United States
Environmental Protection
Agency
Off ice of Water
Regulations and Standards
Criteria and Standards Division
Washington DC 20460
August 1987
SCD# 11
vvEPA
Water
EVALUATION OF THE EQUILIBRIUM PARTITION
THEORY FOR ESTIMATING THE TOXICITY OF THE
NONPOLAR ORGANIC COMPOUND DDT TO THE
SEDIMENT DWELLING ORGANISM
RHEPOXYNIUS ABRONIUS
Organic Carbon (%)
O 0.12
0 0.25
• 0.58
Organic Carbon (%)
O0.60
• 0.80
D 1.75
• 1.92
DDT Equilibration in Interstitial Water with Different
Organic Carbon Content of Sediment a) Three Lowest and
b) Four Highest Percentages
-------
EVALUATION OF THE EQUILIBRIUM PARTITIONING
THEORY FOR ESTIMATING THE TOXICITY OF THE
NONPOLAR ORGANIC COMPOUND DDT TO THE SEDIMENT
DWELLING AMPHIPOD RHEPOXYNIUS ABRONIUS
WA56, Task 1
December 1987
Prepared by:
J. Q. Word, J. A. Ward, L. M. Franklin,
V. I. Cullinan and S. L. Kiesser
Battelle/Marine Research Laboratory
Sequim, Washington
Prepared for:
Criteria and Standards Division
U.S. Environmental Protection Agency
Washington, D.C.
Submitted by:
BATTELLE
Washington Environmental Program Office
Washington, D.C.
-------
ABSTRACT
This report describes research conducted by Battelle/Marine Research
Laboratory to determine if the equilibrium partitioning (EP) theory can be
used to predict acute toxicity of a sediment-associated nonpolar organic con-
taminant (DDT) to the sediment dwelling amphipod Rheopoxym'us abronius. After
considerable protocol development, three research hypotheses were examined.
The first hypothesis stated that DDT in interstitial water was the primary
determinant of acute toxicity. This hypothesis was supported by results of
regression analyses that showed that the interstitial water DDT concentration
was the only significant predictor variable. The second hypothesis asserted
that the organic partitioning model accurately predicted the concentration of
DDT in interstitial water. Again, this hypothesis was supported by research
results. Predicted and measured concentrations were nearly identical with
predicted concentrations exceeding measured values by only a factor of 0.15.
The third hypothesis stated that the source of sediment-bound organic carbon
influenced the toxicity of DDT. This hypothesis was supported by the finding
of significantly different (p<0.05) LC5Q values in bioassays where only the
source of organic carbon was different. The LC50 for interstitial water from
Sequim Bay sediments was 12.5 /tg/L, while the LC5Q for interstitial water
from Point Pulley sediments was 1.9 ng/L. The information gained from these
studies suggested that the EP theory can be used in predicting biological
effects for nonpolar organics having a high affinity for organic carbon, under
static, acute bioassay conditions. Future application of EP theory to
sediment quality criteria development will depend on its ability to predict
effects under field conditions and with chemicals with lower binding
affinities for organic carbon.
m
-------
CONTENTS
ABSTRACT in
1.0 INTRODUCTION 1
2.0 MATERIALS AND METHODS 5
2.1 BIOASSAY MATERIALS 5
2.1.1 Contaminant 5
2.1.2 Test Organism 5
2.1.3 Sediments 7
2.1.4 Laboratory Equipment 9
2.2 GENERAL BIOASSAY PROTOCOLS 11
2.2.1 Randomization 11
2.2.2 Water Quality Parameters 11
2.2.3 Mortality Determination 11
2.2.4 DDT Measurements 12
2.2.5 Dissolved Organic Carbon Measurements 12
2.3 EVALUATION AND DEVELOPMENT OF BIOASSAY PROTOCOLS 13
2.3.1 DDT Transformation 14
2.3.2 Interstitial Water Sampling Methods 16
2.3.3 Sediment and Water Labeling Procedures 17
2.3.4 Time to Equilibrium Experiments 18
2.4 WATER COLUMN BIOASSAYS 19
2.4.1 Acute Toxicity of Water Column Bioassays 19
2.5 SEDIMENT BIOASSAYS 20
2.5.1 Radiolabeling Sediments with DDT 20
2.5.2 Acute Toxicity Sediment Experiments 21
IV
-------
2.5.3 Sediment DOT Concentration 22
2.6 TEST OF RESEARCH HYPOTHESES 22
2.6.1 Toxicity is Determined by the Interstitial
Concentration of DDT 22
2.6.2 Equilibrium Partitioning Model Accurately
Predicts the Concentration of DDT in
Interstitial Water 24
2.6.3 Organic Carbon Source Influences Toxicity 24
3.0 RESULTS 25
3.1 EVALUATION AND DEVELOPMENT OF BIOASSAY PROTOCOLS 25
3.1.1 DDT Transformation 25
3.1.2 Interstitial Water Sampling Methods 26
3.1.3 Sediment and Water Labeling Procedures 27
3.1.4 Time to Equilibrium Experiments 27
3.2 WATER COLUMN BIOASSAYS •. 31
3.3 SEDIMENT BIOASSAYS 34
3.4 TEST OF RESEARCH HYPOTHESES 39
3.4.1 Toxicity is Determined by the Interstitial
Concentration of DDT 39
3.4.2 Equilibrium Partitioning Model Accurately Predicts
the Concentration of DDT in Interstitial Water ... 40
3.4.3 Organic Carbon Source Influences Toxicity 40
3.4.4 Relationship Between Particulate Organic Carbon
and Dissolved Organic Carbon Concentrations in
Interstitial Water 44
4.0 DISCUSSION AND CONCLUSIONS 47
4.1 PROTOCOL DEVELOPMENT 47
4.1.1 DDT Transformation 47
4.1.2 Interstitial Water Sampling Methods 48
-------
4.1.3 Evenness of Labeling in Sediments and Water 49
4.1.4 Determining DDT in All Test Containers 50
4.1.5 Equilibrium Experiments 50
4.2 FINDINGS ON TOXICITY AND EQUILIBRIUM PARTITIONING THEORY . . 51
4.2.1 Research Hypothesis 1 51
4.2.2 Research Hypothesis 2 52
4.2.3 Research Hypothesis 3 52
4.2.4 Additional Experimental Issue 53
4.3 APPLICATION OF RESULTS TO SEDIMENT QUALITY
CRITERIA DEVELOPMENT 53
4.4 FUTURE RESEARCH NEEDS 54
5.0 REFERENCES 57
vi
-------
FIGURES
2.1 Amphipod Dredge Used to Collect Organisms 6
3.1 Rate of Equilibrium Development of DDT Concentration
in Three Independent Water Column Experiments 28
3.2 Percent Coefficient of Variation of DDT Concentration in
Water Column Experiment 1 29
3.3 DDT Equilibration in Interstitial Water (Inter)
and Overlying Water (Over) in a) 5% Glycine and
b) 0.05% Glycine Solutions 30
3.4 DDT Equilibration in Interstitial Water with Different
Organic Carbon Content of Sediment a) Three Lowest and
b) Four Highest Percentages 32
3.5 Amphipod Survival in Water Column Bioassays:
a) First Test, 10 days, b) Second Test, 4 days,
and c) Second Test, 10 days 33
3.6 Survival of Rhepoxynius in a) Exposure Water without Carrier
and b) Exposure Water With Glycine Carrier 34
3.7 Survival of Rhepqxynius in a) Point Pulley Sediment and
b) Sequim Bay Sediment 36
3.8 Survival of Rhepoxvniiis in a) Sequim Bay and West Beach
Sediment Mixture and b) West Beach and Point Pulley
Sediment Mixture 37
3.9 Survival of Rhepoxvm'us in a) Point Pulley/West Beach
Sediments, b) West Beach Sediment and c) West Beach/Sequim
Bay Sediments 38
3.10 Predicted versus Measured Interstital DOT Water
Concentrations 41
3.11 Toxicity Between a) Point Pulley/Sequim Bay Sediments and
b) Sequim Bay/West Beach Sediment with West Beach/Point
Pulley Sediment Mixture 42
3.12 Total Organic Carbon (TOC) to Dissolved Organic
Carbon (DOC) 45
vii
-------
TABLES
2.1 Physical and Chemical Characteristics of Sediments Used for
Definitive Bioassay 8
2.2 Variation in Organic Carbon Concentration Within Mixed
Sediments 10
2.3 Summary of Experimental Conditions for Sediment
DDT Transformation Study 16
3.1 Transformation and Loss of DDT in Seawater 26
3.2 Transformation and Loss of DDT in Sediments 26
3.3 DDT Recovery from Water Using Different
Collection Techniques 27
3.4 Survival of Amphipods in Various Sediment Types; LCso Values Are
Determined Based on Interstitial Water (I.W.) Concentrations and
on I.W. Concentration Normalized to the Sediment Bound Organic
Carbon Concentration 35
3.5 Variation in LCso Concentrations Calculated for Different
Sediment Mixtures and Results of Statistical Tests of
Significance for Influence of Organic Carbon Source 43
viii
-------
1.0 INTRODUCTION
The Criteria and Standards Division (CSO) of the U.S. Environmental
Protection Agency (EPA) is evaluating methods to establish sediment quality
criteria that, when used in association with water quality criteria, will
protect the environment consistent with the provisions of the Federal Water
Pollution Control Act of 1972 as amended by the Clean Water Acts of 1977 and
1978. Extensive data bases already exist from which water quality criteria
were and continue to be developed. Ideally, if a consistent relationship can
be found between the water quality criteria and maximum acceptable concentra-
tion in sediment for a particular chemical, then using the extensive water
quality data base for deriving sediment quality criteria would result in
substantial savings to the agency.
The primary objective of the research conducted by the Battelle/Marine
Research Laboratory (MRL) for the U.S. Environmental Protection Agency was to
evaluate the equilibrium partitioning (EP) theory as a candidate method that
would allow this extrapolation to be made. The theory may provide the theo-
retical basis for understanding how complex interactions between chemicals and
sediment influence toxicity of the chemicals to benthic organisms. Under-
lying assumptions of the EP approach are that a chemical establishes equili-
brium between the dissolved and sediment-associated fractions and that the
distribution between dissolved and sediment-associated fractions can be
described by a partition coefficient and additional information on quantity,
phase, and source of organic carbon or other binding agents in the sediment.
Applying EP theory to sediment quality criteria development is based on
the additional assumption that the only toxic fraction of a sediment
contaminant is that fraction freely dissolved within interstitial waters.
This assumption permits data from water quality criteria, or other aquatic
toxicity data to be used to derive sediment quality criteria. The influence
of sediment-bound and dissolved organic carbon on the dissolved fraction of
nonpolar organic compounds and its resultant toxicity, or bioaccumulation,
has been studied in a few cases (Adams, Kimerle, and Mosher 1985; Knezovich
and Harrison 1986; McCarthy and Jimenez 1985a, b; McCarthy, Jimenez, and
Barbee 1985). In these cases, contaminants bound to organic materials, either
1
-------
dissolved in solution or attached to sediment were less toxic and also less
available for uptake by fish and cladocerans than dissolved contaminants.
Adams, Kimerle, and Mosher (1985) showed that as the concentration of par-
ti culate organic carbon increased, contaminant concentrations measured in
interstitial water became increasingly larger than the concentrations that
were predicted through their partition coefficients. They also showed that
both toxicity and bioaccumulation were related to the concentration of the
contaminant in interstitial water. The difference in the predicted versus
measured interstitial water concentration may have been due to complexation
of the contaminant by dissolved organic carbon in the interstitial water. If
the concentration of organic carbon dissolved in the interstitial water and
its effect on the concentration of the dissolved contaminant are taken into
account, a more accurate model for predicting the dissolved concentration of
nonpolar organic contaminants may be developed.
