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Hicks, M.B., Mays, S.M.,
and Redmond, M.S. 1990. Toxicity of cadmium in sediments: The role of acid
volatile sulfides. Environ. Toxicol. Chem. 9:1487-1502.
20. Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Carlson, A.R., and Ankley,
G.T. 1992. Acid volatile sulfide predicts the acute toxicity of cadmium and nickel
in sediments. Environ. Sci. Tech. 26:96-101.
21. Casas, A.M. and Crecelius, E.A. 1994. The relationship between acid volatile
sulfide and the toxicity of zinc, lead and copper in marine sediments. Environ.
Toxicol. Chem. 13:529-536.
22. Ankley, G.T., Phipps, G.L. Leonard, E.N., Benoit, D.A., Mattson, V.R., Kosian, P.A.,
Cotter, A.M., Dierkes, J.R., Hansen, D.J., and Mahony, J.D. 1991. Acid volatile
sulfide as a factor mediating cadmium and nickel bioavailability in contaminated
sediments. Environ. Toxicol. Chem. 10:1299-1307.
23. Carlson, A.R., Phipps, G.L., Mattson, V.R., Kosian, P.A., and Cotter, A.M. 1991.
The role of acid volatile sulfide in determining cadmium bioavailability and toxicity
in freshwater sediments. Environ. Toxicol. Chem. "10:1309-1319.
-------
24. Ankley, G.T., Mattson, V.R., Leonard, E.N., West, C.W., and Bennett, J.L. 1993.
Predicting the acute toxicity of copper in freshwater sediments: Evaluation of the
role of acid-volatile sulfide. Environ. Toxicol. Chem. 12:315-320.
25. Burgess, R.M. and Morrison, G.E. 1994. A short-exposure sublethal sediment
toxicity test using the bivalve Mulinia lateralis: Statistical design and comparitive
sensitivity. Environ. Toxicol. Chem. In press.
26. Ankley, G.T., Di Toro, D.M., Hansen, D.J., Mahony, J.C., Swartz, R.C., Hoke, R.A.,
Thomas, N.A., Garrison, A.W., Allen, H.E., and Zarba, C.S. 1994. Assessing the
potential bioavailability of metals in sediments: A proposed approach. Environ. Mgt.
18:331-337.
27. Allen, H.S., Fu, G., and Deng, B. 1993. Analysis of acid-volatile sulfide (AVS) and
simultaneously extracted metal (SEM) for the estimation of potential toxicity in
aquatic sediments. Environ. Toxicol. Chem. 12:1441-1453.
28. U.S. Environmental Protection Agency. 1991. Draft analytical method for
determination of acid volatile sediment in sediments. Office of Water, Office of
Science and Technology. August, 1991. 17pp.
29. Smith, E.P., Pontasch, K.W., and Cairns, Jr., J. 1990. Community similarity and
the analysis of multispecies environmental data: A unified statistical approach.
Water Res. 24:507-514.
-------
CHAPTER 8
BIOACCUMULATION OF METALS
Although the presence and/or absence of toxicity can, to some extent, be inferred
as a barometer of contaminant bioavailability, in many instances a more accurate endpoint
for assessing bioavailability is bioaccumulation. Even if a chemical is bioavailable, if the
test species of concern is not particularly sensitive and/or the length of the test is too
short, toxicity may not be manifested. Therefore, bioaccumulation also should be
considered when assessing whether a particular model is appropriate for predicting metal
bioavailability in sediments. This issue is of particular relevance to the bioavailability
paradigm presented in previous Chapters of this document. Basically, if interstitial water
concentrations of metals are small or non-detectable and/or (divalent) metahAVS ratios
measured in appropriate sediment horizons are less than one, significant accumulation of
metals by macrobenthos should not occur. If significant bioaccumulation does occur this
would suggest that, regardless of the results of toxicity studies, the bioavailability model
which has been advanced for metals in sediments is not completely accurate.
Because bioaccumulation is a convenient endpoint for biomonitoring, the literature
is replete with observations of metal bioaccumulation by vertebrate and invertebrate
species associated with metal-contaminated sediments. In many instances, researchers
have attempted to correlate concentrations of metals in field-collected animals with some
measure of sediment metal concentrations. However, as is true for toxicity, total
extractable metal concentrations in sediments are relatively poor indicators of that fraction
of metal "which apparently is available for bioaccumulation (for reviews see Tessier and
Campbell, [11; Luoma, [2]; Hare, [3]). Differential extraction techniques in conjunction
with speciation models sometimes have improved correlations between concentrations of
metals in animals and sediments, however, in most instances these types of models tend
to be empirical and somewhat site-specific.
Definition of the true bioavailable fraction of metals is only part of the difficulty in
quantitatively linking metal bioaccumulation by organisms.to metals in sediments, either
-------
8-2
in the laboratory or field. Although assessing bioaccumulation of metals from sediments
theoretically should be a relatively straight-forward process, if not adequately controlled,
a number of key variables can bias interpretation of observed results. One issue which is
particularly problematic in field studies is the separation of exposure to sediment-
associated metals (either from pore water or particulates) from exposure to metals in the
water column. In the majority of systems contaminated by metals, both the sediments and
overlying water have elevated metal concentrations. Because most benthic organisms
have the potential for exposure via both media, it is difficult to separate the relative
contribution of the two routes. This is compounded by the fact that while relatively
accurate sediment chemistry can be obtained with single samples, it is very difficult to
collect meaningful overlying water quality data temporally. This is not to say that from a
holistic standpoint metals bioaccumulated by organisms from overlying water are not
important; in fact, an overall model for assessing the impacts of metals on benthic
organisms should incorporate exposure both from sediment and overlying water [4,5].
However, to develop a mechanistic understanding of factors mediating metal bioavailability
in sediments, the contribution from overlying water must be monitored or controlled.
A number of biological/physiological factors also can complicate the interpretation
of sediment bioaccumulation studies with benthic organisms. For example, species which
ingest sediment must have their gut contents physically removed, or purged (e.g., by
holding in clean water or sediment) in order to effectively separate metal that actually has
been bioaccumulated from that associated with sediments in the gut [6,7]. Another
potential bias in metal bioaccumulation studies with some species is that a sizable
percentage of the metals measured on a total body basis may actually only be adsorbed
to the outer integument [8]. A final problem with metal bioaccumulation as an endpoint
for assessing bioavailability is that many invertebrate species are capable of regulating
body burdens of essential trace metals (e.g., zinc, copper) to some extent (e.g.,
Timmermans et al., [9]). Hence, the lack of bioaccumulation does not always necessarily
indicate a lack of bioavailability.
-------
8-3
A number of studies have been conducted which permit the critical evaluation of
metal bioaccumulation by macroinvertebrates relative to sediment pore water metal
concentrations and SEMrAVS ratios. These studies include short-term laboratory
experiments with cadmium- or nickel-spiked marine and freshwater sediments, short- and
long-term laboratory experiments with field-collected marine and freshwater sediments
variously contaminated with cadmium, nickel, copper, lead, and zinc, and a long-term field
study with cadmium-spiked freshwater sediments. Species assessed included molluscs,
oligochaetes, polychaetes, amphipods, and chironomids. In these experiments, the
potential confounding variables described above, in particular clearance of gut contents,
were controlled to varying degrees, which in some instances causes ambiguity in the
interpretation of the study results.
Laboratory Spiking Experiments-Freshwater
/
Carlson et at. [10] exposed oligochaetes (Lumbriculus varieoatus) and snails
(Helisoma sp_) to freshwater sediments, containing three different AVS concentrations,
which each had been spiked to achieve nominal cadmiumrAVS ratios of 0, 0.1, 0.3, 1.0,
3.0, and 10. Assays were conducted for 10 d in a system which provided approximately
11.5 turnovers of clean Lake Superior water/d. At test conclusion, mortality of the two
organisms was assessed and tissue samples were collected for residue analysis. As
discussed elsewhere in this document (Chapter 5), significant mortality of the two test
species only occurred at cadmium:AVS ratios greater than one. However, even at
cadmium:AVS ratios less than one, dissolved concentrations of cadmium in pore water
from the spiked sediments were often greater than in unspiked sediments. Cadmium
concentrations in surviving oligochaetes and snails were below those tissue concentrations
expected to result in toxicity based upon comparison to concurrent water-only cadmium
tests with the two species. However, cadmium residues in both species appeared to
increase in a concentration-dependent manner at cadmium:AVS ratios less than one for all
three test sediments (Figure 8-1). Comparison of the actual magnitude of these increases
in worms and snails from the cadmium-spiked sediments is complicated by the fact that
a complete set of cadmium concentrations in organisms from unspiked (control) samples
-------
Pequaywan Sna
D- -
t=
Pequaywan Worr
1
t
i
East River Snai
.--A---
East River Worn
-co
CO
*i3
SC
a v*
CD
CD
U>
I
<
§
West Bearskin Wo
•
\
%
«
\
\
\
\
- O
- co
1 "« *v
\> *•
» *
• »•
D
cd
cc
to
co .2
d £
•o
(0
O
X)
e
.3
T3
(0
O
g
+3
JO
3
O
U
(0
o
15
2
|
Q>
E
TO
0)
V)
(0
£
(0
(D
•o
0)
.*
'5.
jfl
0)
0)
0)
g
'^
(Q
, co
H- o
w* .
§81
e§
•i I
"O Q)
CO
U
•»- TJ
O C
O
O
O
I
o
8
0
s
r
o
o
I
o
(6/fin)
sissy.
col
c" ^
CB ^
O
U
T^
00
-------
8-5
are available for only one of the three test sediments. However, if concentrations of
cadmium in organisms from this one sediment, which were on the order of 0.5 to 5 ug/g
dry weight, are taken to be indicative of those in all the sediments, then cadmium
concentrations were as much as two orders of magnitude greater than background in
surviving test organisms from sediments with SEM:AVS ratios less than one.
An important shortcoming in the study by Carlson et al. [10] was that cadmium
residues measured in the oligochaetes and snails included all gut contents, i.e., the
organisms were removed directly from the sediment without allowance for gut purging.
Data are not available with which to evaluate the potential contribution of gut contents to
the total body weight of the snails; however, recent studies by Brooke et al. [11] indicate
that in L. variegatus which have not been gut purged, approximately 10 percent of the dry
weight of the worms may be sediment. Using this estimate, if it is assumed that the
cadmium concentration in sediment in the gut of the oligochaete was similar to the
cadmium concentration in the surrounding sediment, depending upon the test sediment,
as little as 2 percent or as much as 50 percent of the total cadmium body burden in the
oligochaetes could have been due to gut contents. It should be noted that the assumption
that sediment in the gut is equivalent to that outside the organism becomes increasingly
tenuous as the selectivity of feeding increases.
Laboratory Spiking-Marine
Pesch et al. [12] conducted experiments in which the toxicity and bioaccumulation
of cadmium or nickel spiked into marine sediments were evaluated using the polychaete
Neanthes arenaceodentatg. Two different sediments, with relatively low and high AVS
concentrations, were spiked so as to achieve final SEM (cadmium or nickel):AVS ratios of
0, 0.1, 0.3, 1.0, 3.0, 10, 30, and 100. Exposures were conducted for 10 d in a system
which provided 30 to 50 turnovers of clean sea water/d. At the end of the exposure,
surviving organisms were removed from the test sediments, and placed in clean sea water
for 4 h. At this time, visual inspection suggested that the polychaetes had completely
purged their gut contents. As discussed elsewhere (Chapter 5), no significant mortality
-------
8-6
occurred at SEM-.AVS ratios less than one; at SEM-.AVS ratios greater than one and when
interstitial water toxic units (TU) also were greater than one, the polychaetes either died
or avoided burrowing in the cadmium- or nickel-spiked sediments. At SEM:AVS ratios less
than one, dissolved concentrations of cadmium and nickel in pore water from the two
sediments were comparable to values from unspiked sediments. Bioaccumulation of
cadmium and nickel was most pronounced in sediments with SEM:AVS ratios greater than
one; cadmium concentrations in polychaetes from sediments with ratios greater than one
typically were more than an order of magnitude greater than concentrations in worms from
sediments with ratios less than one (Figure 8-2). Nickel concentrations in animals from
sediments with SEM:AVS ratios greater than one were approximately two- to 10-fold
greater than nickel concentrations in polychaetes from sediments with ratios less than one
(Figure 8-3).
Although bioaccumulation of both cadmium and nickel was most pronounced at
SEM:AVS ratios greater than one, there also was a concentration-dependent increase in
tissue metal concentrations in the polychaetes at SEMrAVS ratios less than one (Figures
8-2 and 8-3). There are a number of possible explanations for this. Metal residues in
polychaetes from sediments with the lower SEM:AVS ratios could have been derived via
ingestion and digestion of contaminated paniculate matter. Alternatively, uptake of the
two metals could have been directly from interstitial water; the burrowing activity of the
animals could have served to effectively oxidize metal-sulf ide complexes thereby releasing
metal to the immediate environs of the polychaetes. Because this phenomenon would
occur only in the microenvironment of the organisms, the release of metals may not have
been manifested in measurements of cadmium and nickel concentrations in the pore water
collected via "peepers." One other possible explanation for the observation of apparent
metal bioaccumulation by the worms in samples with SEM:AVS ratios less than one, and
pore water metal concentrations similar to control values, may be that the metals were
adsorbed to the chitinous sheath of the polychaetes, and were not actually in the
organism. This type of phenomenon has been documented for freshwater invertebrates
such as chironomids [8], but to our knowledge, has not been investigated for NL
arenaceodentata.
-------
CD CO
CO c.
Tissue Cadmium (ug/g)
g. CD
8 " <=
Q) " ™"
»o o
3 O X
il °
b co
Q. ->
CD CO
s §
a)
C* 0
U -K
' CO
m
o"
CO
Q.
3 p _
| "*"
CO
S 0
8 |
5 w
I co
co 2" -* —
^3 ^^to
I 2
3 "^
CO ^
~ ^«
«> CO
CO
1 s1
3 =±
CD O
3 .
co o ""
o
CT
o'
CO
o
o
c
3
c
CO
o'
3
o 0 -
co O
Q.
3.
c'
CT
«d
•» -* "c
' -* o c
«. _L O O C
i i i
r~
0
3
CO
«•»
?• co"
3 m"
DCQ' O3
"Z °-
^ CO
o
D 0 3
Q.
D O
D O
D O
.
on
-------
co -n
CD (5'
3 C
" 5 Tissue Nickel (ug/g)
ID 00
ID ' — *•
= O -k O
u o o O
Pi P
0
3
O
CO
m
2
5"
o'
2.
'•^
< P
CO LL ~
o>
5'
0)
5'
^
0
1 1
Q. ^
3 Z -^ -
Q} ^^^
1" >
% <
S. co
3' 33
CD Q)
5» 6"
o
a-
o
CO .
o O ~
c
1
S*
o'
c^
3
o'
2.
Z 0 -
CD f^
a ^^
3-
CD
1 iillltll 1 Iti
o
o
o
1 1ft
Z
3_
OCQ*
2.
O D
O D
O D
D O
D O
r~
0
3
CO
2L
qj
O 3
Q.
O
C
Q.
-
D 0
D
-------
8-9
Field Sediments-Freshwater
Ankley et al. [13] evaluated bioaccumulation of cadmium and nickel bv L. varieaatus
from 17 sediment samples from the upper (freshwater) end of a marine tidal estuary
contaminated by a battery plant. Exposures were conducted for 10 d in a system which
provided 12 turnovers of clean Lake Superior water/d. At test completion, the surviving
oligochaetes were removed from the test sediments and placed in clean water for 24 h
before residue analysis. Brooke et al. [11] have shown that this sampling regime should
result in > 90 percent clearance of gut contents of L. varieaatus. As described elsewhere
(Chapter 6), toxicity of the 17 sediments to the oligochaete was minimal; although 13 of
the 17 samples had SEM (cadmium plus nickel) :AVS ratios greater than one, only at the
two highest ratios was significant mortality observed, which was consistent with pore
water metal TU calculations for the worm. Bioaccumulation of metals (cadmium plus
nickel) from the test sediments by L. varieaatus was not predictable based upon total
sediment metal concentrations (Figure 8-4a), however, bioaccumulation of metals by the
worm did appear to be related to the sediment SEM:AVS ratios (Figure 8-4b). Metal
concentrations in oligochaetes from sediments with ratios less than one were consistently
small; significant bioaccumulation only occurred in those sediments where SEM:AVS ratios
were greater than one. However, marked bioaccumulation of cadmium and nickel by L.
varieoatus was not observed in all samples with ratios greater than one, suggesting
perhaps the presence of additional binding phases in excess of AVS for the metals in some
of the test sediments.
Ankley et al. [14] also conducted a series of longer term sediment bioaccumulation
tests with L. varieaatus. In these experiments, three sediments from the lower Fox River,
Wisconsin were tested; based upon "typical" background concentrations of metals in
freshwater sediments, the three samples had greatly elevated concentrations of cadmium,
copper, zinc, nickel, and lead. However, the sediments also had very high concentrations
of AVS for fresh water sediments (ca., 20 umol/g), with the net result being that
SEM:AVS ratios were in the range of approximately 0.4 to 0.6. Based upon this, Ankley
et al. [14] hypothesized that (a) pore water concentrations of the metals should be low or
-------
4
o
(a)
• •
0.01
0.1 1 10
umol metal/g sediment
100
1.000
«D
"5
(b)
o
0.01
0.1 1 10
sediment metal/AVS ratio
100
i.ooo
Figure 8-4. Comparison of metal (cadmium plus nickel) bioaccumulation by Lumbriculus
varieaatus to (a) total sediment metal (cadmium plus nickel) concentrations and (b)
SEM{cadmium plus nickel) :AVS ratios in 17 field-collected sediments.
-------
8-10
8-4. Comparison of metal (cadmium plus nickel) bioaccumulation by Lumbriculus
varieaatus to (a) total sediment metal (cadmium plus nickel) concentrations and (b)
-------
8-11
non-detectable, and (b) in laboratory exposures, oligochaetes should not accumulate
significant concentrations of the five cationic metals. Tests were conducted for 30 d with
eight renewals of clean overlying (Lake Superior) water/d. The control for the experiment
consisted of worms held in Lake Superior water only. At conclusion of the test, the
oligochaetes were held for 24 h in clean water to purge gut contents. Sediment SEM: AVS
ratios were relatively constant over the 30 d test; all ratios remained less than one. Pore
water concentrations of dissolved cadmium, lead and nickel were non-detectable
throughout the test; however, concentrations of both zinc and copper were consistently
detectable, and higher than background (Lake Superior water) concentrations. The
explanation for the elevated dissolved .copper and zinc concentrations in pore water, in the
presence of excess concentrations of AVS in the sediments, is uncertain. However, at
completion of the 30 d exposure, concentrations of the five metals in tissues of the
oligochaetes were similar to or smaller than control values, i.e., no apparent
bioaccumulation of any of the metals occurred (Figure 8-5). This suggests that if copper
and zinc in the pore water truly were bioavailable and not simply complexed, for example
with DOC, then L. varieaatus must possess some mechanism for effectively regulating
tissue concentrations of these two metals.