To evaluate the accuracy of EP theory in predicting the toxicity of
sediment-associated nonpolar organic contaminants, three research hypotheses
were examined: 1) the concentration of the contaminant in solution in the
interstitial water is the primary determinant of acute toxicity, 2) the toxic
concentration can be predicted from the organic carbon/water partition coeffi-
cient and the concentration of organic carbon in the sediment, and 3) the
source of the organic carbon influences the toxicity.
To accomplish our objective, experimental protocols needed to be evalu-
ated and new methods developed that would allow us to better understand the
conditions under which the test organisms were exposed. Experiments were
required to demonstrate that the selected toxicant did not undergo chemical
transformation and alter the compound assumed to be accounting for the toxi-
city. Methodology also had to be developed.to label and mix sediments to
ensure even sediment and interstitial water distribution within the experi-
mental habitat of the organisms being tested. In addition, several sampling
methods for collecting interstitial water were evaluated to ensure that the
selected method produced accurate measurements of both the contaminant and
organic carbon in the dissolved phase. To state with confidence the concen-
tration of contaminant to which an organism is exposed, and to compare the
-------
predicted versus the measured concentration in interstitial water, the experi
mental systems had to be at equilibrium before the bioassays were initiated.
Preliminary experiments were also required to determine the time for a con-
taminant to establish equilibrium between the dissolved and particulate
phases.
The amphipod Rhepoxynius abrom'us was selected as the test organism in
these bioassays because of its apparent sensitivity to nonpolar organic pesti-
cides, its ability to survive in a wide range of sediment types, and its
burrowing, sediment-dwelling lifestyle, which ensures exposure predominantly
to interstitial water (Enquist 1954; Word and Striplin 1981; Word 1978, 1980;
Swartz et al. 1985). The organism is also an ideal test subject because
bioassay procedures using this animal are widely recognized and approved as
state-of-the-art techniques for assessing sediment toxicity to marine
organisms (Standard Methods 1985; Swartz et al. 1985, 1986).
DDT was selected as the test contaminant according to three criteria
established by Poston and Prohammer (1986): 1) the reported median lethal
concentration and median effective concentration (LC5Q and EC50) for 48 or
96 h acute toxicity tests with amphipods are less than 20% of the reported
solubility (5.5 /tg/L) of the toxicant in water; 2) because the organic
carbon/water partition coefficient (KQC) has an average value of 3.3 X 105
(Pavlou et al. 1987) there is also reasonable probability that concentrations
could be established in the interstitial water that would elicit a toxic
response in the test organism; and 3) DDT has a low vapor pressure, i.e.,
less than 0.0001-mm mercury, ensuring that excessive volatilization does not
cause a loss of toxicant when amending the sediment. DDT also represents a
nonpolar organic contaminant that is very strongly sorbed to sediments and
one that is a known contaminant of sediments in many places throughout the
world.
Section 2.0 of this report describes the materials and methods. Results
are presented in Section 3.0; Section 4.0 provides a discussion of the
results, our conclusions, and recommendations for future research.
-------
2.0 MATERIALS AND METHODS
Section 2.1 includes descriptions of bioassay materials, test organisms,
the laboratory handling of samples and laboratory equipment used. In Sec-
tion 2.2, the general bioassay and laboratory protocols are reviewed. Evalua-
tion and development of bioassay protocols, sampling methodology, and
experiments for validating particular methods are found in Section 2.3. The
water column and sediment bioassay protocols are contained in Sections 2.4
and 2.5, respectively. Finally, Section 2.6 describes statistical methodology
used to evaluate the results of these experiments.
2.1 BIOASSAY MATERIALS
2.1.1 Contaminant
It was necessary to measure DDT concentrations in a few milliliters of
interstitial water at the parts per trillion level. Because the only avail-
able technique for measuring such low concentrations of DDT in these small
samples is liquid scintillation counting of radioactive decay products, a
mixture of radiolabeled and unlabeled DDT was used in all experiments. The
14
radioactive DDT was labeled with C on the carbon atoms of the central ring.
This radiolabeled DDT was custom synthesized by New England Nuclear Corp., a
division of Dupont Industries. Specific activity of the radiolabeled DDT was
13.5 mCi/mM. Purity of the radiolabel was greater than 98% as determined by
New England Nuclear using silica gel GF autoradiograms. Chemical purity of
the unlabeled DDT (GOLD-LABEL grade) was greater than 99%, as determined by
the manufacturer, Aldrich.
2.1.2 Test Organism
Rhepoxym'us abronius were collected before each test and held for a
period of not less than 5 or more than 10 days before they were used in
experiments. Organisms were collected from control sediments at Whidbey
Island, Washington, at a water depth of approximately 2 m with a towed hand
dredge constructed by Battelle/MRL (Figure 2.1). Aboard ship, the dredge
-------
FIGURE 2.1. Dredge Used to Collect Amphipods
contents were transferred to holding tanks containing native sediments and
covered with water. Any obvious large predators were removed. The holding
tanks were transported to the laboratory where the sediments were screened
and R. abrom'us separated from other species. Identification of R. abronius
was initially made using specialized taxonomic literature (Barnard 1960, 1979;
Barnard and Barnard 1982). Subsequently, it was found that the Whidbey Island
collection site contained very few other species of amphipods, and that the R.
abronius were sufficiently distinct to be easily recognized without the
destructive dissections that are normally required for accurate identifica-
tion. During the recommended holding period, any dead animals or any animals
behaving abnormally were removed from the tank. As recommended by Swartz et
al. (1985), small quantities of dried, flaked fish food were provided during
the holding period, but not during the experiments. The temperature, salinity
and photoperiod under which the organisms were held corresponded to natural
-------
seasonal conditions for Sequim Bay. In July, temperature was 15 * 1°C;
salinity ranged between 28-32 °/oo, and the photoperiod was regulated at 16 h
light to 8 h dark.
2.1.3 Sediments
2.1.3.1 Selection of Collection Sites
The criteria used to select sediment collection sites for the bioassays
included the following: 1) very low levels of contamination, 2) undetectable
DOT concentrations, 3) grain sizes dominated by fine sands and silts,
4) within the correct range of organic carbon, 5) organic carbon sources from
geographically different areas, and 6) areas that were relatively isolated
and under minimal .influence from industrial, agricultural, or domestic efflu-
ent. Based on review of the available literature [Battelle's Eight Bay Study
(Pacific Northwest Laboratory 1986), Municipality of Metropolitan Seattle's
(Metro) Seahurst, Duwamish Head, and Toxicant Pretreatment Planning Study
(unpublished data records, R. C. Farlow, U.S. Environmental Protection Agency,
personal communication)], sites for sediment collection were selected. The
measured concentrations of organic contaminants at each site are contained in
(Table 2.1). These sediments represented normal ranges of test conditions
for R. abronius (Swartz et al. 1984) and had the desired range of organic
carbon contents.
2.1.3.2 Collection and Storage
Sediments were collected with a modified van Veen grab sampler. Station
location was verified by Loran C and radar range information taken from the
Seahurst and Duwamish Head Baseline studies and by range fixes for Sequim Bay
and West Beach sediments. The upper 2 cm of each acceptable grab were removed
using stainless steel implements and transferred into sterile containers.
Sediments were stored on ice, transported to the laboratory the same day col-
lected, and stored in a walk-in cold room at a temperature of approximately
4°C.
2.1.3.3 Mixing Sediments
Intermediate concentrations of organic carbon in sediments were prepared
by mixing sediments with high and low concentrations of organic carbon. In
-------
TABLE 2.1. Physical and Chemical Characteristics of Sediments Used for Definitive Bioassay.
Detection limits were 20 pg/kg dry sediment.
Designation
Native Sediment
Low-Range
Organic Carbon
(0.11-0.13%)
Location
West Beach,
North Whidbey,
Island, HA
Sediment Characteristic
No organic compounds
detected
Concentration Ug/kq dry)
Mean S.D.
00
High-Range
Organic Carbon
(1.69-1.80%)
Station G-780
East Passage off
Point Pulley,
Southern Puget
Sound, WA
Isophorone
Naphthalene
Penantrene
Fluoranthene
Di-N-butyl phthalate
Di-Octyl phthalates
Benzo
Benzo
Benzo
fluoranthene
fluoranthene
pyrene
40
20
40
63
30
55
55
55
40
8
4
8
13
9
16
11
11
8
High-Range
Organic Carbon
(1.86-2.02%)
Station 17
(Battelle Eight
Bay Study)
Sequim Bay
Sequim, WA
No organic compounds
detected
-------
all cases, sediments characterized as having low organic carbon concentration
were West Beach native controls. The quantity of each of the sediments in
the mixture was determined using the following formula:
X (OCL) + (Y-X)(OCH) = Y(OCM) (1)
where Y is the desired weight of the final sediment mixture in grams; X is
the weight of the low organic carbon sediment in grams; and OC,, OCH, and OC»
are the organic carbon contents of the sediment for low, high, and mixed
sediments, respectively.
The initial organic carbon concentrations in the sediments are expressed
on a dry weight basis and were determined by AMTEST, Inc. Global Geochemis-
try provided the measurements for evaluating the success of the mixing
process. The sediments were mixed by continuously rolling for 48 to 96 h.
The evenness of the organic carbon in the mixtures was determined for three
replicate measurements from aliquots of each mixed sample. The target level
of variation was less than 10% of the mean of replicate samples. When the
variation was higher than the target level, the samples were rolled again
until the variation was within the target range. All sediment mixtures except
the West Beach/Point Pulley mixture met this requirement. Final organic
carbon concentrations and the variation in those concentrations are presented
in Table 2.2.
2.1.4 Laboratory Equipment
Wherever possible, only glass, fluoroplastics (Teflon), silicon, and
stainless steel contacted test water. The holding and acclimation tanks were
thoroughly cured and conditioned by continuous exposure to flowing seawater
for the last 10 years. Tests were conducted to determine if treating glass-
ware with an inert coating affected the loss of DDT from solution. Because
HMDC or Prosil coatings of the glassware did not significantly reduce the
amount of DDT lost from the solution, glass containers without coatings were
used for the rest of the experiments.
-------
TABLE 2.2. Variation in Organic Carbon Concentration Within Mixed
Sediments (a)
Mixture
West Beach/Sequim Bay
(93) (7)
West Beach
Point Pulley /West Beach
(42) (58)
West Beach/ Point Pulley
(72) (28)
Sequim Bay /West Beach
(27) (73)
Sequim Bay
Point Pulley
Organic Carbon /.\
(% Dry Weight) Mean S.D.W
0.26
0.24 0.25 0.01
0.26
0.12
0.13 0.12 0.006
0.12
0.84
0.73 0.80 0.064
0.84
0.56
0.71 0.58 0.121
0.47
0.63
0.59 0.60 0.031
0.57
2.02
1.86 1.92 0.085
1.89
1.69
1.80 1.75 0.056
1.76
%CV(b)
4.0
5.0
8.0
20.9
5.0
4.0
3.0
(a) Data provided by Global Geochemistry, Los Angeles, California.
(b) %CV = Coefficient of variation; S.D. = Standard deviation.
(c) Relative percent of sediment in mixture.
10
-------
All glassware and experiment materials were cleaned with hot soapy water,
rinsed three times with distilled water, rinsed three times with methylene
chloride, dried in a controlled environment, and covered with aluminum foil
or used immediately.
2.2 GENERAL BIOASSAY PROTOCOLS
2.2.1 Randomization
The experimental design for the sediment bioassays incorporated a random-
ized block design for distributing the test containers among the water tables.
All containers were assigned a random number generated from a table of random
numbers developed for this design. The only designation on each container
was this random number, thus creating an experimental protocol in which the
analyst's observations were blind to treatment.