Ingersoll et al. [15] conducted a long-term bioaccumulation test with six sediments
from the Clark Fork River, MT using the amphipod Hvalella azteca. SEM:AVS ratios in the
six samples ranged from 0.07 to 960 (Table 8-1), with copper and zinc comprising >97
percent of the SEM. Exposures were conducted for 28 d in a system which provided 1.25
renewals of clean overlying reconstituted water/d. Gut contents of the amphipods were
not purged at test completion. In sediments with SEM:AVS ratios of one or greater,
dissolved pore water concentrations of both copper and zinc were consistently elevated,
while pore water concentrations of the two metals were comparable to control values in
the sediment with the lowest SEM:AVS ratio (Table 8-1). However, as was the case in
the study by Ankley et al. [14], concentrations of both metals also were increased over
control values in two sediments (CF2, CF5) with SEM:AVS ratios slightly less than one.
Concentrations of both copper and zinc were significantly increased in the amphipods from
the three sediment samples with SEM:AVS ratios of one or greater (Table 8-1). Tissue
-------
OlM J9M B/6n)
-------
8-13
concentrations of zinc, but not copper, were significantly higher than control values in tL
azteca from a sediment with an SEM-.AVS ratio of 0.85, while the converse was seen in
sediment with a ratio of 0.75. Neither metal was bioaccumulated in amphipods exposed
to a sediment with a SEM: AVS ratio of 0.07. Because the amphipods were not purged
at test completion, the slight increases in copper or zinc concentrations observed in
organisms from the two sediments with SEM:AVS ratios less than one could have been
due largely to gut contents.
TABLE 8-1 . SUMMARY OF CHEMISTRY AND TOXICITY DATA FOR SIX
SEDIMENT SAMPLES FROM THE CLARK FORK RIVER (CF), AND A
CONTROL.
SEM1:AVS
Site ; (SEM, AVS, //mol/g)
Control
CF1 960 (250, 0.26)
CF2 0.85(16.2,19.1)
CF3 2.15(11.1,5.16)
CF4 0.99(12.9,13.0)
CF5 0.75 (5.88, 7.84)
CF6 0.07 (0.47, 6.66)
Pore Water U/g/LV :
Copper Zincii
4.3
79
36
16
8.7
8.7
1.5
3.9
2603
166
40
28
19
2.0
Tissue (//a/a drv wt)
Copper
80
249*
87
124*
127*
124*
84
j Zinc
57
259*
106*
80*
79*
74
56
] Copper plus zinc.
'Differed significantly from control.
Field Sediments-Marine
Only one study has been conducted using field-collected sediments to evaluate
bioaccumulation of metals by marine invertebrates relative to sediment SEM:AVS ratios
or pore water metal concentrations. Pesch et al. [12] tested tL arenaceodentata with the
same set of samples evaluated by Ankley et al. [13]. Test conditions were essentially the
same as those described above for the polychaete. Results of the study by Pesch et al.
[12] were remarkedly similar to those of Ankley et al. 113]; at SEM (cadmium plus
-------
8-14
nickel):AVS ratios less than one, tissue concentrations of cadmium and nickel were low
and comparable to control values (Figure 8-6). At SEM:AVS ratios greaterthan one, tissue
concentrations of both metals increased markedly. Comparison of the bioaccumulation of
cadmium and nickel by the polychaete also showed a reasonable correspondence with pore
water concentrations of the two metals.
Field Spiking-Freshwater
Hare et al. [4] conducted an experiment in which cadmium was spiked into clean
field-collected sediments, which were subsequently replaced in the field to evaluate trends
in colonization by benthos, as well as bioaccumulation of the cadmium by select taxonomic
groups. The study was conducted in a Precambrian shield lake near Quebec City, Quebec.
Sediments were spiked to achieve nominal SEM:AVS ratios of 0.05 (control), 0.1, 0.5, 2,
and 10. SEM:AVS ratios and pore water cadmium concentrations were measured in 3 cm
horizons of the sediment over the course of slightly more than 1 year, at the end of which
macroinvertebrate samples were collected to evaluate benthic community structure and
cadmium bioaccumulation. Except for the sample with a SEMrAVS ratio of 10, water
overlying the sediments (ca., 1 cm) had non-detectable dissolved cadmium concentrations.
At all the test concentrations, oxidation of AVS in surficial sediments resulted in slightly
greater SEMrAVS ratios in the shallowest horizon than in deeper sections of core samples.
Pore water concentrations of dissolved cadmium in the control sediment and the sediment
with a ratio of 0.1 were consistently low. Cadmium concentrations in pore water from the
sediment with the SEM:AVS ratio of 0.5 were slightly elevated, particularly in the shallow
sediment horizons, likely due to the surface oxidation of AVS. In the two samples with
SEM:AVS ratios greaterthan one, pore water cadmium concentrations in all horizons were
consistently elevated, particularly in the sample with the ratio of 10. Results of the
colonization portion of the study are briefly presented in Chapter 7, and described in detail
by Hare et al. [4]. There was little impact of the spiked cadmium on organism abundance,
however, there was a significant bioaccumulation of cadmium by several different taxa at
SEM:AVS ratios greater than one (Figure 8-7). The source of these metals was not gut
contents, as the animals were purged in clean water for 2 to 5 d before preservation. The
-------
O
h- o
_ o
CP
CO
E
•a
co
DC
CO
T3
O
LJJ
CO
CO)
v
ca
C
to
-a
o
a>
o
ca
£
CO
ca
Q)
a
a
.o
to
i
3
CO
15
"55
CO
D w
||
CO O
CJ 0)
•J5 "P
«
U
CO CM
(6/|oiun)
(IN + PO) I^iaiM enssji
o «-
s =
E.y
4-1
O CO
0 w
cd>
ob<
-------
Abundance Accumulation
600
400
200
0
75
50
25
0
400
Crt 200
o
200
I
I—t-H h-f-H 1—HH 1
10°
C
— o
2000
1000
0
75
50
25
i—n ^ • ' • | ' '^1
Chao §orus
Polycer,tropus
Procladius
•I Ml 111 h-t-
-H-f—l-H-j H-H-j t-H-f-
Procladius sp.c /°"
JOT
Sert'entia
-f-H h-f-H H-H M-f
Chironomus
JP..J
0.6
0.4
0.2
0.0
1.5
1.0
0.5
0.0
10
O
CL
0
u
10
.1 T=
(Q
0.1
10
1
10
1
0.1
0.01 0.1 1 10 0.01 0.1 1 10
Cd/AVS
Figure 8-7. Comparison of SEM(cadmium):AVS to abundance of, and bioaccumulation of
cadmium by six taxa of benthic macroinvertebrates from a field study.
-------
8-17
twotaxa which did not appear to bioaccumulate cadmium (Chaoborus. Polvcentropus) are
both predatory and have limited contact with sediments, while the three taxa which
showed significant cadmium bioaccumulation were all burrowing midges. There also
appeared to be an increase in tissue residues of cadmium in Seraentia from the sediment
with a SEM:AVS ratio of 0.5, which was consistent with the slight increases in pore water
cadmium observed at this ratio. Interestingly, the majority of cadmium accumulated by
this species was associated with gut tissue.
Summary and Conclusions
In an effort to summarize bioaccumulation data from the above studies in a "global"
framework, two different approaches were used. In the first, metal bioaccumulation data
from the various laboratory studies were compared to SEM:AVS ratios in the test
sediments. To normalize for differences in the concentration of metals in control
organisms from the different tests, bioaccumulation data were expressed as the ratio of
the concentration in organisms from test sediments to that in control animals from the
same experiment. Hence, if AVS were critical in controlling metal bioaccumulation by the
various benthic invertebrates tested, this bioaccumulation ratio should be near one at
SEM:AVS ratios less than one, while at S EM: AVS ratios greater than one, bioaccumulation
may be expected to increase. Due to uncertainties about control values and the lack of
gut clearance of the test organisms, data from the study by Carlson et al. [10] were
excluded from this and subsequent analyses.
The bioaccumulation of lead, zinc and copper appears to be explained reasonably
well by a model based on AVS binding, i.e., at SEM:AVS ratios less than one, differences
between control and experimental organisms are minimal, while at ratios between one and
10 bioaccumulation increases (Table 8-2). However, the total dataset for these three
metals, in particular lead, is quite small. By far the most data are available for nickel and
cadmium. The uptake of both of these metals clearly increases with increasing SEM:AVS
ratios; however, the amount of metal bioaccumulated at ratios less than one is higher than
controls (Table 8-2). Interestingly, if data only from field-collected sediments are
-------
. CO
g
§(!
wz
> UJ
< 2
z <
2^
§i
O uj
< cc.
SI
°UJ
CC >•
< cc
sp
WQ
. m
ob -^
UJ
m
h-
CM
00
o
(N
CO
. t
«- CO
cn
CO
5
co SS co
5 S 8
CO
CM
CO
= 8 P
"So
?i "i if
CN CM ^
T^ CO CM
-^ -^ 05
«o t rr
C S •
*= CM
!i 5s.
co ^
-1^1
O o»
0
oi
o
N
•o
s
CO
UJ
CO
^t CM
O CM
i' °'
«- CO
CO
q
3
o
o o
V .1
13
«
m
•O
2
o
o
y
TJ
a
w •;
S >
M V
O 5
o. 5,
m "
*w Ci
> X
°t
*^ >*
2 S
o *••
^5 o
'w JS
•c E E
S o o
Q H- •*-
CO (0 co
+1 "•{5 'i5
1 co re
. ixcc CC
/* T-CS m
-------
8-19
considered, the relationship between metal bioaccumulation and SEM:AVS ratios appears
much more consistent, i.e., little if any bioaccumulation of nickel or cadmium is observed
at SEM:AVS ratios less than one (Table 8-2). It is worth noting that the absolute
concentrations of nickel and cadmium used in the spiking experiments by Pesch et al. [12]
were extremely high; even at SEM:AVS ratios less one, concentrations of the two metals
were as great as approximately 200 and 700 ug/g (dry wt), respectively. Given these test
concentrations, it is possible that even minimal contributions from residual gut contents
and/or surface adsorption could contribute significantly to the total body burden of nickel
or cadmium measured in the polychaete.
Bioaccumulation of the five metals also was compared to measured pore water
concentrations for the various laboratory studies. In this analysis, the use of
bioaccumulation ratios serves not only to correct for different control values for a given
metal, but also to normalize for differences in the propensity for absolute concentrations
of essential trace metals (e.g., zinc) to be naturally higher than those for non-essential
trace metals. Non-detectable pore water concentrations were set at study-specific
detection limits. Regression of the Iog10 bioaccumulation ratios for the five metals versus
the log-io pore water metal concentrations resulted in a significant linear model which
explained approximately 45 percent of the variability in the data (Figure 8-8). The fact that
this model is not more robust could be related to a number of sources of among-study
variation including differences in pore water sampling methodology, differences in
analytical detection limits (particularly between freshwater and marine studies), and
among-species differences in ability to regulate trace metal concentrations.
Although relationships between SEM:AVS ratios and/or pore water metal
concentrations are not as conclusive as for the toxicity data described in Chapter 6, it
appears that trends in the bioaccumulation of divalent cationic metals by benthic
macroinvertebrates in laboratory exposures are clearly related to these two measures of
bioavailability. The one field study that has been conducted to test this hypothesis also
suggests that the long-term bioaccumulation of metals by different taxa can be related to
pore water metal concentrations, which in turn, appear to be controlled by metal-sulfide
-------
Bioaccumulation Ratio
-»• c
P 0 -•• o c
o P -»• o o c
-x -* -*- 0 O 0 C
* - /T^ Itllllll 1 1 1 1 1 1 III 1 Illllll! 1 Illffllll 1 IIIIIIll 1 Illltfl
Tl
(Q*
C
CD
00
t
CO
CD
oaccumulatio
3
O
— *
3
CD
r*
QJ
5T
u
w
•o
o
CD
S
D)
<-f
CD
•^
3
CD
r*
05
^"" *
O
o
0
CD
3
-n
D)
r+
6'
v-> —
P __
^ o -
JD
ff
^^
-30-
13 0
O_
1 —
i
0 _
0
o
.
o
o —
o
o
' ^
0
o
o -
0
o
\J
5 • • £
. — ° ° 5" 8 l
JXD ^_ S «
fs* ' ' tL 73
O ri ^k/^t <"» O m
o ° >*° °° o 5
0 o ^ q.
nifhKD n uQ.
l^r^0* T-l «
°nnb^ 00
O gHP O DO
_Qa !S8n o o° o
D o o
o an
D D D CD
a D D
^03 0
o
D
a
> a o
:H z o
5 ri" W
^ S- &
2. 3
c"
3
-
o
o
DO
o
a o
a
o
-------
8-21
interactions. However, even though this analysis broadly supports the metal bioavailability
model described elsewhere in this document, there are enough inconsistencies in the
various studies described above to warrant future research in this area. In particular, more
studies need to be conducted with field-collected sediments, and in the field. These
studies need to be done with attention to details such as (a) defining appropriate sediment
horizons for sampling pore water metals, and SEM:AVS concentrations, to better
approximate exposure of organisms to metals in aerobic zones of sediments, and
(b)determining the contribution of adsorbed and/or ingested material to the total body
burden of metals in test organisms.
-------
REFERENCES
1. Tessier A. and Campbell P.G.C. 1987. Partitioning of trace metals in sediments:
relationships with bioavailability. Hydrobiologia 149:43-52.
2. Luoma S. 1989. Can we determine the biological availability of sediment-bound
trace elements? Hydrobiologia 176/177:379-396.
3. Hare L. 1992. Aquatic insects and trace metals: bioavailability, bioaccumulation,
and toxicity. Crit Rev. Toxicol. 22:327-369.
4. Hare L, Carignan R. and Hverta-Diza M.A. 1994. A field experimental test of the
hypothesis that acid volatile sulfide (AVS) concentrations improve the prediction of
metal toxicity and accumulation by benthic invertebrates. Limnol. Ocean. Vol. 39.
5. Tessier A., Couillard T., Campbell P.G.C. and Auclair J.C. 1993. Modeling
cadmium partitioning in oxic lake sediments and cadmium concentrations in the
freshwater bivalue Anodonta grandis (Mollusca, Pelecypods). Limnol. Ocean. In
press.
6. Chapman P.M., Churchland L.M., Thomson P.A. and Michnowsky E. 1980. Heavy
metal studies with oligochaetes. in* Aquatic Oligochaete Biology (R.O. Brinkhurst
and D.G. Cook, Eds.). Plenum Publishing Corp., New York, NY. 477-502.
7. Hare L., Campbell P.G.C., Tessier A. and Belzile N. 1989. Gut sediments in a
burrowing mayfly (Ephemeroptera, Hexagenia limbata): their contribution to animal
trace element burdens, their removal, and the efficacy of a correction for their
presence. Can. J. Fish Aquat. Sci. 46:451-456.
8. Krantzberg G. and Stokes P.M. 1988. The importance of surface adsorption and
pH in metal accumulation by chironomids Environ. Toxicol. Chem. 7:653-670.
-------
9. Timmermans K.R., Peeters W. and Tonkes M. 1992. Cadmium, zinc, lead and
copper in Chironomus riparius (Meigen) larvae (Diptera, Chironomidae): uptake and
effects. Hydrobiologia 241:119-134.
10. Carlson A.R., Phipps G.L, Mattson V.R., Kosian P.A. and Cotter A.M. 1991. The
role of acid volatile sulfide in determining cadmium bioavailability in freshwater
sediments. Environ. Toxicol. Chem. 10:1309-1319.
11. Brooke L.T., Call D.J., Poirier S.H., McGovern S.L., Ankley G.T. and Cook P.M.
1993. Gut content weight and content clearance for three species of freshwater
invertebrates. Draft manuscript.
12. Pesch C.E., Hansen D.J., Boothman W., Berry W. and Mahony J.D. 1993. The
role of acid volatile sulfide in determining bioavailability of cadmium and nickel from
contaminated sediments: experiments with Neanthesarenaceodentata(Polychaeta:
Nereidae). Environ. Toxicol. Chem. In press.
13. Ankley G.T., Phipps G.L., Leonard E.N., Benoit D.A., Mattson V.R., Kosian P.A.,
Cotter A.M., Dierkes J.R., Hansen D.J., and Mahony J.D. 1991. Acid volatile
sulfide as a factor mediating cadmium and nickel bioavailability in contaminated
sediments. Environ. Toxicol. Chem. 10:1299-1307.
14. Ankley G.T., Leonard E.N. and Mattson V.R. 1993. Prediction of bioaccumulation
of metals from contaminated sediments by the oHgochaete, Lumbriculus varieaatus.
Water Res. In press.
15. Ingersoll C.G., Brumbaugh W.G, Farag A.M., LaPoint T.W., and Woodward D.F.
1994. Effects of metal-contaminated sediment, water and diet on aquatic
organisms. Water Res. 28:1071-1076.
-------
16. Phipps G.L. Mattson V.R. and Ankley G.T. 1994. Relative sensitivity of three
freshwater macroinvertebrates to five metals and five pesticides Arch. Environ.
Contam. Toxicol. In press.
-------
CHAPTER 9
AVS AND OTHER BINDING PHASES
The previous chapters have addressed the theoretical and experimental evidence in
support of establishing sediment quality criteria for metals using Equilibrium Partitioning.
It was shown that both sediment AVS and interstitial water concentrations are important
in assessing sediment toxicity. Determinations of AVS levels in sediments and interstitial
water concentrations will play an important role in the application of SQC for metals.
This chapter addresses some additional considerations pertaining to characteristics
of AVS sediment distributions and other factors that are important in the application of
SQC for metals. These issues include the seasonal and depth variability of AVS, and the
correlation of AVS to sediment organic carbon. Oxidation kinetics of iron sulf ide and metal
sulfide are presented. Experiments and their results pertaining to organic carbon binding
for copper, cadmium, and lead are discussed. Experimental results for metals in low
organic carbon keep sediments are presented to define a minimum partition coefficient.
Lastly, pore water and SEM and AVS sampling are discussed.
Vertical and Seasonal AVS Distributions
Several factors contribute to variability in AVS distributions. The amounts of
organic matter, sulf ate, iron, and oxygen in sediments affects the potential amount of AVS
that can be formed. An example of this is the difference in AVS concentrations in
freshwater versus saltwater systems. Sulfate is present at higher levels in saltwater
systems and the AVS concentrations can be expected to be higher in these sediments.
A literature review by Leonard et al. 11], found that concentrations of AVS in unpolluted
freshwater sediments were in the range of about 4 to 13 umol S/g, while AVS
concentrations in coastal marine sediments have been reported in the range of about 0.1
to 100 fjmollg 12]. AVS also varies with depth and season. AVS concentrations can
increase with depth up to about 20 cm and then decrease with depth. The following
chapter summarizes three studies that have explored both seasonal and vertical
-------
/„
distributions of AVS. Two studies were done in freshwater systems and one study was
done in a marine system.