2.2.2 Water Quality Parameters
Ambient seasonal conditions for temperature, salinity, dissolved oxygen,
pH, and photoperiod were maintained during bioassays. In July, temperature
was 15 * 1°C, salinity was 30 * 2°/oo, dissolved oxygen was 80 to 100% satur-
ation at 6 * 1 rnL/L, pH was 8.0 * 0.5 units, and photoperiod was regulated at
16 h light to 8 h dark. Dissolved oxygen, salinity, and temperature were
measured daily. Dissolved oxygen was measured using a Yellow Springs Instru-
ment (YSI) dissolved oxygen probe, which was calibrated daily before and after
the measurements. Measurements of pH were made on the first, fourth, and
tenth days of ongoing tests with an Orion pH meter that was calibrated on
each day of use. Salinity was measured using a refractometer, which was cali-
brated monthly against standard Copenhagen seawater. Temperature was measured
using a mercury thermometer that was calibrated monthly against a National
Bureau of Standards (NBS) standard reference thermometer. Before use, all
seawater was filtered through large sand filters and again through a steril-
ized in-line Gel man filter, 0.45 /un.
2.2.3 Mortality Determination
An organism was determined to be dead when there was no movement in pleo-
pods after stimulation of each individual with a glass probe. Verification
11
-------
of the accuracy of these observations was made by re-examining and recounting
organisms in 10% of the exposure containers. This method of determining death
is consistent with recognized procedures (Swartz et al. 1984).
2.2.4 DDT Measurements
For the transformation tests, gas chromatography (GC) quantification of
DDT and DOE in the sediment and water was achieved by high-resolution gas
chromatography using a Hewlett-Packard 5840 gas chromatograph equipped with
an electron capture detector (GC-ECD). Response factors for DDT and DDE were
calibrated using the external standard method. The linear dynamic range of
the GC-ECD was determined to be from 0.002 to 0.2 pg/mL for DDT. Samples
were diluted to be within this range for quantification. A 30 m x 0.25 nrni
inside diameter (i.d.)-fused silica capillary column coated with 0.25 fan DB-5
resin (J&W Scientific) was used. Helium was used as a carrier gas at a linear
velocity of 50 cm/s and argon/methane (95:5 ratio) was used as the make-up
gas. Samples were injected by the split vaporization mode (1:10 split) with
the injector temperature at 285°C. The temperature of the GC oven was iso-
thermal at 250°C.
For scintillation counting, a Beckman LS-150 liquid scintillation counter
was used for counting samples containing radiolabeled DDT. These samples
were placed in 20 ml disposable scintillation vials and then 15 ml of scin-
tillation cocktail (Aquasol-II) was added. Vials were capped and labeled.
The sides of each vial were wiped with 95% ethanol and Kimwipe cleaners.
These vials were then placed into a scintillation counter behind the command
tower. After all vials were placed into the counter, the lid was closed and
samples were incubated in the dark for a minimum of 1 h prior to starting the
counter. Preset errors were established at 2% and preset time at 50 minutes.
2.2.5 Dissolved Organic Carbon Measurements
Dissolved organic carbon was measured using an Oceanography International
Horiba PIR-2000 Carbon Analyzer (McQuaker and Fung 1975). To remove any
potential organic contamination, ampules used in the organic carbon analysis
were precombusted in a kiln at 550°C for 6 h prior to use. One ml of sample
water after centrifugalion (Section 2.3.2) was added to each ampule and
12
-------
diluted to a final volume of 10 ml with distilled water (at 10X dilution).
Then 1 ml of saturated potassium sulfate solution and 0.2 ml of 10% phosphoric
acid solution were added to each sample. Samples were purged of inorganic
carbon by bubbling purified oxygen through the solution for approximately
8 minutes. The ampules were then sealed and stored for later analysis.
Samples were analyzed on the PIR-2000 and the resulting infrared spectrophoto-
metric (IR) units compared with a standard carbon curve. Organic carbon con-
centrations in mg/L were determined after correcting for the organic carbon
concentration in reagent blanks.
2.3 EVALUATION AND DEVELOPMENT OF BIOASSAY PROTOCOLS
To accomplish the objectives of the study, several experimental proto-
cols were evaluated and new protocols developed that would ensure that the
conditions under which the test organisms were exposed could be accurately
determined. Specific protocol evaluations included the following:
• DDT Transformation - Chemicals may change from one form to another
through chemical, physical, and biological processes. This trans-
formation may alter the resulting toxicity. Establishing whether
DDT is transformed and determining its rate of transformation under
conditions similar to those used in the bioassays were crucial to
further experiments. Further experiments with DDT could only be
conducted if little or no transformation took place or if a very
rapid transformation occurred and steady-state conditions could be
ensured.
• Experimental contaminant distribution - It was important to estab-
lish the precise DDT concentrations to which the organisms were
exposed. Therefore, several preliminary tests were run to examine
loss of DDT to containers and evenness of the DDT label within column
water, interstitial water, and sediments.
• Experimental contaminant subsampling from interstitial water - The
materials and techniques used to collect interstitial water samples
were a critical aspect of the bioassay experiments; certain materials
13
-------
and methods could result in loss of the DDT and, thus, inaccurate
evaluation of the exposure concentrations. Preliminary tests were
conducted to assess the collection efficiency of different methods
and select the method that minimized sample loss during the sampling
process.
• Sediment labeling - It was important to ensure that the concentra-
tion of contaminant on the sediment was evenly distributed at a fine-
ness consistent with the area where the organism resided. Several
methods for labeling the sediments were evaluated.
• Equilibration or steady-state rate experiments - Theoretically, an
equilibrium will develop for a contaminant in the dissolved state.
To ensure that the exposure concentrations were known and constant,
we determined toxicity in the water column and sediment bioassays at
this point of equilibration. Moreover, we had to ensure that the
concentration of DDT in the dissolved phase of the exposure system
remained constant throughout the life of the bioassay. If the
equilibration point were to change rapidly during either bioassay
experiment, then the actual exposure concentrations would be unknown
and results invalid.
2.3.1 DDT Transformation
Transformation studies were conducted in water and sediments to deter-
mine whether DDT would transform into DDE under the experimental conditions
and during the bioassay experiments. Reagent grade (99% purity) DDT dissolved
in ethanol was used as a standard solution for GC measurements. After
adjusting the oven temperature and gas flow for the columns in use, so that
4,4' DDT had a retention time of approximately 12 minutes, the injected stan-
dards were found to have DDT degradation products approximating 11 to 21% of
the total DDT. This transformation is relatively common, especially in gas
chromatographs that are used extensively and that accumulate residual contami-
nants near injection ports, and should be considered during analysis. The
DDT did not transform to DDE in measurable quantities after the instrument
was cleaned and after adjustments were made to the gas chromatograph as
indicated in standard EPA Method 8080 (U.S. EPA 1986). It is essential that
14
-------
standards of known quality be compared side by side with experimental samples
to determine if the gas chromatograph is transforming the DOT. This type of
transformation observed with DOT also occurs with other pesticides.
The transformation tests provided information on the following questions:
1) the extraction efficiency through time of DDT from the water and in sedi-
ments, 2) the transformation rate of DDT into DDE after release into seawater
or sediments, and 3) the degree of evenness in labeling sediments using alter-
native mixing strategies.
The first set of tests evaluated the potential for DDT transformation in
water. The stock solution contained 20 mg DDT mixed in 100 ml ethanol (final
DDT concentration of 200 mg/L). For the transformation tests, 20 ml of this
solution was added to 480 ml of seawater in polystyrene test beakers, yielding
an initial concentration of 8 ppm (4 mg DDT, 20 ml ethanol, and 480 mL sea-
water), or an apparent water column concentration of 8000 ng DDT/mL in sea-
water. On days 1, 4, 5, 6, 7, and 10, 10-mL samples of this solution were
extracted for subsequent measurement of the DDT and DDE concentration on the
gas chromatograph. Samples that were expected to be held more than 7 days
were immediately frozen upon removal.
The sediment transformation and all other tests were performed in glass-
ware. Temperature was maintained at 15 * 1°C and photoperiod was regulated
at 16 h light to 8 h dark. Sediment transformation tests used Sequim Bay
sediments, which contain relatively high concentrations of organic matter
(approximately 2%) and which are characterized as fine-grained silts and clays
(Table 2.1). The sediment and contaminants were well mixed before use. Six
experiments were conducted; the experimental conditions are summarized in
Table 2.3; the results of the tests are included in Table 3.2. The sediment
mixtures were placed on a roller for 5 days. After rolling, the sediment was
allowed to settle for approximately 2 h, at which time the carrier (ethanol)
was removed by pipette from the containers of experiments 1 and 2. After the
carrier was removed, 100 ml of seawater was added and the mixture was rolled
overnight. Samples without carriers were continuously rolled. The sediments
were allowed to settle for 2 h, the water removed, 100 ml of seawater added,
and the mixture rolled overnight. This procedure was repeated and the mixture
15
-------
TABLE 2.3. Summary of Experimental Conditions for Sediment
DDT Transformation Study
Ethanol
Experiment
Number
1
2
3
4
5(a)
6
DDT
(mg)
40
40
40
40
40
40
Sediment
(a)
100
100
100
100
0
100
Carrier
(ml)
10
10
0
0
0
0
Seawater
(ml)
100
100
100
100
0
100
(a) Recovery standard, water only.
was rolled for 2 additional days before the DDT concentration in the water and
sediments was determined. Experiment 5 was performed to test the recovery
efficiency of our DDT, DDE analytical method.
Sediments were extracted using the following procedure. First, to remove
water from the sediments, approximately 10 g of sediment was placed in a pre-
viously cleaned 250-mL jar and mixed with 20 g of NagSO^ After this dehydra-
tion step, 100 ml of methylene chloride was added to the jar, and the sample
was rolled overnight. The methylene chloride was then separated from the
sample by pipette and stored. An additional 20 ml of methylene chloride was
added to the container, swirled, and again removed by pipette and then added
to that collected during the previous step. This procedure was repeated
twice. An aliquot of the final volume of methylene chloride was then injected
into the gas chromatograph to determine the concentrations of DDT and DDE.
2.3.2 Interstitial Mater Sampling Methods
Several interstitial water sampling apparatus and methods were evaluated
to determine recovery efficiency and to select a technique that would mini-
mize loss of DDT during sampling. Plastic syringes and glass syringes with
and without glass fiber filters and with or without 0.45-^n filters were used.
Centrifugation was also examined. Recoveries of DDT from spiked water were
determined for the various sampling apparatus. The initial DDT concentrations
16
-------
in 0.45-^m-filtered water, was 4 jtg/L. Each of these techniques provided
approximately the same volume of water per unit volume of sediment processed.
Based on the results of the methods that were tested (see Section 3.1.2),
the method yielding the best results is as follows:
1. The tapered end and 2 ml of each pipette was removed and the ends
fired to make them smooth.
2. The pipette was introduced into the water with the opposing end
capped.
3. When the end of the pipette touched the surface of the sediment, the
capped end was uncapped.
4. The pipette was slowly pushed into the sediment for a distance of
1 to 2 cm.
5. The end of each pipette was again capped.
6. The pipette containing the sediment sample was carefully withdrawn
from the container.
7. The core of sediment was placed into pre-weighed disposable
13 X 100 mm culture tubes and capped.
8. The tubes were centrifuged at 2000 rpm for 10 minutes, which provided
a 760 x g force.
9. The supernatant obtained from these small cores (1 to 2 ml) was
removed with a 1-mL pipette, the volume recorded and identified as
the interstitial water sample.
10. The samples were then analyzed for radioactivity (Section 2.2.4) or
for dissolved organic carbon concentration (Section 2.2.5).
2.3.3 Sediment and Water Labeling Procedures
We considered several candidate techniques that could be used to produce
uniform distributions of DDT in the water column and sediment test systems.