Leonard et al. [11 measured seasonal and depth dependent AVS variations in three
freshwater lakes in northeastern Minnesota; Caribou Lake, Fish Lake, and Pike Lake. The
lakes were sampled approximately monthly from May 1990 to September 1991 for a total
of 16 months. Sediment cores were sectioned into three 15 cm sections to represent the
0 to 15 cm depth, 15 to 30 cm depth and 30 to 45 cm depth. Particle size, total organic
carbon and pore water pH and ammonia were measured for each 15 cm section.
Overlying water was sampled for pH, alkalinity, hardness, conductivity, dissolved oxygen
and primary productivity.
Temporal profiles of overlying water temperature and AVS at the three depths for
each of the three lakes are presented in Figure 9-1. Average AVS concentrations in the
0 to 15 cm segments were <0.1 to 9.8 umol S/g in Caribou Lake, 0.1 to 6.0 umol S/g in
Fish Lake, and 1.3 to 36.2 umol S/g in Pike Lake. Variability in AVS concentrations was
most dramatic in the top 0 to 15 cm with less variability in the 15 to 30 cm and 30 to 45
cm segments in each of the lakes.
Leonard et al. 11 ] found that AVS in the upper two segments of Fish and Pike lakes
was correlated to overlying water temperature and though not statistically significant, AVS
did vary with overlying water temperature in Caribou Lake as well. During periods of ice
cover overlying water temperatures were 0.2 to 1.5°C as shown in Figure 9-1 and AVS
was at lowest levels (0.1 to 2.0 umol S/g). AVS increased to maximum concentrations
as overlying water temperature increased to 20 to 25°C. Generally, when overlying water
temperatures were at highest levels in June through August so were the AVS levels. AVS
decreased with the onset of ice cover. Leonard et al. [1] note that the generation of AVS
in the summer can be expected as sulfate reducing bacteria have an optimal temperature
for growth in the 15 to 20°C range and a minimal temperature for growth at 0°C [31.
Figure 9-1 shows that AVS concentrations in the 15 to 30 cm depths were less influenced
by temperature and almost no changes were seen in the 30 to 45 cm depths.
-------
20
0-15 15-30 30-45 • TempC
o— —a ••••<>••••
Figure 9-1. Seasonal AVS and temperature profiles for Caribou Lake, Fish Lake, and Pike
Lake, Minnesota at 0 to 15 cm, 15 to 30 cm, and 30 to 45 cm. Periods of ice cover are
indicated. Note that sampling dates (xaxis) are not to scale. Source: [1].
-------
9-4
Boothman and Helmstetter [4] studied the vertical and seasonal variability of AVS
in uncontaminated marine sediments. Sediment cores were collected in the
Pettaquamscutt Cove in Narragansett, Rhode island five times between July 1990 and
May 1991 and biweekly in June, July, and August 1991. The top 15 cm were collected
and sliced into sections of 1 cm. Sections of 1 cm were analyzed for the top 5 cm as well
as the 9 to 10 cm and 14 to 15 cm sections.
AVS distributions are presented in Figure 9-2. Replicate cores for 5 of the 11
sampling dates are represented by dashed lines. A general trend in depth variability is seen
for each of the sampling periods. AVS is always lowest in the top 1 cm slice which may
indicate the oxic sediment layer. Then AVS increases generally to the 10 cm slice and
remains fairly constant from the 10 cm section to 15 cm section. Most of the variability
is seen in the top 0 to 5 cm.
The profiles in Figure 9-2 indicate a seasonal variability in AVS levels. AVS
concentrations in the top 5 cm increased from June to August with concentrations of 15
to 35 umol/g. The winter cores (top row) show lower AVS concentrations (10 to 25
umol/g) except for one measurement in the January 1991 core of about 30 umol/g.
Boothman and Helmstetter [4] also found a strong correlation of AVS levels with overlying
water temperature. They attribute vertical and seasonal variability of AVS in an
uncontaminated marine sedimentto two competing processes: microbial diagenetic sulf ate
reduction and oxidation of sulfides. In oxic water bodies diffusion of oxygen into the
sediment from overlying water results in sulfide oxidation. Sulfide oxidation occurs in both
the colder and warmer periods but rates of microbial activity and the production of sulfides
are much greater, so that AVS levels increase in the summer months [ 1 ]. Sulf ide oxidation
is dependent on the presence of oxygen, diffusion rates, bioturbation, and the oxidation
potential of the metal sulfides. A discussion of the oxidation of metal sulfides will be
presented subsequently.
An AVS seasonal and spatial study was done using, three freshwater lakes having
seasonally anoxic hypolimnia and varying periods of stratification; Crosson Lake,
-------
umol AVS / g
JJ25,tS3
3 a
o «2)3}4}0fl2i3}«oi92]a]«otizia«}
-------
9-6
Gullfeather Lake, and Jake Lake, in Ontario [5]. For the spatial comparison sediments in
all three lakes were sampled for AVS at various water column depths from the littoral zone
to the main depositional basin of each lake during periods of anoxia. Crosson and Jake
Lakes were also sampled during periods of no stratification. AVS results were reported
as the mean of three replicates from the top 15 cm of sediment. Figure 9-3 presents AVS
concentrations at various water column depths for the three lakes and comparisons of the
anoxic (September) and oxygenated (May or August) periods at various water column
depths for Crosson and Jake Lakes. Figure 9-3 indicates that AVS concentrations increase
at greater overlying water depth in all three lakes during bathanoxic and oxygenated
periods. AVS concentrations were greatest during periods of anoxia.
For the AVS seasonal analysis, Jake Lake was sampled for AVS bimonthly from
mid-April 1991 to mid-May 1992 at a lake depth of 21m. Figure 9-4 presents a temporal
distribution of AVS. Turnover events and periods of anoxia as measured at the
sediment/water interface are indicated. AVS was highest during periods of anoxia and
began decreasing to a minimum following lake turnover. This is expected since as oxygen
in the overlying water is depleted oxygen diffusion into the sediment ceases and sulfate
reduces to sulf ide thereby increasing the AVS. The key point here is that as the sediment
becomes anaerobic and in the presence of sulfate and sulfate reducing bacteria AVS will
form. This study did not report overlying water temperature. However the lowest AVS
levels occur in May 1991 and December-January 1992 which is in agreement with other
seasonal studies [1,4].
A 40 cm core was taken in Jake Lake on August 12, 1992 when the lake had just
become anoxic to establish a profile of AVS concentration with sediment depth. Figure
9-5 presents this profile. The AVS sample at the sediment/water interface was less than
1.0 umol S/g wet sediment then increased to a high of 7.1 umol S/g wet sediment at
about 8 cm. AVS concentrations showed less variability below 15 cm.
These three studies indicate the variability that can be expected in vertical and
seasonal AVS concentrations. The studies indicate that AVS is low in the top 1 cm of
-------
1.00
o.so
o 0.60
~o
W 0.40
0.20
0.00
Crosson Lake
X
S 10 15 20
Depth at Sampling Site (m)
August 8/91 Sept 30/91
25
c/i
'
o
w
1.00
0.90
0.80
0.70
0.60
O.SO
0.40
0.30
0.20
0.10
0.00
10
Gullfeather Lake
3 6 9 12
Depth at Sampling Site (m). August 8/91
15
Williams Bay
0 5 10 15 20 25
Depth at Sampling Site (m)
Sept 17/91 May 12/92
Figure 9-3. Spatial profiles of the top 0 to 15 cm of sediment AVS from the littoral zone
to the main deposition^ basin in Crosson Lake (top l^^'^
and Williams Bay, Jake Lake (bottom panel), Peterborough, Ontario.
Lake were taken on August 8,1991, before the hypohmnion went anox.c,
30 1991, during the anoxic period. Samples for Gullfeather Lake were taken on August
8, 1991, when the hypolimnion was anoxic. Samples for Williams Bay were taken on
September 17, 1991 during the anoxic period, and May 12, 1992 just after spr.ng
turnover. Error bars represent one standard deviation. Source: [51.
-------
M
5
o>
"o
E
3
00
10
8
-I
AMJJASONDJFMAM
1991 Sampling Date 1"2
Figure 9*4. Seasonal variation in AVS in Williams Bay, Jake Lake at water depth of 21 m.
Periods of anoxia at the sediment water interface are indicated by A and turnover events
are indicated by T. Bars indicate one standard deviation. Source: [5].
o.
0>
O
§ -24
•o
o
W
-32
.40
AVS umol sullide / g w.s.
1234
Figure 9-5. Vertical profile of AVS in Jack Lake at water depth of 21 m. Source: [5].
-------
9-9
sediment. Presumably, oxygen at the sediment/water interface oxidizes the sulfides. The
highest AVS levels are seen in the top 1 to 10 cm but then decrease and show less
variability below 10 to 15 cm. In oxic water bodies AVS generally correlates to overlying
water temperature. Microbial degradation of organic matter to produce sulf ate and the
reduction of sulf ate to sulf ide increases with increasing overlying water temperature and
decreases as overlying water temperature decreases. As a result AVS can be expected
to increase with increased water temperature and decrease as temperature decreases.
Suifide oxidation, which is dependent on the diffusion of oxygen into the sediment and the
oxidation potential of the metal sulfides, also plays a role in vertical and seasonal AVS
variability. In water bodies that have periods of anoxia AVS seems to correlate to
overlying oxygen levels. Overlying water temperatures seem to affect AVS concentrations
in the anoxic system presented.
The Correlation of AVS to Sediment Organic Carbon
Data from the US EPA Environmental Monitoring and Assessment Program
(EMAPH2] were summarized to examine possible relationships between total organic
carbon (TOC) and AVS in marine sediments. Data are from the Gulf of Mexico (Louisiana
Province) and the other from the Mid-Atlantic Coast (Virginian Province). The EMAP data
was collected in the last week of July through the first week of September 1990 through
1992. The data represent the top 2.0 cm.
In the EMAP data set TOC ranges from 0.1 to approximately 10 percent. The
median TOC is approximately 1 percent. AVS ranges from approximately 0.1 to 100
umol/dry weight sediment with a median of approximately 3 umol/g. TOC and AVS
measurements from the Virginian Province and the Louisianan Province had similar
distributions.
X
AVS showed a positive correlation with TOC for the EMAP data. A linear regression
of the log transformed AVS and TOC is shown on Figure 9-6 (solid line). The equation that
relates TOC to AVS derived from this data set is:
-------
(fi
0)
D
O)
u
<
1000
100
10
- A
0.1
0.01
I I I III I I I I I I I t=
0.01
I I I I I I I I I I I I I I I II I I I I I I I I I
0.1 1
Total Organic Carbon (%)
10
in
CD
TD
en
.--i O
4-> e
03 D
u
<
1000
100
10
0.1
0.01
i r i i i
i i i i f i m
r i i i i i
E B
I I I I I I I I I I I 1 L11JJJ L I I I I I I I
0.01
0.1
10
Total Organic Carbon (%)
Figure 9-6. AVS versus percent TOC using EMAP12] data, last week of July to first week
of September 1990 to 1992. Data have been plotted by intervals having equal number
of values. Horizontal bars represent TOC standard deviation and vertical bars represent
AVS standard deviation. Line represents linear regression of log transformed AVS and %
TOC.
-------
9-11
Log AVS = 0.447 + 0.865 Log TOC (9'1)
The individual data from the Louisiana (L) and Virginia (V) Provinces are shown in Figure
9-6A. The data grouped into 10 intervals with an equal number of values and averaged
is shown in Figure 9-6B. Although there is a roughly linear relationship between AVS and
TOC (log slope = 0.87 which is almost 1.0), there is an enormous scatter. The lines in
Figure 9-6 are a factor of 10 and 0.1 of the regression line as a visual aid to illustrate the
variability in the data. Therefore, a prediction of AVS based on sediment TOC is highly
uncertain. These data are in agreement with the results of Ankley et al. [6] who found no
significant correlation between AVS and TOC in 17 sediment samples from an estuarine
system contaminated with cadmium and nickel.
The basis of this analysis is that organic matter contributes to the formation of
sulfides. Organic carbon exists in sediments in refractory and reactive forms. It is the
reactive organic matter present in sediments that contributes to the formation of AVS. It,
in turn, is the net result of the input of TOC by primary production plus terrestrial sources
of TOC less particulate organic carbon (POO loss by burial. Thus, the residual POC is a
good index of the average concentration of reactive organic carbon but not the seasonal
variation. The result In Figure 9-6A is not surprising considering the variability of AVS that
has been shown in the previous discussion. By contrast, organic carbon concentrations
are quite constant with respect to time [7]. The reason is that most of the organic carbon
in sediments is very refractory with only a small percentage that is reactive [8].
Oxidation of Metal Sulfides
The oxidation of metal sulfides is an important component in the analysis of the
ultimate fate and toxicity of metals in sediments. The oxidation of iron sulfide (FeS)
controls the amount that is available to complex newly deposited metals. Also the
seasonal cycle of AVS in natural sediments is controlled by a balance between the
formation of AVS via the oxidation of organic carbon with sulfate as the electron acceptor,
and the oxidation of AVS with oxygen as the oxidant. Thus a knowledge of the oxidation
-------
9-12
kinetics and their interaction with other mass transport mechanisms in sediments is
important to a proper use of sediment quality criteria based on AVS.
FeS(s) and CdS(s) Oxidation Kinetics
The rates of oxidation of iron and cadmium sulfides have been studied extensively
at low pHs in the mining literature [91. However, the oxidation of FeS in more natural
settings has not received much attention with one notable exception. Nelson [10] studied
the oxidation of synthetic FeS under a wide variety of conditions. Variations in pH, oxygen
concentration, ionic strength, temperature, and the presence of catalytic metals were
examined. His focus was on the initial rate of reaction. He proposed a surface
complexation model which fit his experimental results quite successfully.
The entire time course of the reaction for synthetic FeS using Nelson's data and for
AVS in sediments from other experiments, has recently been completed and a model for
the kinetics of the oxidation has been proposed [11]. The model is based on the surface
oxidation of FeS, as proposed by Nelson. The particles are assumed to have various
particle size distributions. In particular a uniform and an exponential distribution of surface
areas are considered. The equations and the solutions are listed in Table 9-1.
The results of fitting the model to synthetic FeS and sediment AVS are shown in
Figure 9-7. The two model parameters are the zero order surface reaction rate, K, and the
coefficient of variation of the particle size distribution. The symbols represent initial
concentrations of FeS and AVS. The results indicate that the reaction rate is virtually the
same for sediments and synthetic FeS at the same pH (Figures 9-7A, C, D). This suggests
that all of the experimental information that has been generated for the pH, temperature,
and 02 dependence of the oxidation rate (Figure 9-8), is applicable to sediment AVS. In
particular, the cadmium and iron oxidation kinetics used below assume that the oxidation
rate is linear with respect to oxygen. The data in Figure 9-8B verify that assumption for
synthetic FeS.
-------
Table 9-1
FeS Oxidation Model
Population balance equation:
where n(A,f) « number of particles with surface area in the interval A and A * dA, at time t.
For a zero order rate of surface oxidation:
the particle balance equation becomes:
— A- —-o
For an initial particle size distribution: n 0 (A), the solution is:
The concentration at any time t is found using the surface area concentration • surface area
relationship:
O
I
Jo
where Y is the volume to surface area exponent For a uniform number density:
-0 elsewhere
The result is:
c(0"c°
where p( /I), the positive function, is the positive portion of its argument A. Using the mean, n,
and coefficient of variation, v, of the uniform density yields:
where fr-fc'/n. The parameters are Y, the volume to surface area exponent; v, the coefficient
of variation of the number density; and k, the decay rate.
-------
Synthetic FeS
pH
K = 0.737 hr
v = 0.0058
-1
2 3
Time (hr)
pH = 9
K = 0.462 hr '
v = 0.382
Sediment AVS
Jamaica Bay (pH=7)
K = 0.658 hr
v - 0.572
-1
Van Cortlandt Pond (pH=7)
Time (hr)
Figure 9-7. FeS oxidation polydisperse particle distribution model fitting synthetic FeS and
sediment AVS. The zero order surface reaction rate, k and the coefficient of variation of
the particle size distribution, v are shown for synthetic FeS at pH 6 and 7 and AVS for two
sediments; Jamaica Bay and Van Cortlandt Pond.
-------
FeS Oxidation Rate
100
i
pH Effect
6 7 8 9 10 11
PH
50
40
30
20
10
Dissolved Oxygen Effect
B
0 10 20 30 40 SO
02 (mg/L)
CdS Oxidation
15
-= 10
Synthetic CdS
K = 0.01 /d
5 10
Time (hr)
15
Cd Spiked Sediments
CO
Figure 9-8. The effect of pH (Panel a) and dissolved oxygen (Panel b) on the FeS oxidation
rate. Oxidation results of metal-sulfide oxidation experiments using synthetic CdS (panel
c) and cadmium spiked sediments.
-------
9-16
The initial results from a series of metal-sulfide oxidation experiments using
synthetic CdS are presented in Figure 9-8C. The reaction rate is much slower than the
rates found for FeS (Figure 9-7A). Cadmium spiked sediments show the same slow
oxidation rate (Figure 9-8D). For Cd/AVS = 0, the AVS is all FeS and the oxidation is
rapid. When a fraction (Cd/AVS = 0.5) or all (Cd/AVS = 1.2) of the FeS is converted to
CdS by adding cadmium to the sediment, that fraction of the AVS does not oxidize in the
time scale of this experiment. Also, Zhuang et al. [12] conducted a series of laboratory
aeration experiments in batch reactors to investigate the effects of aeration of sediment
on the sulfide content of sediment and on the partitioning of cadmium to the sediment.
Aeration of the sediment resulted in a rapid decrease in AVS. Concentations of dissolved
cadmium increased while concentrations of cadmium associated with AVS and with pyrite
decreased.
Sediment Metal Oxidation Model
The rate at which cadmium sulfide oxidizes in sediments depends not only on the
rate it oxidizes in an aerobic environment but also the rate at which FeS oxidizes, which
controls the depth of the aerobic layer. This is the mechanism that can cause the
cadmium-AVS molar ratio in the surface layer of the sediment to change from a value of
less than one to a value of greater than one, with the possible concomitant increase in
toxicity.
The oxidation kinetics of FeS and CdS are a necessary part of a model of the
oxidation and release of metals from sediments. Work has been proceeding to develop
such a comprehensive model. The model is based on a sediment flux model that has been
developed for oxygen and nutrients [61. The formulation for the cadmium flux model is
illustrated in Figure 9-9. The sediment is idealized as having two layers: (1) an aerobic
layer (1) where the oxygen concentration is greater than zero; and (2) an anaerobic layer
where the oxygen is zero. Sulfate reduction in this layer produces sulfide which interacts
with the iron present to form iron monosulfide (FeS).
-------
o *-
CQ
e
a
UJ
5
UJ
CO
3 «
o c
CC UJ
WATER COLUMN
i
SURFACE MASS TRANSFER: KL01 f Cd2+(1)
PARTITIONING: Cd2+
Cd=SS
kCdS,1
OXIDATION: CdS(s) *- Cd2+
PARTICLE MIXING
•• W12
DIFFUSION
KL12
PARTITIONING: Cd2+
Cd=SS
PRECIPITATION:
Cd2+ + FeS(s)
SEDIMENTATION
w2
kCdS,2
CdS(s) + Fe2+
Figure 9-9. Cadmium flux model. The sediment is represented by an aerobic layer and an
anaerobic layer.
-------
9-18
Cadmium enters the sediment either by surface mass transfer from the overlying
water, KL01, or as particles settling to the sediment. The reactions in the aerobic layer are
the partitioning of cadmium to the sediment solids. Particles and dissolved cadmium are
biotically and abiotically, transported to the anaerobic layer where cadmium sulfide is
formed. Cds transport to the aerobic layer is either by particle mixing, w12, or by diffusion
of dissolved cadmium, KL12. Cadmium sulfide oxidation produces Cd2* which, after
partitioning, escapes from the sediment as a flux to the overlying water via surface mass
transfer.
A one dimensional model for cadmium and iron sulfide, oxygen, and dissolved +
sorbed cadmium is based on the mass balance equations listed in Table 9-2. The
mechanisms for cadmium transport and kinetics are shown in Figure 9-9. Models of this
sort have been employed before [13]. The novelty here is the inclusion of the sulfide
reactions. The kinetics of oxidation of both iron and cadmium sulfide are first order in
oxygen and sulfide concentration. The data used to justify these kinetics are shown in
Figures 9-7 and 9-8. Cadmium sulfide oxidizes at a rate: kCdS[02][CdS] (Table 9-1;
Equation 9-1) consuming oxygen (Equation 9-3) and liberating cadmium to the interstitial
water and therefore becoming a source to the dissolved + sorbed cadmium (Equation 9-4).
The fraction that is dissolved is determined by the partitioning expression (Equation 9-7)
which affects the magnitude of the diffusion coefficient of total sorbed + dissolved
cadmium. Iron sulfide oxidizes at a rate: kFeS[O2HFeSl (Equation 9-2) and consumes
oxygen (Equation 9-3). The formulation for particle mixing by bioturbation is conventional
[14] - particle diffusion that exponentially decays in depth with characteristic mixing depth
ZB (Equation 9-5). A source of FeS, J02 from organic matter diagenesis, is added (Equation
9-2) to account for the generation of AVS during the experiment.
The solutions to the equations in Table 9-2 are obtained using an implicit finite
difference formulation with the nonlinear terms lagged by one time step. The model
vertical resolution is 1 mm. The parameters of the model are the oxidation rates: kcds and
kFeS, and the mixing parameters for particles, Dp and ZB, and interstitial water, Dd, and the
AVS source, J02.
-------
_ TABLE 9-2. MODEL OF CADMIUM AND AVS DISTRIBUTION _
Mass balance equations for cadmium sulfide, CdS(z,t), iron sulfide, FeS(z,t) dissolved
oxygen, O2{z,t) and dissolved + sorbed cadmium, CT:
aiCdSl . ^Dp^S! -kcdsK>2HCdS] (1)
_p -kFeS[02]lFeS] * » (2)
_kcdslo2ncds] -kFeS[o2j[FeS]
kcdS[02][CdS] (4)
where kcds and kFeS are the oxidation rates of CdS and FeS, J02, is the source of FeS from
organic matter diagenesis, H is the depth of the sediment, and:
_ z
Dp = DPoe ZB (5)
is the particle mixing diffusion coefficient, ZB is the characteristic depth of bioturbation,
Dd is the diffusion coefficient in porewater, and:
DT = DPfP +Ddfd (6)
is the weighted diffusion coefficient for dissolved + sorbed cadmium. The dissolved, fd,
and paniculate, fp, fractions are:
fd = T— = 1 -fP (7)
1 + m/7
with m = solids concentration and n = dissolved cadmium partition coefficient.
-------
9-20
A set of data from the colonization experiment (Chapter 7) designed to test the
toxicity of the sediment can be used for an initial evaluation of the model [15]. Raw
seawater is allowed to flow over sediments that have been spiked uniformly with various
concentrations of cadmium so that the initial concentrations of AVS and cadmium are
constant with depth. The larval forms of benthic organisms settle and colonize the
sediment. The experiment was carried out for 118 days. The vertical distribution of AVS
and cadmium were measured in 6 mm slices, the smallest practical interval. Four
concentrations of cadmium were dosed into the sediment: 0, 0.1, 0.8, and 3.0 times the
moes of AVS in the sediment. The vertical profiles were measured at day 28, 57, and
118.
The observations and model computations are shown in Figure 9-10 for AVS, Figure
9-11 for cadmium, and Figure 9-12 for SEM to AVS ratio. The columns correspond to
each of the three sampling days. The rows correspond to progressively increasing
cadmium concentrations. The vertical distributions of AVS and cadmium are reproduced
reasonably well by the model except for the bottom row where the Cd/AVS = 3 and the
sediment was toxic to benthic biota. Presumably in this treatment the rate of bioturbation
decreased and less oxidation occurred than predicted. The computed vertical distribution
of dissolved oxygen is also shown as a dashed line in the AVS plot (Figure 9-10). The
critical result is that in order to reproduce the vertical distribution of AVS it is necessary
to mix the particles. Presumably this is the result of bioturbation by the organisms that
colonized the sediment. The presence of significant impacts on many benthic organisms
(Chapter 7) and lack of model fit for Cd/AVS = 3 supports this hypothesis. The model
also reproduces the trend of increasing SEM/AVS ratio that occurs in the top 1 to 2 cm
of the sediment, although the magnitudes of the calculated increases are not as large as
is observed in the topmost 6 mm in the 0.8x treatment and larger than observed in the
0.1x treatment.
The parameters used in this simulation are listed in Table 9-3. In particular the rate
of FeS oxidation is consistent with the kinetics obtained from synthetic and sediment AVS
-------
0
20
I"
80
too
Cd/AVS=0 28 days
Cd/AVS=0 57 days
10 20 30
0
20
I 40
60
80
100
Cd/AVS-O.I
28 days
10
20
30
Cd/AVS=0.8 28 days
u
O
u
20
40
60
80
nn
!>-"V —
*\
4
•
k
t
10
20
30
0
20
<°
60
80
100
Cd/AVS=3.0 28 days
10 20
AVS (umol/g)
30
Cd/AVS=0 118 days
100
0
20
Cd/AVS-0.1
57 days
I
I"
80
100
10
20
30
0
20
I •
Jl
£ 60
O
80
100
Cd/AVS=0.8
57 days
10
20
30
"g
j
O.
V
O
Cd/AVS=3.0 57
20
40
60
80
inn
' "\ *
X *
L \ *
\ *
\
\ *
-
*
0 10 20
days
•
3
20
1 40
60
SO
100
0
20
60
80
100
10
20
30
Cd/AVS=0.1 118 dayt
10
20
30
Cd/AVS=0.8 118 days
0
20
40
60
80
100
Cd/AVS=3.0 118 days
AVS (uraol/g)
10 20
AVS (umol/g)
30
Figure 9-10. Vertical distributions of AVS for the field colonization experiment at days 28,
57, and 118 for nominal treatments of 0., 0.1,0.8, and 3.0 Cd/AVS ratios (symbols) and
cadmium flux model results (solid line). The dashed line is the computed vertical
distribution of dissolved oxygen.
-------
Cd/AVS=0
28 days
Cd/AVS=0 57 days
Cd/AVS=0 118 days
"l
1
1
| 40i
£ 60i
0
80'
100
0
20
1 «
I-
80
,00(
0
20
I «
5
5 60
Q
80
100(
0
20
| 40
| «
80
inn
1 1 2 3 4 5
Cd/AVS-0.1 28 days
*
*
*
*
*
*
112345
Cd/AVS=0.8 28 days
~*\
it
*
*
*
*
*
) 10 20 3
Cd/AVS-3.0 28 days
~A
*
*
*
v:
20
1 ^O
S '
£ 60.
Q
80
,oo€
0
20
1 «
! 60
O
80
,ooc
0
20
1 «o
*»
? 60
Q
80
o 100(
0
20
E 40
g- 60
O
80
inn
-
r
-r
*
12345
M/AVS-0.1 57 days
i
i
*
*
*
*
12343
Cd/AVS=0.8 57 days
"V
>
t
*
u.
20'
1 4°
g- 60
Q
80
100,
0
20
I «
& 60
80
100(
0
20
1 40
g- 60
Q
80
•
1 2 34 5
Cd/AVS-0.1 118 days
V
1
*l
112345
Cd/AVS=0.8 118 days
"\i
\
1
) 10 20 30 """O 10 20 30
Cd/AVS=3.0 57 days
A-
V
\
•1
0
20
B 40
1 "
80
inn
Cd/AVS=3.0 118 days
^\ *
\ *
\ *
\ *
\
A
1
0 20 40 60 80 100
Cd (uraol/g)
0 20 40 60 80 100
Cd (umol/g)
"0 20 40 60 80 100
Cd (umol/g)
Figure 9-11. Vertical distributions of cadmium SEM for the field colonization experiment
at days 28, 57, and 118 for nominal treatments of 0., 0.1, 0.8, and 3.0 Cd/AVS ratios
(symbols) and cadmium flux model results (solid line).
-------
SEM/AVS=0 - 28 days
SEM/AVS=0 118 day*
u
20
! 40
JC
1 60
Q
80
0
0
20
1 «
& 60
O
80
iooo
0
20
| 40
| 60
O
80
iooo
0
20
| 40
8 60
Q
80
inn
•
1 1.0
SEM/AVS-
10
0.1 - 28 days
•
1 1.0
10
SEM/AVS=0.8 - 28 days
*
-
*
\t *
1 1.0
10
SEM/AVS=3.0 - 28 days
*
*
*
*
*
*
-
-
•
•
u
20
I 40
8 60
Q
80
.0
0
20
] 40
8 60
Q
80
100
.0 0
0
20
- I <°
& 60
O
80
.0 100o
0
20
1 <°
& 60
Q
80
inn
i
1
1.0
10
SEM/AVS-0.1 - 57 days
^
1
SEM/AVS=
*
*
*
1
SEM/AVS=
1.0
10
0.8 - 57 days
' -
1.0
10
3.0 - 57 days
*
*
*
*
•4
*
*
•
•
•
•
u
20
1 40
g- 60
Q
80
.0 100o
0
20
i 4o
8 60
Q
80
.0 100o
0
20
I <°
& 60
Q
80
.0 100o
0
20
1 «
8 60
Q
80
inn
i
1 1.0 10
SEM/AVS-0.1 118 days
* .^^
f
•
1 1.0 10
SEM/AVS=0.8 118 days
1 i *
/
J
1 1.0 10
SEM/AVS=3.0 118 days
k
*
i
4
*
: T
i
0.1 1.0 10.0
SEMAVS (umol/umol)
0.1 1.0 10.0
SEMAVS (umol/umol)
0.1 1.0 10.0
SEMAVS (umol/umol)
Figure 9-12. Vertical distributions of SEM/AVS ratios for the field colonization experiment
at days 28, 57, and 118 for nominal treatments of 0., 0.1, 0.8, and 3.0 Cd/AVS ratios
(symbols) and cadmium flux model results (solid line).
-------
9-24
oxidation experiments reported above. The oxidation rate of cadmium sulfide is not
inconsistent with the results of the initial experiments.
A preliminary application of this model to data from a year long field colonization
experiment with zinc contaminated sediments was performed [16]. The purpose of the
field colonization experiment was to investigate the seasonal realtionships between AVS
and the toxicity of zinc to benthic organisms. The FeS oxidation rate and particule mixing
parameters applied in the calibration were similar to those used in the cadmium
colonization model calibration (Table 9-3). The zinc oxidation rate was then calibrated to
the data. Data (symbols) from the sampling after 92 days (Figure 9-13) and near the end
of the experiment (327 days, Figure 9-14) illustrate the fit of the model (solid line) to the
data. A very low zinc oxidation rate (Table 9-3 which is indistinguishable from zero)
provided the best model fit. The results indicate that very little zinc sulfide has oxidized.
TABLE 9-3. SEDIMENT MODEL
Cd (Colonization Cd (Effect of oligochaetes
Experiment) [15) experiment) (171
Dd (cm2/d)
Dp (cm2/d)
ZB (cm)
k FeS (L/mg 02-d)
kMS (L/mg O2-d)
IT, = i72 (L/kg)
Jn? (am/m2d)
1.0
0.1
5.0
3.4
0.01
300
0.2
1.0
.01
5.0
3.8
0.01
300
0
Zinc (Colonization
Experiment)! 16)
0.1
0.1
5.0
3.4
1 x 10'7
300
0.2
An initial experiment which was designed to emphasize the effect of bioturbation
has been performed using oligochaetes [171. It is well known that these organisms can
rework the surface sediment which causes an increase in AVS oxidation. Initial results
from these experiments are shown in Figure 9-15 which presents the vertical profiles of
AVS and cadmium with and without oligochaetes added to the surface of the sediment.
The lines are computed using the same coefficients as before (Table 9-3) with the
exception of particle mixing. For the case without oligochaetes it is zero. For the case
with oligochaetes it is a factor of 10 lower than the colonization experiment, reflecting the
-------
0
2
ZB
r\
*-™ t 1 C
ll 1 -_- D
O
8
10
\*
. V
\
.
0
2
4
6
8
10
1
i
-
0
2
4
6
8
10
y
. r
. / •
.
. .....j
.
»
-
. ...,.
10 15 20
0
2
4
6
8
10
\*
A*
. i*
5 10 15 20
0
2
4
6
8
10
v. -
A
. V
.
0 5 10 15 20
AVS (umole/g)
0 5 10 15 20
ZINC SEM (umole/g)
0.01 0.1
0
2
4
6
8
10
' '•?
i
r
.
^ i HIM! • i iiiw
m
.
.
.
1 1 llflP
0.01 0.1 1
10
0
2
4
fi
8
10
, ™
y
i
.
t i itmJ • i mm
m
.
.
.
i I IIIH
0.01 0.1
10
2
4
6
8
10
(
0
2
4
6
8
10
(
0
2
4
6
8
10
'*\
. *\
. *\
3 5 10 15 2
,...._
: \ :
•
3 5 10 15 2
•\
* 1
•
2
4
6
8
10
0 (
0
2
4
6
8
10
0 (
0
2
4
6
8
10
4
. *
.1
•
) 5 10 15 2
. »
•
-
3 5 10 15 2
•
* j
•
2
4
6
8
10
0 0.
0
2
4
6
8
10
0 0.
0
2
4
6
8
10
7«
/*
r
1 1 lltJ « 1 IIIW
01 0.1
.
•
.j «
01 0.1
.
•
m
.
-
1 1 1""
1 11
_
t
•
1 1
•
I
* .
1 1 INP"
0.01 0.1 1
SEM/AVS
10
Figure 9
the zinc
-14. Vertical distributions of AVS, zinc SEM and SEM/AVS ratios at day 327 of
field experiment for the control and five treatments.
-------
i>\
• \
• V •
. r
.
0
2
. 4
6
8
10
»
I
.
• i i i i i |
5 10 15 20
5 10 15 20
8
10
\
0 5 10 15 20
AVS (umole/g)
5 10 15 20
5 10 15 20
2
4
6
8
10
•
^
4
_
.
.
' , ,
0 5 10 15 20
ZINC SEM (umole/g)
n
2
4
6
8
10
f?
-•/
. 1
.
1 t IIIM^ f 1 II.M
^
.
.
.
.111,.
0.01 0.1
0.01 0.1 1
10
2
-1
UJ.B 6
o
8
10
c
0
2
-1
UJii 6
O
8
10
(
0
2
-1
a u c
LU-hi "
a
8
10
(
0
2
-1
UJ.B 6
0
8
10
:\ ;
) 5 10 15 2
•\
• v
f
•
) 5 10 15 2
:'\ :
.
5 5 10 15 2
:\- !
•
2
4
6
8
10
0 C
0.
2
4
6
8
10
0 C
0.
2
4
6
8
10
0 C
0
2
4
6
8
10
1
I
•
5 10 15 2
i
i }
1
* •
) 5 10 15 2
< i
it
i >
.
) 5 10 15 2
<
t
•
2
4
6
8
10
0 0.
0.
2
4
6
8
10
0 0.
0
2
4
E
8
10
0 0.
0
2
4
6
8
10
if
j _
01 0.1 1
n/»"
- 3
. T
h
i IH..J
01 0.1
:""f
•
01 0.1
;'?
•
-
•
1 1 HIM
1C
.
.
.
•
111 ItH
1 1C
•
-
1 K
•
•
.10
2
4
6
8
10
0.01 0.1 1 10
SEM/AVS
Figure 9-13. Vertical distributions of AVS, zinc SEM and SEM/AVS ratios at day 92 of the
zinc field experiment for the control and five treatments.
-------
SEM/AVS = 0.5 - 75 DAYS
Without Oligochaetes With Oligochaetes
0.
0)
Q
10
~ 20
40
Q.
OJ
a
50
10
30
40
50
0 50 100 150 200 250 300
AVS (umol/g)
i
20
40
60
BO
100
10
20
40
501 L
10
20
30
40
50
0 50 100 150 200 250 300
AVS (umol/g)
Cadmium (umol/g)
20 40 60 BO
Cadmium (umol/g)
100
Figure 9-15. Vertical distributions of AVS (top panels) and cadmium (bottom panels) for
with (left panels) and without (right panels) oligochaetes for cadmium oxidation experiment
and benthic mixing model. Data are plotted as means (symbols) and ranges (vertical and
horizontal bars) for plotting intervals. The model results are represented by the solid line.
-------
9-28
lower benthic biomass. The model is in reasonable agreement with the observations. The
result points out the need for an independent tracer for particle mixing so that the particle
diffussion coefficient can be determined separately.
Conclusions
The oxidation of the metal sulfides is a critical mechanism that can liberate the
metals from the anaerobic layer and cause the SEM/AVS ratio to exceed one and thereby
possibly cause toxicity. The loss of FeS and the liberation of free metal from its metal
sulfide both contribute to this phenomena. The kinetics of FeS oxidation are well
understood. Application of the laboratory measured rates in models of the vertical
distribution of sediments reproduce the observed decline in AVS.
The equivalent definitive laboratory data for the oxidation of the other metal sulfides
of interest are not yet available. Initial results for cadmium suggest that the rate is much
slower than for FeS but still large enough to liberate an amount of cadmium sufficient to
change the SEM/AVS from less than to greater than one in the top layers of the sediment
which would increase interstitial cadmium to amounts of toxicological significance and
result in impacts in benthic biota (see Chapter 7). However a similar experiment for zinc
failed to produce such a result, suggesting that the oxidation rate of zinc sulfide is much
slower than cadmium sulfide.