We did not choose to use solvents as Adams, Kimerle, and Mosher (1985) and
others have used because DDT and other contaminants do not enter natural sedi-
ments in this way. Rather, we chose to use a material that when added to
17
-------
seawater or marine sediment produced a participate and colloidal phase. There
is a large volume of literature documenting the attraction of DDT to dissolved
organic matter and colloidal materials (Pierce, Olney, and Felback 1974; Riley
and Skirrow 1975; Weber et al. 1983a, b; Wijayaratne and Means 1984). We
evaluated two such materials, sucrose and glycine, and selected the latter,
because sucrose was found to decay rapidly in our test system. We also
selected glycine because it is one of the principal amino acids found in
seawater (Riley and Skirrow 1975).
2.3.3.1 Glycine and Nonglycine Carrier Experiments
The general dispersal technique for adding DOT to the aqueous test media
using glycine carrier was to dissolve an appropriate quantity of labeled DDT
in hexane and add an appropriate amount of this mixture to the test container.
We then coated the inside of the test container with this mixture by gently
rotating it until the hexane evaporated. This coating of DDT was then
redissolved in 750 ml of seawater or into carrier and seawater mixtures of
0.05 and 5% glycine. Dissolution of DDT was measured in each treatments.
2.3.4 Time to Equilibrium Experiments
The objective of the equilibration study was to determine when the con-
centrations of DDT in interstitial and overlying water reached steady state.
After equilibrium was achieved, the sediment and water column bioassay began.
By ensuring that the test systems were at apparent equilibrium before toxicity
testing began, the exact concentration to which the test organisms were
exposed was known and not variable. Experimental conditions (temperature,
salinity, pH, photoperiod) were maintained as specified in Section 2.2.2.
2.3.4.1 Water Column Experiments
To determine if DDT equilibrium had been achieved in the water column
experiments, samples of water were removed at specified times and analyzed
for DDT concentration using the liquid scintillation counter. The operation
of this device is explained in Section 2.2.4. Three water column equilibra-
tion experiments were run. In the first experiment, maximum initial concen-
tration of DDT ranged from 0.5 to 50 /ig/L with final concentrations of 0.25 to
10 /ig/L; in the second experiment, concentrations ranged from maximum values
18
-------
of 1.8 to 11.5 pg/L to final concentrations of 0.5 /*g/L to 3.8 /*g/L. In the
third experiment, the maximum initial concentration of DDT ranged from 2.5 to
12.0 pg/L, to final concentrations of 0.8 to 3.0 pg/L. Equilibrium was
assumed when the concentration curve reached an apparent asymptote and when
successive measurements of the column water varied by less than 10%.
2.3.4.2 Sediment Experiments
Concentrations of radiolabeled DDT in interstitial water were measured
regularly to determine when equilibrium was reached. The method for sampling
interstitial water is described in Section 2.3.2. Equilibrium was assumed
when the concentration curve reached an apparent asymptote and when three
consecutive measurements of interstitial water concentrations differed by
less than 10%.
2.4 WATER COLUMN BIOASSAYS
2.4.1 Acute Toxicity Water Column Bioassays
Water column toxicity tests using solutions of DDT with or without gly-
cine were performed using the same experimental techniques as described above.
The concentration of DDT redissolved from the containers' sides was checked
regularly to determine when the concentration within the test containers
reached equilibrium. Amphipods were exposed to six concentrations of DDT
plus a control over 4 and/or 10 days. Three sets of bioassays were performed.
The first set of bioassays was a 10-day acute toxicity test. In this bio-
assay, the DDT concentrations ranged from <1 to 20 pg/L, the glycine carrier
was not added to the beakers. In the second 10-day acute toxicity test, toxi-
city was determined on days 4 and 10; the DDT concentration ranged from 1 to
8 pg/L, and again glycine carrier was not added to the beakers. In the third
set of bioassays, a 4-day acute toxicity test was conducted with glycine (0.05
and 5%) and without glycine added to the beakers. The DDT concentration
ranged from 0.2 to 3 pg/L. These bioassays were conducted to determine if
the LCcQ concentrations for DDT with the glycine carrier differed from the
LCjjg for the DDT .without the glycine carrier. Concentrations of DDT were
determined daily during the experiment.
19
-------
2.5 SEDIMENT BIOASSAYS
2.5.1 Radio!abelinq Sediments with DDT
Labeling sediments with DDT was accomplished by dissolving 5.8089 g of
unlabeled and 0.1817 g of labeled DDT into 100 ml of methylene chloride. A
target concentration of DDT was established for each of the sediment organic
carbon/toxicant concentration levels. The volume of methylene chloride/DDT
mixture needed to produce a given concentration within the sediments was then
determined using the following formula:
V = (Cs * Sm)/ m (2)
where V = volume of stock solution required in mL
Cs = desired sediment concentration in mg/kg
S_ = mass of sediment to be labeled in kg
HI
m = DDT concentration in methylene chloride stock solution
concentration in pg/mL.
The desired sediment-bound DDT concentration of contaminant was chosen
based on the equilibrium partitioning relationship of dissolved to sediment-
bound DDT and the observed water column LC5Q concentrations. This relation-
ship from Staples et al. (1986) is as follows:
Cpw * Cs I (Koc * S * °'c->
or
(o.c.) *C (3)
where C$ a contaminant concentration on sediments
KQC • organic carbon partitioning coefficient (4 X 106)
S = suspended solids concentration (assumed 1 in sediments)
o.c. = organic carbon concentration in sediments (mg/g dry)
C = concentration of contaminant in interstitial water.
20
-------
As in the water column bioassays, the mixture was swirled in the con-
tainers until the methylene chloride evaporated. Seawater was then added to
redissolve the DDT attached to the sides of the container. Glycine at 0.05%
was used as a carrier. The appropriate amount of sediment prepared for the
correct organic carbon levels was added to the seawater/glycine/DDT. These
sediments were mixed in beakers at a 4 to 1 sediment-to-water slurry ratio
and shaken at 120 rpm. Mixing continued until an even distribution was
achieved, which was approximately 7 days.
One hundred and fifty grams of labeled sediments were distributed into
each of the bioassay containers, producing a sediment depth of approximately
2 cm in each container. These bioassay containers were then placed in a con-
trolled temperature bath and supplied with slight aeration through glass
pipette tubes. The sediments were allowed to settle, and DDT measurements
were made in interstitial and overlying waters until equilibrium was achieved
(Section 2.3.4.2).
2.5.2 Acute Toxicity Sediment Experiments
The amphipods were exposed to overlying water and sediments after the
contaminant concentrations had reached steady state in the interstitial and
overlying water. To ensure that any observed toxicity was due to sediment-
associated contaminant concentrations and not the overlying water, one set of
amphipods were exposed to the overlying water for 96 h prior to releasing a
second set into the sediment. The overlying waters in 28 test containers
were sampled and analyzed for DDT within each sediment-organic carbon/
toxicant concentration. These experiments were performed until at least 80%
survival was observed in overlying water samples from each organic carbon
type and DDT concentration. After these water column tests were completed,
the organisms were released into the sediments for 10-day exposure periods.
The sediment bioassays tested seven sediment organic carbon mixtures,
each sediment organic carbon mixture having three replicate samples of six
interstitial water concentrations of DDT and controls. The organic carbon
mixtures produced by mixing sediments with different sources of organic
21
-------
carbon provided three pairs of sediments with similar organic carbon concen-
trations. The desired DOT concentrations in interstitial water were control,
0.3, 0.5, 0.8, 1.0, 2.0, and 5.0 pg/L. The last quantity is near DDT seawater
solubility.
2.5.3 Sediment DDT Concentration
After bioassays were completed, sediments from which interstitial water
was removed were analyzed for sediment-bound DDT using the following method:
1. Sediment plugs were placed in pre-weighed containers.
2. NaSO. was added to the sediment in the container to act as an
absorbent.
3. The sediment was extracted with the addition of 5 ml methylene
chloride.
4. The extraction process was repeated three times.
5. Additional methylene chloride was added to the volume of supernatant
collected to make a total volume of 50 ml.
6. One-mL aliquots were transferred to liquid scintillation vials,
amended with 20-mL scintillation cocktail, and analyzed on the liquid
scintillation detector.
These concentrations were used to test the research hypothesis (Sec-
tion 2.6.1).
2.6 TEST OF RESEARCH HYPOTHESES
2.6.1 Toxicity is Determined by the Interstitial Concentration of DDT
The first research hypothesis is that acute toxicity is a function of
the interstitial water concentration of DDT and not the sediment DDT concen-
tration and organic carbon. The hypothesized response model suggests a fac-
torial design between interstitial water and the sediment concentrations of
nonpolar organic contaminant and dissolved and particulate organic carbon.
Accordingly, data were analyzed using a logistic analytical approach. This
approach compares the parallelism of lines of survival for different sediment
22
-------
organic carbon concentrations. The factorial approach was designed to detect
relationships between survival and participate organic carbon, interstitial
water, and sediment concentrations of toxicants assuming a general linear
response model.
A relationship between particulate organic carbon (o.c), contaminant
concentration in interstitial water (C ) and sediment contaminant concen-
trations (C$) was proposed by Staples et al. (1986) and is represented in
Equation (3) (Section 2.5.1). This relationship suggests specific forms of
the response model.
The survival data from the factorial experiments were analyzed using
logistic response models. Let p. (i = l,...n; where n is the number of treat-
ments) be the observed mortality rate among the 100 test organisms exposed to
a treatment combination. The three explanatory variables measured in the
experiment, o.c., C , and C§ suggest the natural parameterization.
ln r^-pT = 4) * W * ?2(CoJ + Aj(«-c-) «)
Alternatively, taking into account Relationship (3) for C the response
Equation (4) can be rewritten as:
pw} *
Equations (4) and (5) provide equivalent methods for analyzing the same mor-
tality data under the assumptions of Equation (3).
Multiple linear regression analyses were performed assuming Equation (4)
using the measured values of o.c., Cnuil and C. and the results compared with
pw s
a multiple regression based on Equation (5) using the measured values of o.c.
and Cpw.
23
-------
2.6.2 Equilibrium Partitioning Model Accurately Predicts
the Concentration of DDT in Interstitial Water
This research hypothesis was evaluated by comparing the predicted concen-
trations of DDT in interstitial water from Equation (3) with the measured
concentration of DDT in the interstitial water, sediments, and the organic
carbon. These results were then analyzed by linear regression to see if there
was a significant correlation.
2.6.3 Organic Carbon Source Influences Toxicity
Randomly selected replicate samples at each sediment contaminant concen-
tration were evaluated to provide three separate determinations of the LC5Q
concentration. This evaluation estimates the degree of variation around the
LC50 values. The LC50 values were determined using Spearman-Karber estimators
(Finney 1971). Sediments with equivalent mixtures of organic carbon obtained
by mixing sediments from either Sequim Bay or Station G-780 with lower organic
carbon sediments from West Beach were compared by Student t tests corrected
for multiple comparisons. The null hypothesis was: there are no significant
differences in LCcn values for the different locations.
24
-------
3.0 RESULTS
3.1 EVALUATION AND DEVELOPMENT OF BIOASSAY PROTOCOLS
3.1.1 DDT Transformation
The specific results of the DDT transformation tests are summarized in
Tables 3.1 and 3.2. The following discussion presents our observations
concerning DDT transformation, loss from water and sediment, and the extrac-
tion efficiency of the methodology.
Water Tests
• DDT is rapidly lost from solution when contained in polystyrene con-
tainers; therefore, this type of container is inappropriate for the
experiments. All additional tests were performed in glass
containers.
• DDT conversion to DDE was caused by the production of DDE through
DDT transformation within the gas chromatograph. Transformation of
DDT to DDE did not occur in the test containers during our tests.