Organic Carbon Binding
The binding of metals to other phases in sediments also can be important in
determining toxicity when SEM exceeds AVS. This can be seen by considering the data
presented previously (Chapters 5 through 7). If the SEM/AVS ratio is greater than one
then, 56.8 percent of the sediments were toxic. If both SEM/AVS > 1 and the porewater
metals toxic unit concentrations >0.5 then 79.1 percent of the sediments were toxic.
Thus even if the SEM/AVS ratio is greater than one, implying that all the binding capacity
of the AVS has been exhausted, there may be another sediment component that is
-------
9-29
providing binding capacity for the metal and reducing its activity and, therefore, its toxicity
(see Figure 6-15). If the binding capacity is sufficient, the pore water toxic unit
concentration should be insignificant lexicologically. If, however, all the strong binding
components are exhausted, then the sediment pore water toxic unit concentrations will
exceed one and the sediment could be toxic. The data presented previously shows that
this is indeed the case where 79.1 of the sediments tested were toxic when SEM/AVS >
1.0 and IWTU 70.5 (see Table 6-3). For the remaining 20.9 percent it is likely that
dissolved ligands (probably dissolved organic carbon) and dissolved sulfide are providing
additional binding capacity.
An experimental method has been developed to investigate the sediment chemistry
associated with metal sorption under anaerobic conditions at equilibrium. It is named the
Anoxic Sequential Batch Titration (ASBT) method (see Appendix 9A for a description).
Sediment properties other than AVS were correlated to metals binding capacities in excess
of AVS binding capacity (i.e., non-AVS sorptive capacity), and equivalently to metals
porewater activity for copper, cadmium and lead at varying pHs. The results of
partitioning behavior of these three metals to several freshwater sediments under
anaerobic conditions are presented below.
Analysis Framework
The experimental protocol developed in this study was designed to investigate the
sorption of copper, cadmium, or lead under anaerobic conditions at equilibrium. The
experimental procedure allows (a) sediment samples to be contaminated with varying metal
concentrations under anaerobic conditions, (b) direct measurement of aqueous metals
activity, and (c) simultaneous sediment AVS determination on unspiked sediment samples
from the same homogenates.
The pH of the solution phase was controlled using various buffers. Metal
complexation by the buffer matrix was eliminated by using Goode buffers [ 18 through 20].
These buffers are designed to be non-reactive with metal species. The dissolved oxygen
-------
9-30
in the system was eliminated by stripping with nitrogen, and confirmed by direct
measurement. By controlling these two parameters, the water in the ASBT system
provides a chemical environment that is similar to that found in anaerobic sediments. The
exception to this is the dilution of porewater dissolved species.
The ASBT experimental procedure was evaluated with respect to several important
factors impacting sorption processes and anaerobic sediment chemistry. The four
parameters evaluated were (a) deoxygenation efficiency, (b) maintenance of a reducing
chemical environment, (c) sorption kinetics, and (d) particle induced desorption
mechanisms. The results of these evaluations are briefly summarized below and presented
in more detail elsewhere [21].
The primary requirement that sediment AVS not be oxidized was confirmed during
the sorption experiment. The kinetics of copper sorption onto various sediments was
investigated to determine a time to equilibrium. The results of several experiments all
confirmed a time to equilibrium of 1 to 2 hours. Lastly, the solids concentration was
evaluated for its effects on copper partitioning. Sediment samples were spiked and stirred
for one hour, representing a worse case with respect to particle interactions. Sediment
sorption isotherms were obtained at two or three different solids concentrations for three
different sediments. The sorption isotherms were not significantly different at solids
concentrations differing by an order of magnitude, suggesting that partitioning of copper
to anaerobic sediments is not dependent on solids concentration.
The metal in the ASBT system was in the form of (1) free metal, (2) hydrolysis
products of metal, (3) metal complexed with dissolved organic carbon (DOC) released from
the sediment sample and other ligands, and (4) sorbed metal partitioned onto sediment
particles. Loss of metal by sorption onto glass surfaces and by complexation to buffer
matrix was shown to be insignificant [17]. Thus metals sorbed onto the sediment particle
(Cs) can be calculated by difference between the added metal (Cj) and the total aqueous
metal (Cw):
-------
9-31
Cs - CT - Cw 0-2)
Total sediment-bound metal can be divided into that precipitated as metal sulfide by AVS
(Cs AVS), and that bound to other sediment components (CS(Non.AVS):
Cs - CS.AVS + CS.NOH-AVS ^-3)
Measurement of sediment AVS on an experiment-by-experiment basis provides the first
term in this equation. Thus the magnitude of the Non-AVS sorbed metal fraction can be
calculated by difference:
Cs,Non-AVS • C8 - CS(AVS {9"4)
This fraction of the sediment bound metal is analyzed below.
Sorption Isotherm Results
The data analysis procedure for the sorption data is illustrated using the copper data
at pH = 7 (Figure 9-16). The results from the various sediments are distinguished by the
various letters used as plotting symbols. The top figure presents the conventional
isotherm: the relationship of the total sorbed metal (Cs) to the metal activity in the solution
phase as measured using a specific ion electrode (Cf). The middle plot presents the non-
AVS sorbed metal concentration (Cs Non.AVS) versus metal activity. The bottom plot
presents the non-AVS sorbed metal concentration normalized by the organic carbon
fraction of the sediment:
Cs.Non-AVS,OC = Cs,Non -AVS/f oc *9"5'
versus the aqueous metal activity of the sediment where foc is the weight fraction of
organic carbon in the sediment. The extent to which the isotherms for the various carbon
normalized sediments plot over each other is the extent to which the most important
binding component of the sediment is organic carbon. The line fitted to the data is a
Langmuir isotherm model which is discussed below.
-------
UJ
10s
10'
UJ
o>
•* ,«3
„
10'
= i i iniii| i i iniiij i i HIUIJ i i iiiiiij i i HIE
k -
— ^^
10 * I i 11 mill i 11 mill i 11 iinil i iiiinil i iiiiml
,-4
«-3 4«-2
-1
10 J 10 c 10 ' 10'
10*
10
LU 10'
O
a»
l< >°3
cnS1
•* 10'
z
o
10'
mitu| i IIMIII| i iunnj i i HI
i 11 mill i 11 mill i 11 mill i niinil i iiiiml
10
LU 10"
Z
QU
LUO
03 .
LT CD in'
10
6
«-3 *«-2
10 J 10 c 10 * 10" 10
,-1
CO
cn e.
10'
= i i uiui| i i iiiuij n i iiuij t i iuui| i 11UB
i 11 mill i 11 mill i 11 inn! i iiiiml i iiiini
10 J 10* 10 * 10"
C ACTIVITY (mg/L).
101
Figure 9-16. Copper activity versus total sorbed metal (top panel), non-AVS sorbed metal
(center panel) and organic carbon normalized non-AVS sorbed metal (bottom panel) for pH
7.
-------
9-33
Logarithmic isotherm plots revealed non-linear relationships between sorbed metal
and metal activity. Logarithmic isotherm plots of all sediment systems at each pH are
.shown in the top panels of Figures 9-17 through 9-19. The sorbed metal concentrations
represent total bound metal (as calculated by Equation 9-2) on a dry sediment weight
basis. When the measured AVS component of the total sorbed concentration is subtracted
out (Equation 9-4), the resulting isotherms at each pH are shown in the center panels.
From these data it is clear that there is binding capacity for the three metals in these
sediments in excess of AVS. More importantly, this excess binding capacity is not related
to the dry sediment weight. The isotherms cover over a 10-fold range in partition
coefficients.
When the non-AVS binding capacities (Equation 9-5) are normalized to sediment
organic carbon fractions (foc), the isotherms represented in bottom panels of Figures 9-17
through 9-19 result. The isotherms collapse into a single isotherm with residual scatter (a
factor of 2 to 3) with one exception due to binding to other sediment components and
experimental error. The effect of sediment organic normalization is most dramatic at pH
7 although the normalization reduces the scatter at pH 6 also. At pH 8, however, the
precipitation of metal hydroxides is present in addition to the sorption process and this
complicates the analysis. The pH in sediments is typically 6.5 to 8.0. In support of
organic carbon normalization, Allen et al. [22,23] showed that upon oxidation of sediments
residual metal binding remained and that only a single phase was needed to explain this.
In a later paper [24] he showed that the binding of Cd to such sediments was identical in
nature to the binding by humic material extracted from the sediments.
The sorption isotherms can be fit using a number of models. The approximately
straight line behavior for some of the data (e.g., cadmium at pH = 6, Figure 9-18) suggests
a Freundlich isotherm might be appropriate. However, the curvature that exists at higher
concentration suggests that a limiting sorption capacity exists. The simplest model that
includes an upper limit is the Langmuir isotherm. The Langmuir isotherm equation is:
where:
-------
10'
PH 6
»
10'
PH 7
PH 8
'a i
at
a
z
o
10
10'
10'
3*
r e
uJ ••••-
cja i
tu
10
io
QU
UJO
03 A
DC 0> IO4
0>
cn e -
>— io3
o
C* c ee
a e
p
'i**
IO
10"2 IO"1 10° IO1 102 IO3 IO"3 IO"2 IO"1 10° IO1 IO2 IO3 IO"3 IO"2 IO"1 10° IO1 IO2 IO3
C ACTIVITY (mg/L)
C ACTIVITY (mg/L)
C ACTIVITY (mg/L)
Figure 9-17. Copper activity versus total sorbed metal (top panels), non-AVS sorbed metal
(center panels) and organic carbon normalized non-AVS sorbed metal (bottom panels) for
pH 6, 7, and 8.
-------
10'
Sf .
§t'°;
en .5
-J io:
10 J
PH 6
PH 7
PH B
10'
10
<
I-
UJ
2:
a
UJ
i« »•
O)
en
en'
en
10'
10'
lllf"' * Ml1"'
' *l"tl' * *
ca i
e9
-h
UJ
101
10'
ou
LUO
CD
OC C3J 10'
O -V
10'
tn
01
10'
- g
«
b
h d
e a
a
1 t ttttj 1 Illl
10"3 10"2 10"1 10° 101 102 103 10"3 10"2 10"1 10° 101 102 103 10"3 10"2 10"1 10° 101 102 103
C ACTIVITY (mg/L) C ACTIVITY (mg/L) C ACTIVITY (mg/L)
Figure 9-18. Cadmium activity versus total sorbed metal (top panels), non-AVS sorbed
metal (center panels) and organic carbon normalized non AVS sorbed metal (bottom
panels) for pH 6, 7, and 8.
-------
PH 6
PH 7
_J
uj io5
m\ io4
ggi
en£
J io3
H—
O
*~ 10*
, 106
_l
»-
UJ 5
z ios
O^
£* io4
8i>
£ 3
> 103
0 2
Z 10 e
J 10?
w io6
0" .
m° 10
P^ OJ
tn\ in4
a> 1U
en E
< io3
I io2
jh hhS
fc J*?1 B J«
hdtt" e rt«M
hjefi^ . iffltfto
"• tf . . Qd " C ••
3 o * ^* C C
e c)d J c
c
•
• •
a e 1
. 01 eh» :
i i *d ^ c •
e c)a c :
d
c "
1
]
r " -j
1
1
*
hU ^fc j i
c e j •
h
•!
- o^nJ
r i«tt *% j i
h
r i i
c ;
r
i :
r
10"2 10"1 10° 101 102 103
C ACTIVITY (mg/L)
io"3 io"2
1 10° lo1 io2 io3
C ACTIVITY (mg/L)
Figure 9-19. Lead activity versus total sorbed metal (top panels), non-AVS sorbed metal
(center panels) and organic carbon normalized non-AVS sorbed metal (bottom panels) for
pH 6 and 7.
-------
9-37
P _ Cs.OCKd,OC Cf (9-6)
s-oc ~ "r5 Tl? r~
^s,OC * "NJ.OC M
CS.QC = nor|-AVS sorbed metal per weight of sediment carbon (mg M/kg
organic carbon) = Cs, non-AVS (Equation 9-5)
Cf = aqueous metal activity {mg {M2 + }/L), and
Kd oc = partition coefficient (L/kg organic carbon)
CDQC = sorpt'00 capacity (mg M/kg organic carbon).
The properties of the Langmuir isotherm are shown in Figure 9-20A. The isotherm is linear
(slope = 1) on a log-log scale and approaches a constant as the concentration exceeds
C§,OC/Kd/oc.
If organic carbon is the only significant sorption phase, then one would expect that
the total sorption capacity of this phase is independent of pH within the narrow range
tested. On the other hand, the partition coefficient for metal sorption to natural and
synthetic particles is expected to vary with pH over a wider range. It arises because metal
sorption is a competition between the metal ion and the hydrogen ion. As the pH is
increases the concentration of hydrogen ion decreases, lowering the competition for
sorption sites, so that the quantity of metals sorbing increases. The isotherms that result
from this type of model are shown in Figure 9-20B. The increasing partition coefficient
as pH increases results in an increased paniculate concentration for a fixed metal activity.
The Langmuir parameters KdJOC for each pH, and the binding capacity (Cso Oc) can
be obtained using a non-linear regression of the log-transformed sorption data for each
metal. The resulting capacity and partition coefficients are listed in Table 9-4. Figures
9-21 through 9-23 present the Langmuir model and the data for copper, cadmium and lead
at the various pH's. For each pH, the data are summarized and plotted in the following
way. The sediment concentrations are sorted from low to high and divided into seven
groups with an approximately equal number of concentrations in each group. Then the log
mean of both the sediment concentration and aqueous activity are found. These mean
-------
The Lanamuir Isotherm:
C°,oc + Kd/ocCf
C°
Cf s'oc
K
d,oc
Kd,occf cf
Kd/oc
O)
10
iiiiiuu MI Him i 1 1 mm 1 1 1 mm i i HIE
= A
io2 I iiiiHui i ilium i ilium i ill mm tiiniin itiim
io"3 io"2 10" ' 10°
io2 io3
Model:
C ACTIVITY (mg/L)
Co oc Independent of pH
•s,oc
Kd oc Depends on pH
10'
— IO5
u
o
01
O)
10'
10'
= 11iiinu i ilium riunni rnnnir iiiiuiii IIIIIBJ
- B
pH 8
JiiifflB i iiniin 111 mm i ilium i iiiimi i limn
io"3 IO"2 10'1 10° IO1 IO2 IO3
C ACTIVITY (mg/L)
Figure 9-20. Properties of the Langmuir isotherm (top panel, A) and resulting isotherms
at varying pHs (bottom panel, B).
-------
LU
au
UJO
CD
DC cn
o.*
10'
icr
10'.L.
cn
cn £
102I_L
= I I Illlllj
^ PH5
> iiuii| i i IIIMI| i i mui| i i iuui| i i nig
fiinil i i mini i i mini iininil iniinil i mini
lO" 10'3 10'S 10'1 10° 101 102
C ACTIVITY (mg/L)
E i iiiniij i iiuiii| i iiiuiij i i UIMIJ i ininij i HUB
= PH7
2 i 11 mill i 11 inn i 11 mill i 11 mill i miml t i
10 J 10 * 10 ' 10" 10'
C ACTIVITY (mg/L)
UJ
ou
LU O
03
a: a>
cn
cn e
10'
10'
10'
10'
10'
= i i iuiii| i i niiiij irninij i i iiiiii
= PH8
i 11 mill i 11 mill i 11 mill i iiinnl i iiinnl i iniiiil
I
"3 "2 "1 °
10" 10" 10
10° 101 102
C ACTIVITY (mg/L)
Figure 9-21. Copper sorption isotherms for pH 6, 7, and 8. Data for plotting intervals is
represented by symbols (mean) and bars (standard deviation). The Langmuir isotherm
model is represented by the solid line.
-------
10'
OCJ
UJO
m .
g °) 10 *.
CD
W3
<~ IO3
z
O
10'
= i i uiiii| i i Mini] i 111 mi] i 111 mi] i 111 mi] i i HIES
- PH6
11 mill i i mini i iiiiml i iiiinil i iininl i i
ililli
10
10"2 10"1 10° 101 102 103
tu
ou
UJO
m
cc en
O.*
1/1
en
101
10'
10
10-
10'
= 1 lllllll| 1 IIIIMI| i illllll| i lllllll| I ITnillj I HUB
I PH7
2 i i mini iiniiiil iniiinl iiiiinil iniiiiil 11
linn
io"3 io"2
10
io2 io3
10'
UJ 10'
00
UJO
o: o> 10"
10-
en
O
o>
10'
M Illlllj I I llllll| I I llllll| I I llllll| I I llllll| I I lilt
I PH8
i mini i i mini iiiiinil iinniil iniiinl ininnl
io"3 io"2 io"1 10° ioj io2 io3
C ACTIVITY (mg/L)
Figure 9-22. Cadmium sorption isotherms for pH 6, 7, and 8. Data for plotting intervals
is represented by symbols (mean) and bars (standard deviation). The Langmuir isotherm
model is represented by the solid line.
-------
10'
UJ 10
ou
UJO
CD _
DC CT io5
10'
01
01 E
o
z
10'
= 111 nni| i i mill] i 11 mii| i 11 uiii| i 11 iiuij i 11 nm
PH6
i 11 mill i 11 mill i 11 inn! i imiiil t imiiil i i
mil
IO"3 IO"2 IO"1 10° IO1 IO2
C ACTIVITY (mg/L)
10
UJ 10'
ou
LUO
CD
OC O. 105
O X
10'
tn
<
z
o
10'
= i 1 1 iiuij i 1 1 mil] rriiiiiij \ 1 1 inn]
PH7
i ITTW
3 i 11 mill i 11 iiinl i 11 mill i imiiil i i mini i 11
Mil
io"3 io"2
10
io2 io3
C ACTIVITY (mg/L)
Figure 9-23. Lead sorption isotherms for pH 6 and 7. Data for plotting intervals is
represented by symbols (mean) and bars (standard deviation). The Langmuir isotherm
model is represented by the solid line.
-------
9-42
concentrations are plotted as the filled symbol. The standard deviation of the data in each
group is indicated by the lines in both the x and y axis directions. The model fits the data
for each metal surprisingly well, especially at the low metals activity where the isothems
will be used for setting Sediment Quality Criteria.
TABLE 9-4. ORGANIC CARBON BINDING CAPACITY AND
PARTITION COEFFICIENTS FOR COPPER, CADMIUM, AND LEAD.
Copper
Cadmium
Lead
Capacity
(L/kgOC
117,500
54,450
339,400
••'* .1:-: •--
pH6
390,400
20,740
248,700
Kd (L/kg OC)
.• ":
pH7 •
2,731,000
250,700
346,400
pH8 !