Sediment Tests
• DOT mixed with ethanol carrier adsorbed to sediments and was recover-
able with our extraction procedure. We recovered 46 and 63% of the
spiked material (X = 54.5; S.D. = 12.0; n = 2).
• DDT mixed with sediments without ethanol carrier appears to be bound
to sediments and is as available to our extraction procedure as the
spiked carrier tests. Recovery ranged from 47.5 to 85.4% of the
spiked material (X = 67.6; S.D. = 19.1; n = 3).
• DDT mixed in sediments with an ethanol carrier results in only a
slight transformation. In comparable experiments without an ethanol
carrier, no transformation to DDE was detected. Measurements indi-
cate that the DOT-DDE transformation observed within the gas chroma-
tograph was entirely due to cleanliness, pressure, and temperature
effects within the machine.
25
-------
TABLE 3.1. Transformation and Loss of DDT in Seawater
Day
1
4
5
6
7
10
TABLE 3.2.
Experiment
Number
2(b)
3
4
5(d)
6
(ng/mL)
DDE & DDT
5633.0
13.9
13.7
11.6
64.0
2.2
Transformation and
(see Table 2.3 for
Actual (a)
Concentration
328
210
293
184
4.5
402
% Initial % DDE
69.71 4.3
0.17 16.6
0.17 16.8
0.14 31.9
0.79 11.6
0.03 13.6
Loss of DDT in Sediments
experimental conditions)
Expected
Concentration
s DDE (/*g/g)
4 523
5 457
ND^ 416
ND 387
ND 3.8
ND 471
(ng/mL)
% of
Expected
63
46
70
47.5
125
85.4
(a) Concentration = DDT + DDE.
(b) Ethanol carrier present.
(c) ND - not detected.
(d) Recovery standard, water only.
• Consequently, transformation does not occur during the period of
these experiments so that the toxicity expressed by the organisms is
directly related to 4,4' DDT.
3.1.2 Interstitial Water Sampling Methods
We found that 86 to 92% of the DDT was lost when the water was collected
in plastic syringes, while glass syringes under similar exposure conditions
removed 30%. Glass wool plugs placed in the glass syringe removed an addi-
tional 51%, while adding a 0.45-/»m filter to the syringe removed another 8%
26
-------
(Table 3.3). These results indicated that dissolved DDT concentrations,
measured after sampling with plastic syringes or glass syringes with glass
wool plugs and in-line filters, could be underestimated by as much as 92%.
The loss was only 14.5% for centrifugation. These results demonstrated that
plastics and filters should not be used during subsampling; the final method
chosen for sampling interstitial water and minimizing loss of DDT is described
in Section 2.3.2.
3.1.3 Sediment and Water Labeling Procedures
The non glycine carrier labeling resulted in coefficients of variation
of 99%, while under the same conditions labeling with the glycine carrier
resulted in coefficients of variation ranging from 22 to 24%. Thus, using a
glycine carrier provided a 4-fold improvement in the evenness of labeling.
3.1.4 Time to Equilibrium Experiments
3.1.4.1 Water Column Experiments
The concentration of DDT in the water column fluctuated for approximately
10 days before stabilizing. The concentrations in all test containers
decreased after the amphipods were added (Figure 3. la, b, c.). Bottles with
higher initial concentrations of DDT took longer to reach a steady-state or
an equilibrium point.
TABLE 3.3. DDT Recovery From Water Using Different Collection
Techniques
Water Concentration
_ Collection Technique
Initial water concentration 4.00
After glass syringe exposure 2.80
After glass syringe and glass 0.76
wool exposure
After glass syringe, glass wool, 0.44
and 0.45-^m filter exposure
After plastic syringe exposure 0.32
Centrifugation 3.46
27
-------
o
a
a>
3.
Concentrations (ug/L)
1
15
a
a
o
12
10
8
6
4
2
0
Maximum
Concentrations
15
Maximum c
Concentrations l^g/D
A 12
05
• 2.5
DO
11
13 15
Days
FIGURE 3.1. Rate of Equilibrium Development of DDT Concentration in Three
Independent Water Column Experiments (Vertical line indicates
the day amphipods were added.)
28
-------
Variation in concentration between beakers with the same initial DDT
concentration was larger than expected. The general trend in the data was a
decrease in variation through time (Figure 3.2). However, this was not the
case at the highest concentrations of DDT in water, which may be related to
this concentration exceeding the aqueous solubility of DDT. It was also not
true for the lowest concentration of DDT, which showed decreasing variation
for the first few days and increasing variation after that period. Because
the observed variation averaged more than 30%, sampling only a subset of the
exposure beakers would not have given precise estimates of the exposure
levels.
3.1.4.2 Sediment Experiments
Two preliminary tests to evaluate changes in DDT concentrations in inter-
stitial and overlying waters at two concentrations of the glycine carrier
50
30
20
10
Maximum
Concentrations l/jg/L)
D50
• 12
06
A3
O 1.5
A 0.5
L i L i I i I i I
1 2345 6 7 8 9 10 11 12 13 14 15 16
Days
FIGURE 3.2. Percent Coefficient of Variation of DDT Concentration in
Water Column Experiment 1
29
-------
(5 and 0.05%) indicated that overlying water concentrations reached equili-
brium and minimum intersample variability in about 10 days (Figure 3.3a, b).
The final overlying water concentration was approximately one half of the
I
o
«3
I
I
o
CJ
140
120
100
80
60
40
20
. D Over
— O Inter
8 12
January 1987
16
20
10
20
30
December 1986
40
50
FIGURE 3.3. DDT Equilibration in Interstitial Water (Inter) and Overlying
Water (Over) in a) 5% Glycine and b) 0.05% Glycine Solutions.
Confidence intervals (95%) are also shown.
30
-------
final concentration in the interstitial water. After 10 days, the concentra-
tions of DDT in overlying water varied by 2 to 5%. However, the interstitial
water concentrations did not reach equilibrium until at least 1 month had
passed. Intersample variation in the interstitial water concentration collec-
ted from beakers containing the same target concentrations ranged from less
than 10% to more than 40% at 20 days. The instability in'the DOT concentra-
tions in the interstitial waters suggested that the time to reach equilibrium
would be greater than 1 month.
For the third experiment, with glycine carrier added at 0.05% (Fig-
ure 3.4a, b), the concentrations of DDT in interstitial water from all test
beakers fluctuated greatly over a minimum period of 40 days. After 40 days,
the interstitial water concentrations of DDT stabilized in the higher DDT
and/or higher sediment organic carbon beakers. The concentrations in the
lower DDT and lower organic carbon concentration beakers took longer to sta-
bilize, approximately 60 to 75 days. In the lowest organic carbon and lowest
DDT concentration beakers, nearly 120 days were required for the interstitial
water concentration to stabilize.
i
3.2 WATER COLUMN BIOASSAYS
In the first 10-day acute water column toxicity tests (without glycine
carrier), survival was less than 25% at all DDT concentrations above 5 pq/L
(Figure 3.5a). In the second 10-day acute toxicity test (without glycine
carrier), survival was determined at both 4 and 10 days (Figures 3.5b, c). At
4 days, 80% survival (Figure 3.5a, b) was observed at approximately 4 ftg/l.
At 10 days, all concentrations above 2 pg/L showed less than 10% survival
(Figure 3.5c).
To determine if the addition of the glycine carrier would affect the
toxicity of the DDT, 4-day water column toxicity tests with and without the
addition of glycine were conducted. The results are presented in Figure 3.6a,
b. The concentration causing 50% mortality (LCg0) was 1.2 pg/L in the
noncarrier experiments and 1.3 pg/L in the glycine carrier experiments
(Figure 3.6a, b).
31
-------
16
14
^2
Organic Carbon (%)
O 0.12
D 0.25
• 0.58
I
c
I
0>
i
O
O
Organic Carbon (%)
O0.60
• 0.80
D1.75
• 1.92
FIGURE 3.4.
DDT Equilibration in Interstitial Water with Different
Organic Carbon Content of Sediment a) Three Lowest and
b) Four Highest Percentages
32
-------
***.
c
18
8 '"
•S 14
1 12
B 10
i 8
i 6
4
2
0
S6
&
-CD O
-O
- ^^ °
- o o
- 0
-
-
CL
:, , , , ,0, , , , , & ?%, , ,
02 4 6 8 10 12 14* 16
DDT Concentration (fjg/L)
.Ji O «D O
3 16
o 14
1 12
O
s 10
.1 8
i 6
(0
4
2
o
: °
- o o
i ®
w
-
-
—
M
-
-
~ i i ii ii i
0248
nnY f* T.^ • LI? f u««/i i
OUT concentration (fig/u
18?
8 16F
*. 14
-------
3
•6
t;
o
s
o
.2
3
20 r
6- a
18T
16
14
12
10
8
6
4
2
0
> O
—
-
_
o
: ° o
O
O
o
- o
1 1 1 1 1 1 1 1 1 1 1 1 f\
0.8 1.2 1.6 2.0
DDT Concentration (pg/L)
2.4
—
18
-------
carbon concentration are summarized in Table 3.4. The observed mortality
data for all DDT exposures and all sediments and sediment mixtures are shown
in Figures 3.7, 3.8, and 3.9. The interstitial water LCen concentrations
were within a factor of 4, irrespective of sediment source and organic carbon
concentrations.
Normalizing the sediment LCrn values to the total quantity of organic
carbon in the sediments did not decrease, but rather increased, the
coefficient of variation between sediment types as compared with the LCcg
values based on interstitial water. Survival of control organisms was low in
some bioassays (e.g., Figure 3.7a) and would require according to standard
TABLE 3.4. Survival of Amphipods in Various Sediment Types; LCso
Values Are Determined Based on Interstitial Water (I.W.)
Concentrations and on I.W. Concentration Normalized to
the Sediment-Bound Organic Carbon Concentration
Sediment Type
Water
Water + glycine carrier
West Beach Sediment
West Beach/Sequim Bay
Sequim Bay/West Beach
West Beach/Point Pulley
Point Pulley/West Beach
Point Pulley
Sequim Bay
Mean
SD
N
%CV
Organic Carbon
Sediment
Concentration
(mg/q dry)
na
na
0.125
0.25
0.60
0.58
0.80
1.75
1.92
I.W. Sediment
LC50 LC50
(ftq/L) (jtq/q carbon)
1.20 na
1.30 na
0.69 5.52
3.13 12.51
3.27 5.45
2.67 4.61
2.55 3.18
1.96 1.12
4.28 2.23
2.34 4.95
1.15 3.72
7 7
49 75
na = not applicable
35
-------
s
•S
0
2
"1
w
S
•5
o
§
'g
3
2U
18
16
14
12<
a
-
I 0
- 0
- 00
4 "
£ 00
To o
6ir o
^9" o
4
2
n
—
- 0
- , , , ,0 , ,^P,o , , , ^
0 24 6" 8 10 12
DDT Concentration (fjg/L)
20,
,8,
16
14
12
10
8
4
2
0
^
-
o
- o oo o
00 0 °
o
I o
0
o
1 1 1 1 lr\ 1 1 1 I I 1 1
2 4 ~ 6 o, 10 12
DDT Concentration
FIGURE 3.7. Survival of Rhepoxynius in a) Point Pulley Sediment and
b) Sequim Bay Sediment
36
-------
3
8 14
1 12
c 10
I 8
1 6
4
2
rt
^°
- 0
-
-
-
_
-
_
~" i i
°0 2
„,
16
8 14
•*;
12
*•
0 10
s
9 8
•5
i 6
W
4
2
0
E
--
^
- O
- O
- O
_
-
—
_
1
0
FIGURE 3.8. Survival
Organic Carbon (0.6 mg/g dry) a
O
O
O
0
0 0
0
ii i V T i i^vft i i 1 i i |rHL%
4 6 8 "10 12 14 16 U
DDT Concentration (fjg/L)
Organic Carbon (0.58 mg/g dry) b
O
1 f+\. 1 ^"^^ f^ 1 1 ^A— /"\/^*^ ' * x^^
24 6 8 10 12
DDT Concentration (jjg/U
of Rhepoxym'us in a) Sequim Bay anc
Sediment Mixture and b) West Beach and Point Pulley
Sediment Mixture
37
-------
3
•5
5
o
&
.i
1
£\1
18
16
14 (
12
10
8
f
6
4
2
n
- Organic Carbon (0.80 mg/g dry) a
_
5
- o
: o
-
>-
-
- o
• i ^°i 9 9 rc.hS i K_K_
FIGURE 3.9.
a
"o
**
0
Q
8
•5
3
O
e
o
3
iva
0
*
12
10
8
6
4
2
0
\
_
_
-
—
_
-
™
—
_
-
0
ieS
14
12
10
8
6
4
2
0
-
_
_
-
-
-
_
_
—
0
1 of
2 4 ^ ^B iflr^iS
DDT Concentration {/ug/D
Organic Carbon (0.12 mg/g dry) b
O
0
O
o
Q O O O
246
DDT Concentration (/jg/L)
Organic Carbon (0.25 mg/g dry) c
O
O
0 0
o
2 4 6 8 10 12 14 16
DDT Concentration (jig/U
Rhepoxym'us in a) Point Pulley/Wes
b) West Beach Sediment, and c) West Beach/Sequim Bay Sediments
38
-------
protocols that the experiment be redone. The standard acceptance level is
mortality less than 20% in the control. However, given time and resource
constraints, redoing the sediment bioassay experiment was impossible. The
LCeg values calculated by Spearman-Karber would not be affected by low control
survival, and the LCgQ values (Table 3.4) show little variability.
3.4 TEST OF RESEARCH HYPOTHESES
3.4.1 Toxicity is Determined by the Interstitial
Concentration of DDT
The correlation of the dependent variable, survival, with each of the
independent variables showed an insignificant correlation to particulate
organic carbon (o.c.), and significant (p<0.05) correlations with interstitial
water DDT concentration (C ), and sediment DDT concentration (Cs) (r = 0.118,
0.478, and 0.308, respectively; n = 75). These correlations indicate that a
significant relationship exists between the variable survival with the vari-
ables interstitial water DDT and sediment DDT concentrations, but with a large
degree of variation. The correlation of sediment DDT concentration with the
variable (sediment organic carbon concentration * interstitial water DDT
concentration) as suggested by Relationship (3) in Section 2.5.1 was also
found to be significant (r = 0.526; n = 75; a = 0.05). Thus, a positive
linear trend was found; however, the large variability implies a relatively
weak relationship. The correlation of survival with this same variable, as
expected, was also significant (r = 0.363; n - 75; a = 0.05), but the low
correlation again implies a weak relationship.
Multiple linear regressions were compared using the hypothesized rela-
tionships (4) and (5), which are described in Section 2.6.2 as competing full
models and their associated reduced models created by successively removing
terms from the model. The assumptions of normality and constant variance for
the dependent variable, survival, were checked visually using a histogram and
residual plot analyses (results not shown). One highly influential data point
and several points associated with only a single survivor were removed from
the analysis to ensure unbiased regression and normality.
39
-------
Tests of the full model depicted in Relationship (4) were significant;
however, the variables of particulate organic carbon and sediment DDT concen-
tration did not significantly add to the relationship. Therefore, the best
model was the reduced form incorporating only the dissolved interstitial water
concentration of DDT. The resultant regression equation is
= -0.764 + 0.173(Cpw) = survival (6)
2
which is significant at p<0.01 (R = 0.229). The competitor model depicted
in Equation (5), as expected from the above results, was only significant
because of the interstitial water concentration of DDT. Therefore, both
Models (4) and (5) were reduced to Equation (6), and the interstitial concen-
tration of the toxicant was the only measured variable of significance.
3.4.2 Equilibrium Partitioning Model Accurately Predicts
the Concentration of DDT in Interstitial Water
The concentration of DDT in the interstitial water of the various sedi-
ments was compared with the predicted concentration based on the equilibrium
partitioning model, the total organic carbon content of the sediment, and the
KQC value of 3.3 x 10 , which is the average KQC value reported in Pavlou et
al. (1987). Predicted and measured concentrations were nearly identical with
predicted concentrations exceeding measured values by only a factor of 0.15
(Figure 3.10). The linear regression between these values provides an r value
of 0.563, which at 140 degrees of freedom is highly significant at p<0.01.
However, this low r value implies a relatively weak relationship and high
variability (as shown in Figure 3.10). The range of reported KOC values is
1.5 x 104 to 4 x 106 (Pavlou et al. 1987). When the KQC value of 4 x 106
(Staples et al. 1986) is used to predict the DDT concentration, the measured
and predicted concentration differ by a factor of approximately 10.
3.4.3 Organic Carbon Source Influences Toxicity
The results of sediment bioassays indicate that the source of organic
carbon influences the toxicity of the contaminant. Sequim Bay sediment, or
mixtures that contained Sequim Bay sediment, exhibited greater survival at a
40
-------
Measured DDT Concentrations (^g/U
FIGURE 3.10. Predicted versus Measured Interstitial DDT Water Concentrations
Predicted = 1.084 (measured) + 1.057, r = 0.563, df = 140.
given interstitial water concentration of DDT than other sediment sources or
mixtures with similar concentrations of organic carbon (Figure 3.11). Sequim
Bay sediments alone exhibited nearly 50% survival at concentrations in excess
of 12 ng/L. This concentration was nearly 6 times greater than the LC5Q value
of Point Pulley sediment (1.96 pg/L). Three pairs of LC50 values, derived as
described in Section 2.6.1, for similar contaminant concentrations and organic
carbon concentrations were compared using Student t tests. The results of
these comparisons demonstrated that test interstitial water from Sequim Bay
sediment required significantly higher concentrations of DDT before test
organisms showed the same toxicity observed in other sediments and mixtures
(Table 3.5).
41
-------
o
CM
•5
O
s
>
1
CO
20
18<
16
14
121
10
q
8
6
c
4
2
0
2. a
>_ ° o Point Pulley
0 Sequim Bay
D
- Qo o
HO 0
o
— o oo o
> a a o
— 00 O
~ O Q O
— a o
>• a
— o
o
— a
Lj_l_i_la, I°,DI , I ,_i , I , I ,
4 6 8 10 12 14 16 18
DOT Concentration (pg/L)
p- a b
18^ D
16
•5 14
5 12
0
2 10
3
2
co1 6
4
2
0
»- & Sequim Bay/West Beach
I ° o West Beach/ Point Pulley
_
-
- o a
• o
a
• 0
a
a
a a
- a
68 10 12 14 16 18
DOT Concentration (/jg/L)
FIGURE 3.11. Toxicity Between a) Point Pulley/Sequim Bay Sediments
and b) Sequim Bay/West Beach Sediment with West Beach/
Point Pulley Sediment Mixture
42
-------
TABLE 3.5. Variation in Interstitial Water LCso Concentrations Calculated for
Different Sediment Mixtures and Results of Statistical Tests of
Significance for Influence of Organic Carbon Source
S.D.
Sediment Type
West Beach
replicate 1
replicate 2
replicate 3
Sequim Bay
replicate 1
replicate 2
replicate 3
Sequim/ West Mix
replicate 1
replicate 2
replicate 3
West/Sequim Mix
replicate 1
replicate 2
replicate 3
Point Pulley
replicate 1
replicate 2
replicate 3
Pulley/West Mix
replicate 1
replicate 2
replicate 3
Replicate LCRn
Ufl/L) 50
0.97
1.14
0.63
3.71
4.26
4.56
4.39
4.62
3.79
1.55
2.61
1.32
2.36
2.82
1.97
1.36
1.10
1.42
Mean
(gg/D
0.91
4.18
4.27
1.83
2.38
1.29
0.21
0.35
0.35
0.56
0.35
0.14
43
-------
TABLE 3.5. (contd)
Sediment Type
West/Pulley Mix
replicate 1
replicate 2
replicate 3
Statistical Tests
T-test Statistics t4
t4
Hypothesis Testing
Replicate LCen
(ng/D 50
1.32
1.04
1.37
Mean
3.747 @ a
1.533 @ a
0.01
0.10
Ho: LCso Sequim Bay = Point Pulley LCso
Ha: Sequim Bay LCso > Point Pulley LCso
Ho: LCso Sequim/West mix = Pulley/West mix
Ha: LCso Sequim/West mix > Pulley/West mix
Ho: LCso West/Sequim mix = West/Pulley mix
1.24
S.D.
0.14
t = 6.285, therefore
Ho rejected
t = 13.718, therefore
Ho rejected
t = 1.741, therefore
Ho accepted
3.4.4 Relationship Between Participate Organic Carbon and Dissolved Organic
Carbon Concentrations in Interstitial Water
Higher concentrations of dissolved organic carbon were measured in inter-
stitial waters of sediments with higher organic carbon contents (Figure 3.12).
Also, the concentration of dissolved organic carbon in the interstitial waters
appears to be directly related to the concentration of total organic carbon
(TOC) in the sediments, although there is considerable variability. Linear
regression of the concentration of dissolved organic carbon (DOC) versus the
sediment organic carbon provides the following equation:
DOC = (2.94 x TOC) + 2.32
(r = 0.735, df = 53, p<0.01)
(7)
44
-------
14
13
I"
I11
10
8
7
o
« 8
}.
1 4
0.1 0.3 0.5 0.7 0.9 1.1 1.3 1.5
Total Organic Carbon (mg/g dry)
8
1.7
1.9
FIGURE 3.12. Total Organic Carbon (TOC) to Dissolved Organic Carbon (DOC)
DOC = 2.94(TOC) + 2.32, r = 0.735, df = 53, p<0.01.
45
-------
4.0 DISCUSSION AND CONCLUSIONS
4.1 PROTOCOL DEVELOPMENT
Considerable effort was expended to develop methods that would enable us
to accurately evaluate the EP approach. Based on that research, we recommend
that protocols used to evaluate the EP approach for other nonpolar organic
chemicals include provisions for: 1) determining the rate and toxic signifi-
cance of transformation products of test compounds, 2) sampling interstitial
waters using a glass pipette and centrifugation, 3) using a glycine carrier
in sediment labeling, 4) measuring contaminant concentrations in all experi-
mental beakers, and 5) attaining equilibrium before the toxicity tests are
initiated. The reasons for these recommendations are discussed below.
4.1.1 DDT Transformation
DDT has five principal potential metabolic transformation routes in
various organisms: DDA, Kelthane, DDE, ODD, or dicholorobenzophenone (O'Brien
1967). The principal transformation product in water and in sediments is DDE
(Wolfe et al. 1977; Gould 1966). The concern about this transformation pro-
cess is that toxicity is known to vary with each metabolic product (Sunshine
1969; Murty 1986); accordingly, if the toxicant added in our experiments were
transformed, we would not be sure which of the products produced the observed
toxicity.
Our experiments revealed that transformation of DDT to DDE can occur
during gas chromatography. This process is known and has been clearly des-
cribed in the chemical analytical literature (U.S. EPA 1986). However,
bioassay analysts are not generally aware of this potential and should
carefully analyze external standards of high purity compounds at the same time
experimental samples are analyzed.
Our experiments were also designed to determine if the rate of DDT trans-
formation was slow, indicating that the toxicity was attributable to only
DDT, or fast, indicating that the toxicity could be associated with one of
the other breakdown products. Our evaluation revealed the rate of trans-
formation of DDT to its principal metabolite, DDE, was slow under these
47
-------
experimental conditions. Because DDE was not detected in the sediments at
the end of the transformation experiments, transformation was not a factor of
concern for these experiments.