2,003,000
914,400
-
The sorption capacities for copper, cadmium and lead are shown in the top panel
of Figure 9-24. The capacities for copper and lead are similar whereas the capacity for
cadmium is lower. The partition coefficients as a function of pH are shown in the bottom
panel of Figure 9-24. They are almost equal for copper and lead but are an order of
magnitude lower for cadmium. A line with slope = 1 is included in the figure for
comparison. The relationship between log Kd Oc apd pH is essentially slope one between
pH 6 and 7. This corresponds to a replacement of a single hydrogen ion with each metal
ion sorbed, presumably as the metal hydroxide, MOH*.
These results are used in deriving sediment quality criteria in Chapter 11. However,
a simple case is examined in this chapter in order to assess the importance of sediment
organic carbon binding. The sediment quality criteria for a single metal is:
SQC = AVS + Kd
(9-7)
where Kd is the partition coefficient and CFCV is the final chronic value for that metal. It
should be pointed out that this equation is valid for a given metal only if the concentrations
of other metals is small relative to the AVS. The partition coefficient can be expressed in
-------
10'
CAPACITY
cj
o
en
o
E
D
CJ
O
in
CJ
10° _
10'
Cd
Cu
Pb
PARTITION COEFFICIENT
CJ
o
CO
cj
o
en
o
Cu
Pb
Cd
7
PH
8
Figure 9-24. Calculated capacity (top panel) and organic carbon normalized partition
coefficients (bottom panel) for cadmium, copper, and lead.
-------
9-44
terms of the organic carbon partition coefficient and the fraction organic carbon in the
sediment:
SQC = AVS + Kd/ocfoc CFCV {9'8>
The importance of organic carbon binding can be assessed by comparing the
magnitude of the term in the above Equation (9-8) that corresponds to organic carbon
binding:
Kd.oc foc CFCV (9'9)
with typical AVS concentrations found in sediments. This is demonstrated in Figure 9-25
using the fresh water final chronic values (at hardness = 100 mg/U and Kd oc for pH 7
for copper, cadmium and lead. The dashed horizontal lines show the typical range of AVS
and the dashed vertical lines give a typical range of foc. As foc and AVS increase so does
the allowable sediment metals concentration. Even with no AVS binding, sediment metals
levels of up to 100 umol/g may be acceptable depending on the organic carbon content.
Least Sorptive Phase
The extent of partitioning between sediments and interstitial water is a critical
component in establishing SQCs. Equation 9-8 points out the importance of the partition
coefficient, Kd. In the absence of AVS and significant organic carbon in a sediment, the
partitioning would be established between the mineralogical phases and the intersitital
water.
A series of experiments have been initiated to determine the partition coefficients
for metals using "clean" sediments which contain no AVS and no appreciable organic
carbon. The idea is to measure partition coefficients that can be used as minimum values
to determine the extent of partitioning. These partition coefficients would be used to
compute lower bounds of the SQC. For sediments with metal concentrations below these
values, the sediments would be judged to have satisfied the SQC.
-------
s
10'
PH 7
O)
I—I
O
CJ
CJ
O
a
O
= i 1 1 nun i 1 1 nun i 1 1 nun i i mini i i mini i i mini 1 1 l
eu
-1
in~2l i 11 nun i i mini i iniini i iiinin i iininl i iiiiini i iinin
10"3 10"2 10'1 10° 101 102 103 104
OC
Figure 9-25. Magnitude of organic carbon binding (y axis) as foc (x axis) varies for
cadmium, copper and lead, pH 7. Typical ranges of foc and AVS are represented by
dashed lines. The freshwater FCVs at hardness of 100 mg/l are 1.1 ug/l for cadmium,
12.0 ug/l for copper and 3.2 ug/I for lead.
-------
9-46
Determinations of the partiton coefficeint, Kd> for cadmium, copper, nickel, lead,
and zinc were done using two adsorbents with very low organic carbon content. The
analyses were done at the U.S. EPA Environmental Research Laboratory, Athens, Georgia
[25]. The adsorbents were a commercially obtained washed sea sand and a sample of
natural sand from Ona Beach, Oregon. The washed sea sand was used as provided. Most
of the aqueous phase was removed from the Ona Beach sand samples. Remaining water
content was 16.7 percent. Total organic carbon content was 0.006 percent, for the
washed sea sand, and 0.019 percent for the Ona Beach sand.
About 5 g of sediment and 30 ml of a serial dilution of the five mixed metals were
mixed and allowed to equlibrate. The samples were then centrifuged and the supernatants
were analyzed for pH and remaining metals using the graphite furnace atomic adsorption
spectrophotometer (GFAAS) 125]. The adsorbed amounts of each of the metals, on a
gram weight basis were computed from the difference between the initial metals
concentration and the remaining metals concentration in the supernatant divided by the
initial volume of 30 ml. The amounts absorbed (gram) were then divided by the amount
of sediment added (gram) to compute the adsorbed solid phase concentration, Cs. The
partition coefficient, Kd was then computed as follows:
Kd = -£1 (9-10)
Cw
where Cw is the remaining metals concentration in the supernatant. Table 9-5 presents
the mean Kd and statndard error for each of the metals.
These initial results suggest that the partition coefficients are varying somewhat,
but not by orders of magnitudes. Therefore, it is probable that minimum partition
coefficients can be established which would provide a lower bound for the Sediment
Quality Criteria on an SEM basis. The details of this formulation are discussed in Chapter
11.
-------
9-47
TABLE 9-5.
Sea Sand
Ona Beach Sand
aStandard error,
PARTITION COEFFICIENT, Kd, FOR LOW ORGANIC CARBON
CONTENT SEDIMENTS
Copper
163.05
(6.97,5)a
265.08
(16.1,6)
Nickel
21.17
(4.15,5)
34.58
(1.87,6)
Kd(L/Kg)
Cadmfum
48.36
(7.34,5)
71.09
(5,44,7)
Zinc
1,847.67
(137.95,6)
2,183.70
(260.53,5)
Lead
273.65
(24.1,5)
579.10
(33.5,4)
number of values
Pore Water and SEM/AVS Sampling
The sampling methods used for AVS and interstitial water have been studied and
recommendations are available that appear to be the optimal choices at present.
Pore Water Sampling
Bufflap and Allen [26] reviewed the four commonly used methods for collection of
pore water and potential artifacts from their use, particularly in the preparation of samples
for trace metal analysis. Two of the methods, centrifugation and squeezing, are ex-situ,
requiring the removal of sediment from the natural environment. The other two, dialysis
and suction filtration, are used in-situ. Their work has been included in Volume II of this
submission. In addition to each method having its own advantages and disadvantages,
there are several general sources of error that can alter pore water chemical
concentrations. A summary of their findings is presented below.
Several sources of error in sampling can lead to erroneuos porewater
measurements. A primary source of error is the oxidation of anoxic pore waters. If a
sample is allowed to oxidize the speciation of iron and other trace metals will be altered.
Anoxic sediments should be handled in a glove box or glove bag when extracting pore
water. Another source of error that can occur during sampling is pore water oxidation as
-------
9-48
a result of the mixing of oxic and anoxic sediments during sampling. Also metal
contamination should be avoided when sampling by avoiding contact with metal parts of
the sampling device that may contaminate the sample. Some studies indicate that sample
extractions at temperatures higher than in-situ may change the composition of the pore
water however there is no direct evidence that extraction temperature plays a significant
role in trace metal concentrations in pore water [26]. Pore water samples should be
filtered during or after extraction to remove residual particles that can interfere with both
analytical procedures or alter trace metal concentrations due to adsorbtion/desorbtion of
the metal to the residual particles. Buff lap and Allen [26] discuss the three most common
sampling techniques; dredging, grab sampling and coring, and give recommendations to
minimize sampling errors.
The primary concern with analyzing sediment pore water is finding an extraction
technique that will produce samples that best represent the natural environment. To
accomplish this goal, the technique that is used should have the lowest potential for
producing sampling artifacts. The ex-situ techniques, centrifugation and squeezing, require
the removal of sediment samples from the natural environment. The squeezer method
employs various apparatus to pressurize a sediment sample which forces the pore water
through an exit port. The squeezer apparatus are known as core section or whole core
squeezers. Core section squeezers employ either gas pressure or a mechanical means of
pressurizing the sediment sample and forcing the pore water through an exit port. Core
section squeezers are an inexpensive and simple means of extracting sediment pore water.
They also offer immediate filtration of the water samples, thus eliminating a handling step
which may introduce contamination to the samples. The disadvantage of core section
squeezers is that their use requires handling the sediment which may introduce artifacts
resulting from oxidation or temperature differences.
Whole core squeezers may help to remove the possibility of artifacts that may result
when using the core section squeezers because the sediment remains in the core liner with
which it was removed from the natural environment. These squeezers apply pressure to
-------
9-49
the sediment by the use of plungers. A problem with the whole core squeezers is that
solid phase pore water interactions may alter pore water concentrations during squeezing.
Centrifuging is another widely used ex-situ simple technique to obtain pore water.
Centrifuging can be conducted at in-situ temperatures and handling the sediment samples
can be done in an inert atmosphere to avoid artifacts that may change pore water
concentrations. One problem with Centrifuging is that some fine particulates may still
remain in pore water. Fine particulates can be removed either by using a built-in filter at
the top of the centrifuge cup or by displacement of the pore water in the sediment by an
inert solvent placed in the centrifuge tube. The dense solvent replaces the pore water in
the sediment forcing the less dense pore water to the top. Of the two existing techniques,
squeezing has the lower potential for artifacts because all handling steps can be conducted
in an inert atmosphere contained in a glove bag in order to avoid oxidation artifacts. In
addition, pore water filtration can be conducted in-line, thus eliminating a handling step
that is required in centrifugation, and lowering the potential for artifacts.
Squeezing and Centrifuging are discussed by Bufflap and Allen [26] in more detail.
In-situ techniques, such as dialysis and suction filtration, have less potential for producing
sampling artifacts than ex-situ techniques because pore water samples are extracted
directly from the natural environment. The general principle of dialysis sampling involves
allowing a volume of deionized, distilled water to come to equilibrium with the sediment
pore water in order to determine chemical concentrations. One problem with dialysis
samplers is that the chamber water must be deaerated before insertion into anoxic
sediemnts to avoid oxidation of the sample. Dialysis has limitations because equilibration
times can last several weeks. In addition, the volume of sample is limited by the size of
the sample chambers, not by the physical features of the sediment. Lastly dialysis
samplers generally require placement and retrieval by SCUBA divers, thus increasing study
costs.
Because there are different techniques available for extracting sediment pore water,
it is often difficult to compare data from different laboratories. What is needed to limit
-------
9-50
these discrepancies is to compare the existing techniques and to develop a sampling
methodology that will produce pore water samples that best resemble the in-situ conditons
and can be easily utilized by all researches. Studies done to compare the existing
techniques are discussed by Buff lap and Allen [26]. However, more research is needed
in this area.
AVS/SEM Sampling
A draft analytical method for the determination of AVS and SEM in sediment has
been proposed [27,281. This method describes procedures for the determination of acid
volatile sulfides (AVS) and for metals that are solubilized during the acidification step
(SEM). The conditions used have been reported to measure amorphous or moderately
crystalline monosulfides. Because the relative amounts of AVS and SEM are important in
the prediction of potential metal bioavailability, it is important to use the SEM procedure
for sample preparation for metal analysis. This uses the same conditions for release of
both sulfide and metal from the sediment and thus provides the most predictive means of
assessing the amount of metal associated with sulfide. The method is included in Volume
II and a summary is provided below.
The AVS in the sample is first converted to hydrogen sulfide (H2S) by acidification
with hydrochloric acid at room temperature. The H2S is then purged from the sample and
trapped. The amount of sulfide that has been trapped is then determined. The SEM are
metals liberated from the sediment during the acidification. These are determined after
filtration of the supernatant from the acidification step.
Two types of apparatus for sample purging and trapping of H2S are described. One
uses a series of Erlenmeyer flasks while the other uses flasks and traps with ground glass
stoppers. The former is less costly. The latter is less prone to leakage that causes low
recovery of AVS. The latter is recommended when higher degrees of precision are desired
and for samples containing low levels of AVS (~ 0.1 ^mol/g).
-------
9-51
Three means of quantifying the H2S released by acidifying the sample are provided.
In the gravimetric procedure, the H2S is trapped in silver nitrate. The silver sulfide that is
formed is determined by weighing. This procedure is recommended for samples with
moderate or high AVS concentrations. In an alternative prpcedure the H2S is trapped in
an antioxidant buffer before using an ion-selective electrode. After release of the H2S, the
acidified sediment sample is membrane filtered before determination of the SEM by atomic
absorption or inductive coupled plasma spectrometric methods. In the colorimetric
method, the H2S is trapped in sodium hydroxide. The sulfide is converted to methylene
blue that is measured. This procedure is recommended for samples that have low to
moderate AVS concentrations.
Using the apparatus described by Allen et al. [28], the colorimetric method of
analysis is capable of detecting AVS at concentrations normally encountered with a
recovery of sulfide of at least 90 percent. High precision is possible if the Allen et al. [24]
apparatus is used with a limit of detection of approximately 0.01 //mol/g dry sediment.
-------
APPENDIX 9A
EXPERIMENTAL PROCEDURES
Sediment samples with overlying water were collected in one gallon plastic
containers from various locations, transported under ice, and stored at 4°C throughout the
study period. Subsampling procedures attempted to minimize exposure times to air and
elevated temperatures.
Sorption isotherm data were obtained by titrating sediment with either copper,
cadmium or lead in batch mode (Figure 9A-1). The Anoxic Sequential Batch Titration
(ASBT) method consists of a series of 250ml, Florence flasks (typically 10) stoppered gas-
tight with two-hole rubber stoppers fitted with glass tubing. The entire train of flasks is
connected to a source of purified nitrogen gas (Matheson, Prepurified, 99.998 percent
minimum) which is bubbled through a vanadate/HCI/amalgamated- zinc oxygen-stripping
solution as a polishing de-oxygenation step. Into each flask, a wet sediment sample is
introduced, a known volume of stock metal solution is added, and a known volume of
deaerated buffer solution is added. Each flask is similarly prepared and sequentially
attached to the flask train. An Acid Volatile Sulfide (AVS) determining apparatus is
attached to the sorption flask train to measure the sediment AVS simultaneously. The
sediment is then titrated by varying the copper, cadmium, or lead stock solution volume
added to each flask and thereby obtaining a range of sediment-bound and aqueous copper,
cadmium, or lead concentrations.
Total aqueous metal concentrations were determined by Atomic Absorption
Spectroscopy (AAS; Perkin Elmer model PE-3030) in samples of the overlying water of the
ASBT system which were filtered through glass fiber filter discs and acidified to 1 percent
HN03 by volume. Metal concentrations below 1 mg/L required the use of AAS in Graphite
furnace mode (model HGA-400). Concentrations greater than 1 mg/L were analyzed by
AAS in Flame mode (air-acetylene flame; Perkin-Elmer lamps).
Metal activity in the overlying water of the ASBT system was measured directly by
an ion-selective electrode (Orion, #90-29) in conjunction with a reference electrode (Orion,
-------
c
o
03
CO
15
0
CT
0
CO
o
"x
o
•••
o
CO
CO
a
CO
H
i_
a
CO
£
CD
TO
a
CO
H
CD
2
CD z: z:
£- 3 3
3 CO CO
OCTOW
O) D>
X.
CO
CO
"5
"CD
a> r= -= -
£ "•
o
e
Q.
E
co .2 co
jo ^ CO
o I ^
CD
E
V
.£ E
CO O H= "O
co in 3 o
O CM CD co
«. •• ** • •
•O T3 OJ
C C JZ
CO CD
'C .£
g « E
E
T"
CD
W T "O
3 t-g
.2 o ja
o>
O
k_
•*-• r-
eg
3
•£
CD
•o S-
« =5 «
^ w -e
3 w £
£g I
O 0}
dlg-3-
gS g
oJ o c
CO C 0)
i-ili
2 I- J2 o
.t; c ^
*-««*-
£ •* Q) C
o « jr «
!S ^S ^ E
« ^ _ .=
•Q c .E -o
To .2 c w
CD £ CO W
(O -C CD C
•5 E g =
i |-S «
< c go
. SE§
*- >• to -^
S _3 CD E
LL 0)
-------
9A-3
#90-02) and a voltage meter (Orion SA-720). Daily metal activity standard curves
generated at the appropriate pH were used to calculate a regression line on the linear
portion of the curves. It was assumed that metal activity was linearly related to electrode
response at all activities. This allowed a given day's standard curve regression equation
to be used to calculate the metal activity of that day's ASBT overlying water samples,
even though the millivolt reading of some samples fell below the lowest standard. Copper,
cadmium or lead activities calculated in this way over-estimated the metal activity of
samples below 0.01 mg {Me2 + }/L. The ionic strength of both samples and standards was
adjusted to 0.005M NaN03 prior to determining metal activity.
The pertinent sediment characteristics relating to metal partitioning were determined
as follows. Sediment dry solids were determined by weighing a sediment sample before
and after drying at 103°C overnight. Sediment carbon (total, organic, and inorganic) was
measured using a LECO model CHN-800 Carbon-Hydrogen-Nitrogen analyzer. Sediment
dried at 103°C was analyzed for paniculate total carbon (PTC). A subsample of this
material was acidified with 0.05M HCI for one hour and dried at 103°C. This treated
sediment was analyzed for paniculate organic carbon (POO, and the difference between
PTC and POC was attributed to paniculate inorganic carbon (PIC; i.e., carbonates).
The pH of the overlying water was measured using a Beckman (ALTEX) 060 pH
meter with a Fisher Standard Polymer-Body Gel-filled combination pH electrode (#13-640-
108). The pH of the ASBT systems were buffered using Goode buffers (0.005 to 0.01 M
solutions adjusted to desired pH with NaOH). At pH 6, MES (2-[n-
morpholinolethanesulfonic acid, sodium salt; pKa = 6.1; Sigma M-3885) was used, at pH
7, MOPS (3-[Morpholino]propanesulfonic acid, sodium salt; pKa = 7.2; SIGMA M-9381)
was used, and at pH 8, HEPES (N-[2-Hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid.
Sodium salt; pKa = 7.5; SIGMA H-2393) was used. These buffers are designed to be non-
reactive with metal species [17,18,19].
-------
REFERENCES
1. Leonard, E.N., Mattson, V.R., Benoit, D.A., Hoke, R.A., and Ankley, G.T. 1993.
Seasonal variation of acid volatile suitide in sediment cores from three northeastern
Minnesota lakes. Hydrobiologia. 271:87-95.
2. U.S. Environmental Protection Agency. Environmental Research Laboratory,
Narragansett, Rhode Island. EMAP-Estuaries Virginian Province data. Unpublished.
3. Morita, R.Y. 1975. Psychrophlic bacteria. Bacteriological Rev. 39:144-167.
4. Boothman, W.S. and Helmstetter, A. Vertical and seasonal variability of acid
volatile sulfides in marine sediments. Manuscript.
5. Howard, D.E. and Evans, R.D. 1993. Acid-volatile sulfide (AVS) in a seasonally
anoxic mesotrophic fake: Seasonal and spatial changes in sediment AVS. Environ.
Toxicol. Chem. 12:1051-1057.
6. Ankley, G.T., Phipps, G.L., Leonard, E.N., Benoit, D.A., Mattson, V.R., Kosian,
P.A., Cotter, A.M., Dierkes, J.R., Hansen, D. J. and J.D. Mahony. 1991. Acid-
volatile sulfide as a factor mediating cadmium and nickel bioavailability in
contaminated sediments. Environ. Toxicol. Chem. 10:1299-1307.
7. Di Toro, D.M. and Fitzpatrick, J.F. 1993. Chesapeake bay sediment flux model.
HydroQual, Inc. Mahwah, NJ. Prepared for US Army Corps of Engineers,
Waterways Experiment Station. Report EL-93-2
8. Berner, R.A. 1980. A rate model for organic matter decomposition during bacterial
sulfate reduction in marine sediments. Biooeochemistrv pi organic matter aj the
sediment-water interface. France, Comm. Natl. Recherche Scientific. 35-44.
-------
9. Forssberg, K. S. 1985. Flotation of Sulphide Minerals. New York. NY: Elsevier
Science Publishing Co. Inc.
10. Nelson, M. B. 1978. Kinetics and mechanisms of the oxidation of ferrous sulfide
Ph.D Thesis. Stanford University, Palo Alto, CA
11. Di Toro, D.M., Mahony, J.D., and Gonzalez, A.M. Particle oxidation model of
synthetic FeS and sediment AVS. Environ. Tox. Chem., 1994. Submitted.
12. Zhuang, Y., Allen, H.E., Fu, G. 1994. Effect of aeration of sediment on cadmium
binding. Enivon. Toxicol. Chem. 13:717-724.
13. Nyffeler, U. P., Santschi, P. H., & Li, Y.-H. 1986. The relevance of scavenging
kinetics to modeling of sediment-water interactions in natural waters. Limnol.
Oceanogr., 31:277-292.
14. Boudreau, B. P. 1986. Mathematics of tracer mixing in sediments: I.
Spatially-dependent diffusive mixing. Am. J. Sci., 286:161-198.
15. Hansen, D.J., Mahony, J.D., Berry, W.J., Benyi, S., Corbin, J., Praft, S.D., Able,
M.B. 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. Manuscript.
16. Liber, K., Ankley, G., Call, D., Markee, T., and Schmude, K. 1994. Seasonal
relationships between acid volatile sulfide concentrations and toxicity of zinc to
benthic macroinvertebrates. Manuscript In Preparation.
17. Mahony, J.D. 1993. Personal communication on oligochaet experiments.
18. Goode, N.E., et.al., 1966, Biochemistry, 5, 467.
-------
19. Goode, N.E. and S. Izawa, 1972, Methods Enzymol., 24(Part B), 53.
20. Ferguson, WJ. and .Goode, N.E. 1980. Analytical Biochem., 104, 300.
21. Gonzalez, A.M., 1992, An Experimental Study of Sediment Organic Carbon as a
metal-binding phase under anoxic conditions, Master's Thesis: Manhattan College,
Riverdale, N.Y.
22. Fu, G., Allen, H.E., and Cowan, C.A. 1991. Adsorption of cadmium and copper
by manganese oxide. Soil Science. 152:72-81.
23. Fu, G., and Allen, H.E. 1992. Cadmium adsorption by oxic sediments. Water
Research. 26:225-233.
24. Fu, G., Allen, H.E., and Cao, Y. 1992. The importance of humic acids to proton
and cadmium binding in sediments. Environ. Toxicol. Chem. 11:1363-1372.
25. Garrison, W. 1994. Personel communication on metal sorption on low sorbitivity
phases.
26. Buff lap, S.E. and Allen. H.E. 1994. Sediment pore water collection methods: A
review. Water Research. In press.
27. Allen, H.E., Fuj, G., Deng, Baolin. 1993. Analysis of acid-volatile sulfide (AVS) and
simultaneously extracted metals (SEM) for the estimation of potential toxicity in
aquatic sediments. Environ. Toxicol. Chem. 12:1441-1453.
28. U.S. Environmental Protection Agency. 1991. Draft analytical method for
determination of acid volatile sulfide in sediment. Office of Science and
Technology, Health and Ecological Criteria Div., Washington, D.C. 20460.
-------
CHAPTER 10
CONSIDERATIONS FOR ASSESSING METAL BIOAVAILABILITY IN SEDIMENTS
Based on the studies described in the preceding chapters, it is apparent that
evaluation of pore water metal concentrations and/or SEM:AVS ratios can provide
significant insights concerning metal bioavailability in sediments. We feel that the two
techniques are complementary and should be used in conjunction with one another
as an approach to providing assessments of the potential ecological impacts of metals
in sediments. However, when using these measures of metal bioavailability, it is
important to recognize the limits of applicability of the techniques. These are
discussed below.
Interstitial Water
Comparison of metal concentrations in pore water to water-only toxicity data
can be used to predict not only the presence, but also the extent of, metal toxicity in
sediments. The ability to actually quantify bioavailable metals in sediments is
attractive for a number of reasons. For example, quantification of bioavailable metal
facilitates the evaluation of differences in relative species sensitivity and thus, enables
the identification of species at risk. This is not yet possible with SEM and A VS.
When SEM and AVS, due to other possible binding phases, it is not yet possible to
predict actual pore water concentrations of metals. Another advantage to monitoring
pore water metal concentrations is that they should be useful for predicting the
toxicity of metals, such as chromium, which do not form insoluble sulfides. Finally,
because AVS is readily oxidized, it is not an important binding phase for metals in
completely aerobic sediments, however, the bioavailable fraction of metals should still
be approximated by using the pore water concentrations.
There also are disadvantages to solely using pore water concentrations to
quantify metal bioavailablity. First, because pore water is operationally defined (i.e.,
-------
10-2
there is no standard method for isolation although we recommend a procedure in
Chapter 9), there is a valid concern that laboratory variations in preparation may result
in significant differences in metal concentrations found in pore water. A second
disadvantage to using pore water metal concentrations to predict toxicity is that, if
one accepts the paradigm that pore water is indeed a major route of contaminant
exposure for epibenthic and benthic invertebrates, it may be difficult to account for
the effects of the pore water matrix (e.g., dissolved organic carbon, hardness, salinity)
on metal complexation and bioavailability. This is, of course, also an issue of on-going
concern in the area of WQC issued by the US Environmental Protection Agency. A
final potential complication is that for species-specific assessments, it is necessary to
have a water-only effects data base for comparative purposes for the metal and the
species of concern. However, for the purpose of SQC derivation for metals in
sediments, target values for pore water metals could be obtained from appropriate
WQC documents in a fashion similar to that for nonionic organic chemicals.
Acid Volatile Sulfide and SEM
The studies described in previous chapters have clearly demonstrated that AVS
can be a key factor influencing interstitial water concentrations and bioavaiiability of
metals in sediments. In virtually no instance have we seen metal toxicity when SEM
is less than AVS and SEM greater than AVS often has been predictive of the presence
(but not extent) of metal toxicity. The use of SEM and AVS concentrations alleviates
the need for water-only effects data in an assessment since no bioavailability is
expected at SEM less than AVS. A further advantage to measuring sediment SEM
and AVS in sediments is that it gives an indication as to the relative size of the pool
of both components. This is not possible through monitoring pore water metal
concentrations. Pore water metal concentrations should be low in sediments with
SEM very much less than AVS to, theoretically, Sem equals AVS. Yet, sediments
with SEM close to AVS would be of more potential concern than those with SEM < <
AVS. In the absence of other metal binding phases, slight increases in SEM or
-------
10-3
decreases in AVS could cause the SEM to exceed the AVS and thereby result in
toxicity.
There are a number of limitations inherent in using either pore water or SEM
and AVS to predict bioavailability. First, because AVS varies seasonally in a system-
specific manner, it is desirable that SEM and AVS and pore water metals be measured
over time, or at least when AVS is expected to be minimal (e.g., late winter in our
studies). A single sampling is only a snap-shot of what occurs through the course of
the year.
At present, investigations are ongoing to assess the role of AVS in deeper
sediments relative to metal partitioning at the sediment surface, where most biological
acitvity and exposure occurs. This is significant because the AVS pool in deeper
sediments appears to remain relatively constant, as opposed to AVS in surficial
sediments. It may be, for example, that as surficial sediments are depleted of AVS,
metals will subsequently bind to AVS in deeper sediments. Alternatively, as AVS
concentrations are depleted in surface sediments, other binding phases for metals may
become important in determining bioavailability. In any instance, it is important that
pore water metal and SEM and AVS measurements be made at all relevant points over
the vertical gradient of the sediment cores; for example, if concern is only for
exposure of current benthic communities, the measurements can be made in the
surficial sediment horizon. On the other hand, if the assessment is focused on
possible impacts of deeper sediments (e.g., for dredging), appropriate measurements
should be made throughout the core.
Neither pore water metal concentrations nor SEM and AVS can be used to
assess potential metal bioavailability in situations where sediments are expected to
be altered and become aerobic through physical disturbance (e.g., storms, boat
traffic). In fact, in these cases, it may be possible in a worst case evaluation to
"exhaust" sediment AVS (e.g., by aerating the sample) before attempting to determine
-------
10-4
the presence of bioavailable metals, possibly through evaluation of pore water metal
concentrations.
Field validation of sediment quality criteria is an important component of the
establishing their validity. To date, the most exhaustive studies have been conducted
using spiked sediments in the laboratory and field which focus on changes in benthic
community structures and the bioaccumulation of metals by benthos (Chapters 7and
8). These studies have been consistent with predictions based on SEM and AVS
ratios, and /or a pore water exposure model [1,2]. However, further work in this area
is needed, in particular with in situ sediments contaminated by point or nonpoint
source anthropogenic inputs of metals.
Based on the technical considerations described above, we present the
following recommendations/caveats for assessing the potential bioavailability of
metals in sediments.
1. Both SEM and AVS and pore water metal concentrations should be measured
in sediment assessments focused on defining bioavailability. A standard
method for the extraction and measurement of SEM and AVS has been
described [3]. For the studies described above, pore water was isolated using
either of two different techniques: dialysis chambers (peepers) or
centrifugation. Other pore water isolation techniques also may be useful;
however, we have had little experience with them.
2. If AVS is used as a normalization phase, it should be used only for cadmium,
nickel, lead, zinc, and copper, and only for these metals when simultaneously
extracted with the AVS. Molar concentrations of the metals then can be
summed to generate SEM and compared to AVS ratios. Theoretically,
however, it is possible to utilize pore water measurements of metals, other than
the five listed above, to evaluate their potential bioavailability.
-------
10-5
3. It is strongly recommended that cadmium, nickel, copper, lead and zinc all be
measured when evaluating SEM and AVS and pore water metal concentrations,
at least in initial test samples. This is because although all five of these metals
have a higher affinity than iron for sulfide in monosulfide complexes,
individually they also have varying affinities (solubility products) for the sulfide.
Thus, for example, cadmium will displace nickel from sulfide, and if excess
sulfide is not available, nickel will be released to the pore water. If only pore
water metals were measured, or only nickel was measured in the solid phase,
the analyst would erroneously conclude that nickel was the only problem in the
sediments, when in fact, elevated concentrations of cadmium also were
present. In order to have a complete understanding as to why a specific metal
is present at elevated concentrations in pore water, it is necessary to know the
molar concentrations of all the SEM. This is particularly true when considering
the fact that metal concentrations often covary in contaminated aquatic
sediments, that is, rarely is only one metal of concern.
4. In fully aerobic sediments (e.g., sand), AVS concentrations should not be used
to attempt to predict the bioavailability of metals in sediments. Theoretically,
however, it should be possible to infer bioavailability based on pore water metal
concentrations. Moreover, as described in Chapter 9, significant progress is
being made in identifying alternative normalization phases, such as carbon, for
metals in aerobic sediments.
5. Only a limited amount of research has been conducted to assess the utility of
SEM and AVS or pore water concentrations for predicting metal bioavailability
in long-term exposures. Given the uncertainties in kinetics of metal and AVS
interactions in temporal cycles, and the lack of information on the importance
of other metal binding phases relative to these cycles, extrapolations of the
exposure model to long-term situations should be made with care. Further
information also is needed concerning the nature of the microhabitat of
-------
10-6
invertebrates relative to long-term changes in metal bioavailability in sediments.
In any case, it is important that pore water metal concentrations and SEM and
AVS be measured in sediment horizons appropriate for defining exposure to
species of concern. Based on our research, it appears that exposure of benthtc
organisms to surficial sediments can be defined reasonably well by
measurement in the 0-2 cm horizon.
6. As with any chemical-specific monitoring method, the analyst should be aware
that: (a) not all chemicals of possible toxicological concern can be measured in
environmental samples; and (b) in most instances, it is difficult to account for
possible toxicological interactions among measured toxicants. For these
reasons, we strongly recommend toxicity tests as an integral part of any
assessment concerned with the effects of sediment contaminants.
-------
References
1. Hare, L, CarignarvR., and Huerta-Diaz, M.A. 1994. A field experimental test
of the hypothesis that acid-volatile sulfide (AVS) concentrations improve the
prediction of metal toxicity and accumulation by benthic invertebrates. Limnol.
Oceanog. In Press.
2. Liber, K., Call, M., Markee, T., Schmude, K., and Ankley, G. 1994. Seasonal
relationships between acid volatile-sulfide concentrations and toxicity of zinc
to benthic invertebrates. Draft Manuscript.
3. Allen, H.E., Fu, G., and Deng, B. 1993. Analysis of acid-volatile sulfide (AVS)
and simultaneously extracted metals (SEM) for the estimation of potential
toxicity in aquatic sediments. Environ. Toxicol. Chem. 12:1441-1453.
-------
CHAPTER 11
PROPOSED SEDIMENT QUALITY CRITERIA
Introduction
Sediment quality criteria are intended to be the U.S. Environmental Protection
Agency's best recommendation of the concentration of a metal in a sediment that will
be protective of benthic organisms. The sediment quality criteria for the five metals:
cadmium, copper, nickel, lead, and zinc, will be based on the Equilibrium Partitioning
model of bioavaiiability. EqP asserts that the bioavailability of a chemical is related
to the chemical activity of the sediment - interstitial water system. For these metals
it has been shown that biological effects correlate with free metal activity in either
water only exposures or in sediment - interstitial water exposures. Thus the SQC for
these metals are based on insuring that the free metal activity is below levels that can
cause undesirable biological effects.
SQC's for the five metals being considered can be derived using four
procedures:
(1) By comparing the molar concentrations of cadmium, copper, lead, nickel, and
zinc to the molar concentration of AVS in sediments;
(2) By comparing the measured interstitial water concentrations of metals to the
water quality criteria final chronic values (FCVs) for the metals;
(3) By using organic carbon based partition coefficients in addition to the AVS to
compute the interstitial water concentrations and compare them to the water
quality criteria FCVs;
-------
11-2
(4) By using minimum partition coefficients and AVS to compute lower bound
sediment concentrations that are unlikely to cause toxicity.
These procedures are described in more detail below. No citations to the
literature are included in the chapter since these have been provided in the previous
chapters. We believe that the technical basis for implementing procedures (1) and (2)
are presently supportable. Initial data for implementing procedures (3) and (4) have
also been presented. However, additional research is required to complete the
required data sets, as discussed below.
In the following sections we discuss the application of these methods to
deriving an SQC for a single metal. Then we continue with the more common
situation where appreciable concentrations of all the metals are present. The
application of these principles to the derivation of a sediment quality criteria for a
single metal is included for illustrative purposes only. It is instructive to present the
logic for this case as a prelude to the derivation of the multiple metal criteria.
However, as will become clear subsequently, single metal criteria are not usually
applicable to field situations since there is always a significant quantity of more than
one metal to be considered. In fact it is misleading to think of the criteria one metal
at a time. As we shall see, single metal criteria are inherently underprotective because
of the additive nature of AVS binding. Nevertheless the following sections are
included because of their instructional value.
One final point should be made with respect to nomenclature. When we use
the terms non-toxic or having no effect, we mean only with respect to the five metals
considered in this document. The toxicity of field collected sediments can be caused
by other chemicals. Therefore not violating the SQC for metals does not guarantee
that the sediments are non-toxic, only that the five metals being considered will not
have an undesirable biological effect.
-------
11-3
Single Metal Sediment Quality Criteria
Sediment quality criteria for metals will be expressed in molar units. These are
the natural units because of the one to one molar stoichiometry of metal binding to
AVS. Thus solid phase constituents: AVS and simultaneously extracted metal, SEM,
are in //mol/g units. The interstitial water concentrations are in //mol/L units as are the
metal activities. To be consistent with the usual chemical notation, metal activities
are denoted by curly brackets {} and metal concentrations are denoted by square
brackets []. The partition coefficients have units L/g consistent with the above solid
and aqueous phase concentration units. Table 11-1 summarizes these conventions:
; Species :
AVS
SEM
Metal activity
Dissolved Metal cone.
Total Metal cone.
TABLE 11-1.
Phase
Solid
Solid
Aqueous
Aqueous
Solid + Aqueous
Notation
[AVS]
[SEM]
{M2+}
[Md]
[SEMT]
Units
//mol/g
//mol/g
//mol/L
//mol/L
//mol/L(bulk)
The subscripted notation, Md, is used to distinguish aqueous phase molar
concentrations from solid phase molar concentrations with no subscript. For the total
concentration, [SEMT], the units are the moles of metal per volume of solid + liquid
phase. Since metal activity is used below only relative to the aqueous phase no
subscript is needed.
AVS Criteria
It has been demonstrated that if the SEM of a sediment is less than the AVS:
-------
11-4
[SEM]*[AVS] (11-1)
;
then no toxicity effects are seen. This is consistent with the results of a chemical
equilibrium model for the sediment - interstitial water system. The resulting metal
activity {M2+} can be related to the total SEM of the sediment and water, and the
metal sulfide (KMS) and iron sulfide (KFeS) solubility products. In particular it is true
that if [SEM] < [AVS] then:
(11-2)
[SEMT]
Since the ratio of metal sulfide to iron sulfide solubility products (KMS/KFeS) is very
small « 10*5) even for the most soluble of the sulfides (Table 4-2, Chapter 4), the
metal activity of the sediment is at least five orders of magnitude smaller than the
SEM. This guarantees that no biological effects would be seen if this sediment were
tested. Therefore the condition [SEM] < [AVS] is a no effect sediment quality
criteria.
The reason we use the term "no effect" criteria is that for the condition [SEM]
< [AVS] we expect no biological impacts. For [SEM] > [AVS], which would normally
be considered a criteria violation, there are cases where we would expect no biological
impacts, for example, where significant organic carbon partitioning is occurring. The
most dramatic examples are for sediments with low AVS concentrations (Fig.6-5).
interstitial Water Criteria
The condition [SEM] < [AVS] guarantees that the metal activity of the
sediment - interstitial water system is very low and therefore, below any effect level
of concern. Another way of guaranteeing this is to place a condition on the interstitial
water activity directly. Let us suppose that we knew the metal activity, denoted by
-------
11-5
{FCV}, that corresponded to the Final Chronic Value of the Water Quality Criteria:
[FCV]. Then the SQC corresponding to this effect level is:
{M2*k(FCV} H1-3)
It is quite difficult to measure and/or calculate the metal activity in a solution phase,
{M2+}, at the low concentrations required since it depends on the identities,
concentrations and thermodynamic affinities of other chemically reactive species that
are present. Also the WQC on an activity basis, {FCV}, is not known.