This finding is contrary to what is normally seen in the environment,
where DDE is the dominant marine degradation product in sediments. This
breakdown apparently occurs in anaerobic (reducing) environments or when
passed through a human digestive system (O'Brien 1967; Gould 1966). Although
our results indicated that we did not have to be concerned with DDT transfor-
mation, we should be aware that under certain environmental conditions, the
most likely metabolite of DDT in sediments is DDE.
4.1.2 Interstitial Water Sampling Methods
Dissolved components in overlying or interstitial waters are operation-
ally defined as those components that are contained in water, but that will
pass through filters of 0.45-/mi pore diameter. They can also be operationally
defined as the materials remaining in water after centrifugation at 760 x g
for 10 minutes. Centrifugation at this gravitational force is sufficient to
remove particles of 0.45 ;im diameter if composed of materials with a specific
gravity of 1.2 g/cm3 (Gschwend and Wu 1985).
The problem associated with these two methods is that nonpolar organic
toxicants have a high affinity for sorption to surfaces of any type (O'Brien
1967). Our results showed that the use of plastics and glass fiber filters
could result in removal of as much as 92% of the DDT from pre-filtered and
spiked seawater. This removal would result in a measured concentration
approximately an order of magnitude less than the actual concentration. This
error in the measured concentration could result in an error in estimating
the toxicity of the DDT and an incorrect evaluation of whether partitioning
is accurately represented by the EP method. Underestimation of the intersti-
tial water concentration of nonpolar organic contaminants caused by sorption
of these compounds onto glass filters or plastics could have resulted in the
differences observed between predicted and measured interstitial water concen-
trations as determined in the field (Socha and Carpenter 1987; Kadeg and
Pavlou 1987). As a result of these experiments, it appears that the practice
48
-------
of filtering overlying or interstitial water samples, although common, is an
inappropriate procedure for obtaining the dissolved fraction of organic
contaminants.
4.1.3 Evenness of Labeling in Sediments and Water
The objective of the bioassays was to relate the observed toxicity to a
specific concentration of contaminant. Uniformity of toxicant exposure of
the organism is important because it is possible that some test organisms can
avoid certain areas of high concentration. Such behavioral responses have
been observed for several types of amphipods, including R. abrom'us, which has
been observed to come out of sediments if the sediments are highly toxic
(Erdem and Meadows 1980; Swartz et al. 1984). If the test organism is able
to sense and avoid areas of high concentration, and the variation in concen-
tration within the test container is large, then using the highest concentra-
tion, or an average concentration, would not correctly represent the exposure
concentrations. It is important to ensure that the sorptive materials in the
sediments are evenly mixed throughout the sediments.
Several sediment-labeling techniques were evaluated to ensure that the
toxicant was uniformly dispersed within the organism's mobility range. Com-
parison of five replicate samples of sediment (2-3 ml) indicated that
labeling using the noncarrier procedures resulted in coefficients of variation
of 99%. Under the same conditions, the use of the glycine carrier resulted
in coefficients of variation ranging from 22 to 24% for sediment contaminant
concentrations covering three orders of magnitude. For sediment labeled using
glycine as a carrier, the coefficients of variation demonstrated a 4-fold
improvement in the evenness of labeling, which provides greater confidence in
specifying the exposure concentration.
The improved evenness of labeling using glycine, the fact that glycine
is a natural constituent of the ocean's dissolved organic carbon sink (Riley
and Skirrow 1975), and that dissolved organic carbon has been shown to bind
nonpolar organic toxicants (Carter and Suffet 1982; Hassett and Anderson 1982;
Caron, Suffet, and Belton 1985), suggest that the glycine carrier labeling
technique should be considered for all sediment bioassays using nonpolar
49
-------
organic contaminants. These results also demonstrate the importance of eval-
uating labeling techniques before conducting sediment bioassays.
4.1.4 Determining DDT in All Test Containers
Another technique commonly used in sediment bioassays to minimize costs
is to use separate beakers for chemistry and toxicity measurements. The
results shown here indicated that the DOT behaved differently in each con-
tainer and that the concentration of DDT had to be measured within each
container during the bioassays to interpret the toxicity results properly.
Thus, when conducting bioassays with hydrophobia compounds such as DDT, using
separate beakers for chemical analyses could result in misinterpretation of
the toxicity results. We feel that determining the toxicant concentration in
each of the toxicity test containers throughout the experiment is necessary.
4.1.5 Equilibrium Experiments
Ensuring that DDT contaminants dissolved in the water in the water column
test and in interstitial water and the sediment-bound forms for the sediment
test were at equilibrium was necessary for two reasons. First, the equilib-
rium partitioning theory requires that the interstitial water concentrations
be at equilibrium before the method can be evaluated. Second, we observed
the concentrations in overlying and interstitial water to range over several
orders of magnitude in the first few days after DDT was added to the test
beakers. Because of this large variation, interpretation of the results of
toxicity tests conducted during these few days would be impossible. There-
fore, before conducting either the water column or sediment toxicity
bioassays, it was necessary that the time to equilibrium in overlying and
interstitial water be determined.
Previous studies (Adams, Kimerle, and Mosher 1985) determined that the
time for nonpolar organic compounds in interstitial water to come to equili-
brium with surrounding sediments might be long (>30 days). However, experi-
ments were not specifically conducted with DDT to determine the rate of this
equilibration. It was found during one experiment that interstitial water
concentrations reached equilibrium with sediment concentrations more rapidly
in the presence of higher concentrations of toxicant and organic carbon.
50
-------
When the toxicant or organic carbon concentrations decreased, the time to
equilibrium increased. The times required for equilibrium to be established
ranged between 30 and 120 days.
Another factor that may affect the rate of equilibrium is the form of
the toxicant that is found in interstitial water. Many investigators have
demonstrated that nonpolar organic toxicants bind to dissolved organic matter
(Carter and Suffett 1982; Hassett and Anderson 1982; Caron and Belton 1985).
Researchers have also seen a relationship between dissolved organic matter
and resultant toxicity. However, our measurements of DDT in interstitial
water included both freely dissolved DDT and DDT bound to dissolved organic
carbon. Unfortunately, we do not know which equilibrium determines the rate
of equilibration, the equilibrium between sediment-bound DDT and free DDT or
that between sediment-bound DDT and DOC-bound DDT.
4.2 FINDINGS ON TOXICITY AND EQUILIBRIUM PARTITIONING THEORY
Three research hypotheses and an additional issue were investigated
during these studies, as described below.
4.2.1 Research Hypothesis 1
• The toxicant in the interstitial water is the primary determinant of
acute toxicity.
The results of the bioassay tests showed that the concentration of the
nonpolar organic contaminant in interstitial water was closely correlated to
the estimation of toxicity for the organisms. The regression relationship
for alternative models showed that the interstitial water DDT concentration
was the only significant predictor variable. Sediment concentrations of DDT
did not improve the significance of the regression. Even with the noted
variation of Sequim Bay sediments (Section 4.2.3), interstitial water DDT
concentrations resulted in LC50 values that varied within a range of 0.69 -
4.28 itg/l with a mean of 2.34 /»g/L. This variation in interstitial water LC5Q
concentrations was probably due to complexation of DDT with dissolved organic
carbon, which was also found in the interstitial waters.
51
-------
4.2.2 Research Hypothesis 2
• The organic carbon partitioning model accurately predicts the concen-
tration of toxicant in interstitial water.
It was found that the concentration of DDT in the interstitial water
could be predicted based on the total organic carbon and the total nonpolar
contaminant concentration in the sediment. The predicted concentration
exceeds the measured concentrations by only a factor of 0.15 for the average
KQC value (Pavlou et al. 1987). At a maximum, the measured and predicted
concentrations differ by a factor of 10 if the KOC values of Staples et al.
1986 are used. Discrepancies between predicted and measured concentrations in
interstitial water have been reported in the literature (Briggs 1981; Carter
and Suffet 1982; Di Toro 1985; Adams, Kimerle, and Mosher 1985). The large
variation noted in the measured DDT values may be due to the quantity of dis-
solved organic matter present in interstitial water and uncertainty in K
value. In the future, it is recommended that dissolved organic carbon be
analyzed for each individual measurement of interstitial water.
4.2.3 Research Hypothesis 3
• The toxicant concentration in interstitial water results in different
LCgQ values if different sources of sediment-bound organic carbon
are used.
The source of the organic carbon in the sediment had a statistically
significant (p<0.05) effect on the toxicity of the contaminant. Sequim Bay
sediments were observed to produce less toxicity than would be expected based
solely on the measured concentration of DDT in the interstitial water.
Bioassay tests that contained Sequim Bay sediments required significantly
higher concentrations of DDT in the interstitial waters before the organisms
showed the same toxicity as that observed for other sediments and mixtures
tested.
The reason for this finding is unknown at this time. One possibility is
that the DOC concentrations or the available binding sites ["omega factor" of
Lambert (1968) and "active fraction" of Shin, Chodan, and Wolcott (1970)] of
the organic carbon are greater for sediments from Sequim Bay than from the
52
-------
other sediments tested. Another possibility is that the mortality for the
Point Pulley sediments and mixtures containing these sediments may have been
caused by other unmeasured toxicants in interstitial waters. Because the
observed toxicity based on the interstitial water concentration of DOT is
also lower in Sequim Bay sediments than in the water column tests, this latter
explanation is unlikely.
4.2.4 Additional Experimental Issue
An additional experimental issue is related to the influence of dissolved
organic carbon on interstitial water toxicity. It was determined that the
quantity of dissolved organic carbon contained within interstitial waters is
highly variable and is generally related to the total particle-bound organic
carbon fraction. It is probable that this latter fraction also influences
contaminant concentrations in the interstitial waters and account for the
large variability in measured DDT concentrations. These observations and
conclusions are consistent with data provided by Adams, Kimerle, and Mosher
(1985).
4.3 APPLICATION OF RESULTS TO SEDIMENT QUALITY CRITERIA DEVELOPMENT
These experimental results will increase our ability to predict labora-
tory concentrations of DDT that will result in acute toxicity for the amphipod
R. abronius. which dwells in the interstitial waters and sediments. But can
these laboratory results be extrapolated to answer environmental questions?
To answer this question, we must first examine the laboratory protocols to
identify possible limitations affecting these extrapolations. The following
lists those limitations.
• The tests are acute, not chronic; therefore, extrapolation to the
environment should be based only on short-term exposures to rela-
tively high levels of the chemical that might occur during a spill
event or when the organisms are new adult or juvenile recruits to
the contaminated sediment.
• Vie assumed that equilibrium between the dissolved interstitial water
concentrations of the contaminant and the contaminants attached to
53
-------
sediments occurs naturally in the environment and that the equili-
brium establishes itself rapidly enough that organisms will only be
exposed to equilibrated levels of the contaminants. Therefore, we
did not include a potential for nonequilibrium situations that occur
through major changes in the environment such as through human-
induced activities; i.e., dredging sediments.
• We used organisms that live freely within the sediments and that do
not sequester themselves from exposure by forming burrows or other
dwellings.
• The tests were performed on relatively short time scales with organ-
isms that do not ingest sediments. This forces the exposure route
to only be through contact with interstitial waters.
• The tests that were performed were static bioassays that force the
water overlying the sediments to be somewhat contaminated. This
means that organisms could not temporarily escape contamination as
they may be able to do in some lentic environments.