An approximation to this condition is:
[MJ^FCVJ (11-4)
That is, we require that the total dissolved metal concentration in the interstitial water
[Md] be less than the Final Chronic Value from the WQC applied as a dissolved
criteria. Although this requirement ignores the effect of chemical speciation on both
sides of the equation - compare Equation (11-4) to (11-3) - it is the approximation that
is currently being suggested by the EPA for the WQC for metals. That is, the WQC
should be applied to the total dissolved- rather than the total acid recoverable - metal
concentration. Hence, if this second condition is satisfied it is consistent with the
level of protection afforded by the water quality criteria.
In situations where the SEM exceeds the AVS ([SEM] > [AVS]) but the
interstitial water total dissolved metal is less than the final chronic value ([Md] <
[FCVd]), this sediment does not violate the criteria. These cases occur when
significant binding is occurring to other phases. It should be noted that using the final
chronic value for metals requires that the hardness of the interstitial water be known
since the criteria vary as hardness varies.
-------
11-6
AVS and Organic Carbon Criteria
For sediments with an appreciable AVS concentration relative to SEM, the SQC
requirement that [SEM] < [AVS] is a useful result. However if the AVS of the
sediment is small, then the condition is of little value. The reason is that other
sorption phases are present that affect the activity of the system. Similarly, even in
situations where significant AVS occurs in sediments, other sorption phases may
significantly limit the metal activity even if the SEM exceeds the AVS.
Consider the organic carbon in the sediment. It is demonstrated in Chapter 9
that a relationship exists between the SEM that is in excess of the AVS and the
interstitial water metal activity {M2+}:
[SEM]-[AVS]=K*diOCfoc{M2*} (11-5)
where K*d oc is the partition coefficient between the organic carbon of the sediment
and foc is the weight fraction of organic carbon of the sediment. If we require that
the metal activity be at the FCV metal activity, then the SQC for SEM would be:
(11-6)
If the activity is replaced with the WQC total dissolved FCV, then the criteria
becomes:
(11-7)
where Kd oc is the partition coefficient between organic carbon and total dissolved
interstitial water metal concentration. Note that the organic carbon based partition
coefficients vary with respect to pH so that the pH of the interstitial water must be
either measured or estimated. In addition, the FCV for the five metals is hardness
dependent in freshwater so that the hardness is also required.
-------
11-7
This is the third condition from which a sediment quality criteria can be derived.
For sediments where organic carbon provides all the excess binding capacity, it is a
criteria in the ordinary sense. That is, exceeding the criteria would imply that
unacceptable biological impacts would occur. Since the analysis of sediment binding
data and the estimation of the Koc's in Chapter 9 attributes all the binding to organic
carbon thereby indexing the binding to organic carbon, using these constants would
imply that the criteria (Equation 11-8) is the boundary between no effects and effects.
There are situations, however, for which the assumption that organic carbon is the
only important phase may not be correct, in these cases, the criteria becomes a no
effect criteria. Of course, using this as an effect criteria also assumes that applying
the FCV as a total dissolved criteria is appropriate. If, in fact, a significant fraction
of the interstitial water metal is not bioavailable, then again this criteria would be a
no effect criteria.
AVS and Minimum Partitioning Criteria
It would be useful to have a solid phase criteria that would effectively screen
sediments for which the metals concentrations are low enough so that no problem is
anticipated. The idea is to examine sediments for which the partition coefficients are
likely to be quite low. From these sediments it would be possible to establish
minimum partition coefficients (Kd min) which could be applied to any sediment. Then
the no effect SQC would be:
(11-8)
This would also ba a no effect criteria since it is established using a minimum partition
coefficient. No AVS term is included unlike Equation (11-7), because it is assumed
that these sediments have no appreciable AVS. If there was a significant AVS
concentration, then the AVS criteria (Equation 11-1) would apply.
-------
11-8
These single metal criteria are derived only for illustrative purposes. Single
metal criteria are misleading and should not be applied. Only criteria based on the five
metals taken together are valid. They reduce to the single metal criteria in the unlikely
situation where only one metal is present to any significant degree.
Multiple Metals Criteria
The previous section presents the derivation of criteria if only one metal is
present in significant quantities in a sediment. In the usual case, however, it is
insufficient and inappropriate to consider each metal separately. This is of particular
concern for the AVS criteria.
AVS Criteria
The results of calculations using equilibrium chemical models indicate that
metals act in an additive fashion when binding to AVS. That is, each of the five
metals: Cu, Pb, Cd, Zn, and Ni will bind to the AVS and be converted to CuS, PbS,
CdS, ZnS, and NiS in this sequence; i.e., in the order of increasing solubility. The
mixed metals experiment (Figure 5-3) and the Foundry Cove data (Chapter 6) confirm
this behavior. Therefore, the five metals must be considered together. There cannot
be a criteria for just nickel, for example, since 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 SEM, then the SEM will be misleadingly small, and it may appear that
[SEM] < [AVS] when in fact that is not true if all the metals are considered together.
We restrict the discussion below to the five metals listed above. In special situations
where other sulfide forming metals (e.g., Co, Hg, Ag) are in high concentrations they
also must be considered.
-------
11-9
The equilibrium model prediction of the metals activity if a mixture of the metals
are present is similar to the single metals case. If the sum of the SEM's for the five
metals is less than the AVS; i.e.:
£,[SEM,]*[AVS] (11-9)
then:
(11-10)
[SEMliT]
where [SEMi/r] is the total SEM (//mol/L(bulk)) for the ith metal. Thus the activity of
each metal, {Mj}, is unaffected by the presence of the other sulfides. This can be
understood as follows. Imagine that the chemical system starts initially as iron and
metal sulfide solids and that the system proceeds to equilibrium by each solid
dissolving to some extent. The iron sulfide dissolves until the solubility product of
FeS is satisfied. This sets the sulfide activity. Then each metal sulfide dissolves until
it reaches its solubility. Since so little of each dissolve relative to the FeS, the
interstitial water chemistry is not appreciably changed. Hence the sulfide activity
remains the same and the metal activity adjusts to meet each solubility requirement.
Therefore, each metal sulfide behaves independently of each other. The fact that they
are only slightly soluble relative to FeS is the cause of this behavior. Hence the AVS
criteria is easily extended to the case of multiple metals. It is only necessary to sum
the molar concentration of each metal SEM and compare it to the AVS (Equation 11-
9).
Interstitial Water Criteria
The application of the interstitial water criteria to multiple metals is complicated
not by the interactions of the metals chemistry of the sediment - interstitial water
system as in the case with the ,AVS criteria, but rather their possible toxic
interactions. Even if the individual concentrations do not exceed the FCV of each
-------
11-10
metal (FCVj), their additive effect might be toxic. Therefore, to address this additivity,
the interstitial water metal concentrations are converted to toxic units and these are
summed. Since the effects concentration to be used are the final chronic values from
the water quality criteria, we call these toxic units the Interstitial Water Criteria Toxic
Units (IWCTU). For freshwater sediments, the FCV's are hardness dependent for all
of the metals being considered and they need to be adjusted to the hardness of the
interstitial water from the sediment being considered. For the ith metal with total
dissolved concentration [Mifd] the IWCTU for the ith metal is:
(11-11.
The sediment quality criteria requires that the sum of the interstitial water criteria
toxic unit concentration is less than one:
__*1 m.12)
-1
Hence the multiple metals criteria is quite similar to the single metal case (Equation
11-4) except that the criteria is expressed as toxic units and summed.
AVS and Organic Carbon Criteria
The case for which the sediment organic carbon needs to be considered in
addition to AVS is more complicated. Consider, first, a single metal. The relationship
between interstitial water concentrations and sediment concentration for the ith metal
is given by the equation:
[SEMHAVShK^/ocIMJ (1 1-13)
where Kd OC(j is the metal specific partition coefficient between sediment organic
carbon and interstitial water, and [Mj|d] is the total dissolved interstitial water metal
concentration. For this case the interstitial water concentration is predicted using the
-------
11-11
SEM in excess of the AVS and the partition coefficient between the excess SEM and
the interstitial water:
•SJ,OC,POC
In order to apply this equation to the case of multiple metals, it is first
necessary to identify and quantify the metals which are not entirely present as metal
sulfides. The best way to do this is to establish which metals are present as the metal
sulfides and in what quantity. The procedure is to assign the AVS to the metals in
the sequence of their solubility products from the lowest to the highest: SEMCu/
SEMpb, SEMCd, SEM2n, and SEMNj. That is, the AVS complexed metals would be Cu,
followed by Pb, followed by Cd, etc. until the AVS is exhausted. The remaining SEMs
are present in excess of the AVS.
To be specific, let A[SEMj] be the excess SEM for each of the ith metals. The
least soluble metal sulfide (of the five metals being considered in this document) is
copper sulfide (CuS). Thus if the copper SEM is less than the AVS ([SEMCu] <
[AVS]), then all of the copper SEM must be present as copper sulfide (CuS) and no
additional SEM is present so that A[SEMCu] = 0. The remaining AVS is A[AVS] =
[AVS] - [SEMCuJ.
This computation is repeated next for lead because lead sulfide (PbS) is the next
least soluble sulfide. Suppose, unlike copper, the lead SEM is not less than the
remaining AVS ([SEMZn] > A[AVS]). Hence only a portion of the lead SEM is present
as PbS and the remaining SEM, which is denoted as A[SEMpb], is the difference
between the remaining AVS, A[AVS] and the lead SEM: A[SEMPb] = [SEMPb] -
A[AVS]. Thus a portion of the lead is present as lead sulfide, and a portion is excess
SEM.
-------
11-12
Since the AVS has been exhausted by the lead SEM, the remaining metals are
all present as excess SEM so that: A[SEM]Cd = [SEMCd]; A[SEMZn] = [SEMZn]; and
A[SEMNi] = [SEMZn].
For each of these metals, the interstitial water concentrations can be
determined from the appropriate partition coefficients:
(11-15,
•Si.oc.roc
This equation is analogous to Equation (1 1-13) for the single metal case. Note that
if A[SEMj] = 0 then so also is the interstitial water metal concentration. The
interstitial water criteria toxic units are computed using this equation for the interstitial
water concentrations. That is:
FCVJ
where Equation (1 1 -1 4) is used to compute the interstitial water concentrations. Note
that the organic carbon based partition coefficients vary with respect to pH so that
the pH of the interstitial water must be either measured or estimated together with
the hardness if necessary.
The sediment quality criteria requires that the computed total interstitial water
criteria toxic unit concentration is less than one:
1[SEMJ
This criteria is simply the interstitial water toxic unit criteria, Equation (11-12), with
the interstitial water concentrations calculated from the excess SEM for each metal
and the appropriate partition coefficients.
-------
11-13
AVS and Minimum Partition Coefficient Criteria
The no effect criteria that use the minimum partition coefficients (Kd Min {) is
analogous to that using the organic carbon based coefficients:
1 (11-18)
Since the minimum partition coefficients are being used, this would correspond to the
upper bound estimate of the interstitial water criteria toxic units.
Criteria Summary
The proposed Sediment Quality Criteria is as follows. The sediment passes the
SQC if any one of these conditions is satisfied:
(1) AVS Criteria:
£, [SEMJsJAVS] (11-19)
(2) Interstitial Water Criteria:
-------
11-14
(3) AVS and Organic Carbon Criteria:
MSEMJ 2
(4) AVS and Minimum Partition Coefficient Criteria:
If any one of these criteria are violated, this does not mean that the sediment is toxic.
For example, if the AVS in a sediment is virtually zero, then Condition (1) will be
violated. However, if there is sufficient organic carbon sorption so that either
Condition (3) or Condition (4) is satisfied then the sediment is non-toxic.
•
If all - not any but all - of these conditions are violated then there is reason to
think that the sediment may be unacceptably contaminated by these metals. Further
testing and evaluations are therefore required in order to assess the actual level of
toxicity and its causal relationship to the five metals. These may include acute and
chronic tests on species that are sensitive to the metals suspected to be in excess of
the AVS and causing the toxicity. Also in situ community assessments and seasonal
characterizations of the SEM's, AVS, and interstitial water concentrations would be
appropriate.
Sediment Quality Criteria Uncertainty
The methodology for obtaining sediment quality criteria relies on certain
simplifications and, for conditions 3-4, empirical partitioning models to insure that the
metal activity of the sediment - interstitial water system is below effects levels. As
-------
11-15
a consequence, there are uncertainties associated with their use. It is anticipated that
when final SQC are generated, confidence limits will be generated as well. This was
the case for the non-ionic organic chemical SQCs. It is anticipated that the derivation
of the uncertainty limits will be derived in the same way, namely by quantifying the
predictive power of the methods.
For the condition (1 ) criteria relating total SEM and AVS , and for the condition
(2) using measured interstitial water concentrations, sufficient data currently exists
to derive the confidence limits. Since these are both no effect criteria, the confidence
limits are set so that the predictive power of no effect has a high probability of being
the case. That is the criteria requirement would be:
(1 1-23)
where [SEM]SQC 95 is computed so that 95 percent of the tested sediments are
correctly classified as non toxic. Based on the results in chapters 5-7, [SEM]SQC 95
« 0, the theoretical value based on one to one stoichiometric binding of the divalent
metals by AVS.
A similar analysis will be applied to the measured interstitial water toxic unit
condition. The condition:
V .^wcTUsocx ( 1 1 -24)
[FCFJ 5QC'*
where IWCTUSQC 95 is set so that 95 percent of the cases are correctly predicted.
Since we have no chronic data that would show effects at the criteria level, the acute
data base would be analyzed instead:
-------
11-16
/ ^ . —;••»•">'•« v
-------
11-17
The experimental data is available to compute the confidence limits for the no
effect conditions (1) and (2). However, the experiments required to compare
predicted and observed LC50's are not available. The spiked experiments presented
in Chapters 5 to 7 were all dosed with too wide a spacing of concentrations (0.1, 0.3,
1.0, 3.0, and 10.0 multiples of the AVS) to permit calculating a reliable LC50.
Therefore, a selected number of sediments, with widely varying carbon
concentrations, would need to be tested with a much more narrow range of
concentrations. And the spacing would be computed using the predicted LC50 from
the AVS and the organic carbon concentration.
For condition (4), more sandy low carbon sediments need to be evaluated and
the metal partition coefficients measured, from which a probability distribution
analysis could be made. From the small amount of data already available, it is
interesting to note that even these low carbon sediments have Kd oc's that are in the
same order of magnitude as found in the high carbon sediments.
Research Recommendations
There are a number of unfinished areas of research that need to be completed.
1. The organic carbon partition coefficients have been developed for three metals
in freshwater: Cu, Cd, and Pb. The remaining metals need to be completed.
A similar set of experiments are needed for saltwater.
2. Additional experiments need to be conducted, as outlined above, for the
uncertainty analysis of condition 3. These are analogous to the set of
experiments performed for the non-ionic organic chemicals.
3. Additional partition coefficients are needed for the low carbon sediments.
-------
11-18
4. An explicit procedure is needed to handle the cases for which the no effect
criteria are exceeded. Both field and laboratory testing methods need to be
elaborated into a staged investigation.
There are the other metals for which criteria a needed. These include
chromium, arsenic, and selenium. Some initial work has been done for chromium and
arsenic. It appears that AVS can reduce Cr(VI) to Cr(lll), and it may be able to do the
same for the other redox sensitive metals. Further work is necessary to identify the
controlling phases for these metals.
Conclusions
A proposal for establishing sediment quality criteria for cadmium, copper, nickel,
lead, and zinc has been presented. It is based on the Equilibrium Partitioning Model.
The criteria are based on keeping the activity of the sediment - interstitial water
equilibrium system below effects levels. The criteria presented in this report are all
lower bound criteria. That is, if the criteria are satisfied then no effects are expected.
If the criteria are exceeded then further study is required. The difficulties are related
to the presence of multiple binding species in both the solid phase and in the
interstitial water.
The initial solid phase criteria is based on the strongest binding phase, namely
the AVS. If sufficient AVS is present then no effects are expected. If the
simultaneously extractable metal exceeds the AVS, then other binding phases become
important. The next most important phase is organic carbon. A partitioning model
has been suggested that can be used to develop criteria. It is analogous to the
organic carbon normalized model used for the non-ionic organic chemical SQCs. It is
uncertain at present whether any other solid phases need to be considered.
-------
11-19
The interstitial water criteria are based on total dissolved interstitial water
metals concentrations. Therefore, they are analogous to the water quality criteria.
They do not take explicit account of metal speciation and so they are not regulating
the metal activity. Therefore, these are also lower bound criteria. An exceedence of
the water quality criteria in the interstitial water may or may not signal a toxicity
problem. However, if the concentrations are below the WQC then no effects are
expected.
It must be stressed that the sediment quality criteria are aquatic life criteria that
apply only to benthic organisms. They do not address the water column
consequences of contaminated sediments. Water column concentrations are
determined by the transport of metals from the sediment to the overlying water. The
resulting concentrations would be compared to the water quality criteria for metals.
This is a separate evaluation that needs to be made if water column effects are
suspected.
-------
15. Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Hicks, M.B., Mays, S.M.
and Redmond, M.S.. 1990. Toxicity of cadmium in sediments: The role of acid
volatile sulfides. Environ. Toxicol. Chem. 9:1487-1502.
16. Ankley, G.T., Thomas, N.A., Di Toro, D.M., Hansen, D.J., Mahony, J.D., Berry,
W.J., Swartz, R.C., Hoke, R.A.. 1994. Assessing potential bioavailability of metals
in sediments: A proposed approach. Environ. Mgt. 18:331-337.
17. ASTM. 1993. Standard guide for conducting 10-day static sediment toxicity tests
with marine and estuarine amphipods. 1993 Annual Book of ASTM Standards,
Chapter 11, Water and Environmental Technology. American Society for Testing
and Materials. 11.04:1139-1163.
18. Mearns, A.J., Swartz, R.C., Cummins, J.M., Dinnel, P.A., Plesha, P., and Chapman,
P.M. 1986. Inter-laboratory comparison of a sediment toxicity test using the
marine amphipod, Rhepoxvnius abronius. Marine Environ. Res. 19:13-37.
19. Cornwall, J.C. and Morse, J.W. 1987. The characterization of iron sulfide minerals
in recent anoxic marine sediments. Mar. Chem. 22:55-69.
20. Boothman, W.S. and Helmstetter, A. Vertical and seasonal variability of acid
volatile sulfides in marine sediments. Manuscript.
------- |