\
4.4 FUTURE RESEARCH NEEDS
Extrapolation of laboratory results to field situations should always be
pursued carefully. The experiments described in this report demonstrated
that the equilibrium partitioning theory adequately estimated interstitial
water toxicities under closely controlled experimental conditions. A new set
of experiments should be pursued that could demonstrate whether the effect of
the discrepancies outlined in the previous paragraphs are either minor or
predictable. One such experiment would be a disturbance experiment where the
equilibrium of interstitial water concentrations of the contaminant is dis-
rupted. This experiment would attempt to model the effect of bioturbation
and disturbance that is commonly encountered in the environment. Another
experiment would be to examine the differences in toxicity that would result
from exposing an organism that would receive doses of the contaminant not
only from the interstitial water route, but also through ingesting the sedi-
ments. Chronic toxicity experiments need to be included to determine if the
54
-------
mechanisms of acute toxicity of interstitial waters are the same as those
that affect long-term survival, reproduction, and recruitment of sediment
dwellers.
In addition to the above potential problems in extrapolating our results
to the environment, there are other questions that have not been addressed.
Can these results be extended to a wider range of toxic chemicals? We believe
that the results are transferable to other organic nonpolar chemicals that
have a high affinity for organic carbon and that will not be rapidly trans-
formed. However, experimental confirmation is recommended. Other experiments
should also examine the toxicity of chemicals that are less strongly bound to
organic carbon. The reasons for the influence of organic carbon source on
toxicity should also be examined.
55
-------
5.0 REFERENCES
Adams, W. J., R. A. Kimerle, and R. G. Mosher. 1985. "Aquatic Safety
Assessment of Chemicals Sorbed to Sediments." In Aquatic Toxicology and
Hazard Assessment; Seventh Symposium, eds. R.D. Cardwell, R. Purdy, and R.
C. Banner, ASTM STP 854, pp. 429-453. American Society for Testing and
Materials, Philadelphia, Pennsylvania.
Barnard, J. L. 1960. "The Amphipod Family Phoxocephalidae in the Eastern
Pacific Ocean, with Analyses of Other Species and Notes for a Revision of
the Family." Allan Hancock Pacific Expeditions 18(3), 375 pp.
Barnard, J. L. 1979. "Revision of the American Species of the Marine
Amphipod Genus Paraphoxus (Gammaridea: Phoxocephalidae)." Proc. Biol. Soc.
Wash. 92(2):368-379.
Barnard, J. L. and C. M. Barnard. 1982. "The Genus Rhepoxynius (Crustacea:
amphipoda: Phoxocephalidae) in American Seas". Smithsonian Contributions
to Zoology No. 357, Smithsonian Institution Press, Washinton, O.C.
Briggs, G. G. 1981. "Theoretical and Experimental Relationships Between
Soil Absorption, Ocatanol-Water Partition Coefficients, Water Solubilities,
Bioconcentration Factors, and the Parachor." J. Aqric. Food Chem.
29:1050-1059.
Caron, G., I. H. Suffet, and T. Belton. 1985. "Effect of Dissolved Organic
Carbon on the Environmental Distribution of Nonpolar Compounds."
Chemosphere 14(8):993-1000.
Carter, C. W., and I. H. Suffet. 1982. "Binding of DDT to Dissolved Humic
Materials." Environ. Sci. Techno!. 16(11):735-740.
Di Toro, M. 1985. "A Particle Interaction Model of Reversible Organic
Chemical Sorption." Chemosphere 14(10): 1503-1538.
Enquist, P. 1954. "Studies on the Soft Bottom Amphipods of the Skagerrak."
Zool. Bidrag. Fran Uppsala 28:297-492.
Erdem, C. and P. S. Meadows. 1980. "The Influence of Mercury on the
Burrowing Behavior of Corophium volutator." Marine Biology 56:233-237.
Finney, D. J. 1971 Probit Analysis. Third Edition, Cambridge University
Press. Cambridge, Massachusetts.
Gould, R. F. Ed. 1966. Organic Pesticides in the Environment. Advances in
Chemistry Series No. 60. American Chemical Society, Washington D.C. 309
pp. Second Printing 1970.
57-
-------
Gschwend, P. M., and S. Wu. 1985. "On the Constancy of Sediment-Water
Partition Coefficients of Hydrophobic Organic Pollutants." Environ. Sci.
Techno!. 19:90-96.
Hassett, J. P., and M. A. Anderson. 1982. "Effects of Dissolved Organic
Matter on Adsorption of Hydrophobic Organic Compounds by River- and Sewage-
Borne Particles. Water Res. 16:681-686.
Kadeg, R.D., and S. P. Pavlou. 1987. Reconnaissance Field Study for
Verification of Equilibrium Partitioning; Nonpolar Hydrophobic Organic
Chemicals"Prepared for the U.S. Environmental Protection Agency by
Battelle, Washington Environmental Program Office, Washington, D.C.
Knezovich, J. P., and L. L. Harrison. 1986. Transfer Coefficients of
Selected Sediment-Bound Organic Chemicals in a Model Aquatic System.
Lawrence Liver-more National Laboratory, Livermore, California.
Lambert, S. M. 1968. "Omega, a Useful Index of Soil Sorption Equilibria."
J. Agr. Food Chem. 16:340-343.
McCarthy, J. F.( and B. D. Jimenez. 1985a. "Interactions Between Polycyclic
Aromatic Hydrocarbons and Dissolved Humic Material: Binding and Dissocia-
tion." Environ. Sci. Techno!. 19:1072-1076.
McCarthy, J. P., and B. D. Jimenez. 1985b. "Reduction in Bioavailability to
Bluegills of Polycyclic Aromatic Hydrocarbons Bound to Dissolved Humic
Material." Environ. Toxicol. Chem. 4:511-521.
McCarthy, J. F.( B. D. Jimenez, and T. Barbee. 1985. "Effect of Dissolved
Humic Material on Accumulation of Polycyclic Aromatic Hydrocarbons:
Structure-Activity Relationships." Aquat. Toxicol. 7:15-24.
McQuaker, N. R. and T. Fung. 1975. "Determination of Carbonaceous Material
in Sediments by Reductive Pyrolysis and Spectrophotometry". Anal. Chem.
47(8):1435-1438.
Murty, A. S. 1986. Toxicitv of Pesticides to Fish Vols 1 and 2. CRC Press
Inc. Pensacola, Florida 142 pp.
O'Brien, R. D. 1967. Insecticides; Action and Metabolism. Academic Press,
New York. 332 pp.
Pacific Northwest Laboratory. 1986. Reconnaissance Survey of Eight Bays in
Puget Sound. Prepared for the U.S. Environmental Protection Agency, Region
X, Seattle, Washington, PNL-SA-15393A by Pacific Northwest Laboratory,
Richland, Washington.
58
-------
Pavlou, S., R. Kadeg, A. Turner, and M. March!ik. 1987. Sediment Quality
Criteria Methodology Validation Uncertainty Analysis of Sediment
Normalization Theory for Nonpolar Organic Contaminants^Prepared for the
U.S. Environmental Protection Agency by Battelle, Washington Environmental
Program Office, Washington, O.C.
Pierce, R. H., Jr., C. E. Olney, and G. T. Felbeck, Jr. 1974. "pp'-DDT
Adsorption to Suspended Particulate Matter in Sea Water." Geochim.
Cosmochim. Acta 38:1061-1073.
Poston, T. M., and L. A. Prohammer. 1986. Protocol for Sediment Toxicity
Testing of Nonpolar Organic Compounds. Prepared for the U.S. Environmental
Protection Agency by Battelle, Washington Enviromental Program Office.
Riley, J. P., and G. Skirrow, eds. 1975. Chemical Oceanography. Academic
Press, London. Vol. 1, 606 pp.
Shin, Y. 0., J. J. Chodan, and A. R. Wolcott. 1970. "Adsorption of DDT by
Soils, Soil Fractions, and Biological Materials." J. Agr. Food Chem.
18(6):1129-1133.
Socha, S. B., and R. Carpenter. 1987. "Factors Affecting Pore Water
Hydrocarbon Concentrations in Puget Sound Sediments." Geochim. Cosmochim.
Acta. 51:1273-1284.
Standard Methods. 1985. Standard Methods for the Examination of Water and
Wastewater. American Public Health Association, Washington, D.C.
Staples, C. A., K. L. Dickson, J. H. Rodgers, Jr., and F. Y. Saleh. 1986.
"A Model for Predicting the Influence of Suspended Sediments on the
Bioavailability of Neutral Organic Chemicals in the Water Compartment." In
Aguatic Toxicology and Hazard Assessment; Seventh Symposium, eds. R.D.
Cardwell, R. Purdy, and R. C. Bahner, ASTM STP 854, pp. 417-428. American
Society for Testing and Materials, Philadelphia, Pennsylvania.
Sunshine, I. 1969. Handbook of Analytical Toxicology. The Chemical Rubber
Company, Cleveland, Ohio.1081 pp.
Swartz, R. C.. W. A. DeBen, J. K. P. Jones, J. 0. Lamberson, and F. A. Cole.
1984. "Phoxocephalid Amphi pod Bioassay for Marine Sediment Toxicity." In
Aquatic Toxicology and Hazard Assessment; Seventh Symposium, eds R. D.
Cardwell, R. Purdy and R. C. Bahner, ASTM STP 854, pp. 284-306. American
Society for Testing and Materials, Philadelphia, Pennsylvania.
Swartz, R. C., D. W. Schults, G. R. Ditsworth, W. A. DeBen. 1985. "Sediment
Toxicity, Contamination, and Macrobenthic Communities Near a Large Sewage
Outfall." In Validation and Predictability of Laboratory Methods for
Assuring the Fate and Affects of Contaminants in Aguatic Systems, ed. T.P.
Boyle, ASTM STP. 865 pp.
Swartz, R.C., G. R. Ditsworth, D. W. Schults, and J. 0. Lamberson. 1986.
59
-------
"Sediment Toxicity to a Marine Infaunal Amphipod: Cadmium and its
Interaction with Sewage Sludge." Mar. Environ. Res. 18:133-153.
U.S. Environmental Protection Agency (EPA). 1986. "Method 8080. Determina-
tion of Pesticides and PCBs in Water and Soil/Sediment by Gas
Chromatography/Mass Spectrometry." In Test Methods for Evaluating Solid
Hastes. EPA-SW-846, Third Edition, September 1986.U. S. Environmental
Protection Agency, Cincinnati, Ohio.
Weber, W. J., Jr., T. C. Voice, M. Pirbazari, G. E. Hunt, and D. M. Ulanoff.
1983a. "Sorption of Hydrophobic Compounds by Sediments, Soils and Suspended
Solids - II. Sorbent Evaluation Studies." Water Res. 17:1443-1452.
Weber, W. J., Jr., T. C. Voice, M. Pirbazari, G. E. Hunt, and D. M. Ulanoff.
1983b. "Sorption of Hydrophobic Compounds by Sediments, Soils and Suspended
Solids - II." Water Res. 10:1443-1452.
Wijayaratne, R. 0. and J. C. Means. 1984. "Sorption of Polycyclic Aromatic
Hydrocarbons by Natural Estuarine Colloids." Mar. Environ. Res. 11:77-89.
Wolfe, N. L., R. G. Zepp, D. F. Paris, G. L. Baughman, and R. C. Hollis.
1977. "Methoxychlor and DDT Degradation in Water: Rates and Products."
Env. Sci. and Tech. 11(12):1077-1081.
Word, J. Q. 1978. "Changes in Indicator Organisms as Reflected by the
Infaunal Index." Presented in hearings before the Subcommittee on Water
Resources of the Committee on Public Works and Transportation, U.S. House
of Representatives, May 24-25, 1978, Washington, D.C.
Word, J. Q. 1980. Marine Environmental Studies and the Use of the Infaunal
Trophic Index. Seminars in Ecology, University of Victoria, British
Columbia.
Word, J. Q., and P. L. Striplin. 1981. "Increases in Benthic Microcrustacea
and Echinoderms on the Palo Verdes Shelf." In Annual Report; Southern
California Coastal Water Research Pro.iect. Long Beach, California.
60
------- |