evaluation of Chester River oyster mortality

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

an eva ua ion o COASTAL ZONL
CHESTER RIVER 'NFORmATION CENTER.
OYSTER MORTALITY

sw n.. ck%QvWthWv%
Pt.

ir
.kove

0.r lea

QL
430.7
.09
E75
1979
I t f

MARYLAND DFPAPTMFNT OF NATURAL RESOUCES

AN EVALUATION OF CHESTER RIVER OYSTER MORTALITY

Joseph J. Cooney, F, Douglas Martin, and W. H. Roosenberg
Chesapeake Biological Laboratory, University of Maryland
Solomons, Maryland 20688

and

David H. Freeman
,Department of Chemistry and Chesapeake Biological Laboratory
College Park, Maryland 20742

and

Charles R. Bostater, Jr.
Project Manager, Maryland Water Resources Administration
Annapolis, Maryland 21401

Funded By

Chesapeake Bay Program Grant #805976

and

Maryland Department of,Natural Resources

Lowell Bahner, Project officer
EPA Chesapeake Bay Program
Gulf Breeze Environmental Research Laboratory
Annapolis, Maryland 21401

GULF BREEZE ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
GULF BREEZE, FLORIDA 32561

DISCLAIMER

This report has been reviewed by the Chesapeake Bay Program, U.S.
Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and
policies of the U. S. Environmental Protection Agency or the Maryland
Department of Natural Resources, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

PREFACE

The Maryland Water Resources Administration received a
grant from the U.S. EPA Chesapeake Bay Program to conduct a
study concerning oyster mortality in the Chester River, a sub-
estuary of the Chesapeake Bay. The State of Maryland contracted
with the University of Maryland to perform a three-part study to
evaluate the reported oyster mortality. Due to the fact that
this was a retrospective study, the evaluation of past large
instananeous inputs of chemicals and the resulting potential
impact was not made. The results of the University of Maryland's
research follows.

The results of the three studies involving aquatic bioassays
and analysis of phthalate esters, tin and organotin compounds in
various environmental media did not show conclusive cause-and-
effect relationships concerning the oyster mortality.

The bioassay study showed no significant point source or
ambient acute aquatic toxicity to organisms tested; however,
chronic stress was indicated by growth reduction of oysters
during the study period and by occasional low levels of dis-
solved oxygen in the lower estuary. The point source bioassay
observations showed mortality of organisms as well as controls;
however, the controls and point source tests were not signifi-
cantly different.

The analysis of phthalate esters in alluvial sediments
showed a decreasing trend downstream from a point source. A
pond receiving an industrial discharge showed extremely high
concentrations of phthalate esters. The levels found in the
vicinity of the pond warrant further consideration. Levels of
Phthalate esters in the estuarine sediments are low. The tox-
icity of the levels found is unknown, especially to oysters.
A review of current literature shows no documentation of the
toxicity and dynamics of phthalates to exposed oysters. The
oyster concentration data in this report are the first published
data available for di (2-ethylhexyl) phthalate (DEHP) in the
American oyster. Although the toxicity of phthalates to the
?yster would be expected to be low, because of the economic
importance of the oyster, further laboratory bioassays are
warranted to determine acute-and chronic-effect concentration
levels.

Other modeling efforts conducted by state and Federal
workers haveestimated that the major transport pathway of

phthalate esters is in the dissolved fraction in the water
column. It is not known at this time if the water phase trans-
port of phthalates out of the system and other environmental
degradation processes are great enough to preclude further
buildup of DEHP in the estuarine sediments. Based on the levels
of DEHP found in the pond receiving an effluent, further accumu-
lation of phthalates in the estuarine sediments may be possible.
Additional modeling could address these issues if sediment
deposition and resuspension calculations were included in a
modeling framework.

Tink microorganisms resistant to tin, and microorganisms
capable of transforming inorganic tin to organotin(s). were
present at the sites sampled. However, the analysis of tin
and organotin compounds in Chester River media did not result
in any data to support or preclude the possibility of oyster
mortality due to organotin compounds.

In summary, no significant mortality of oysters was
observed during the course of this study. However, there are
indications of chronic stress in the estuarine system based on
the results of this study. Should mortality be observed again
at any time in the future, it is recommended that oyster samples
should be taken immediately, stored, and later analyzed for
suspected xenobiotics. At the same time, water quality variables
such as dissolved oxygen, salinity and other environmental
variables should be observed as soon as possible after the
mortality is reported. It should be noted that the greater the
time period between when mortality of organisms is reported, and
the analytical observations are made, the more difficult it will
be to evaluate possible cause-and-effect relationships.

In addition, the acute and chronic effects of phthalate
esters to estuarine organisms is relatively unknown. Additional
monitoring of phthalates in this River is recommended as well
as laboratory aquatic bioassay tests. Monitoring of oyster
growth and mortality, dissolved oxygen and salinity over a
several year period is required to determine if the observed
high mortality has subsided. Such monitoring would help to
identify causes of chronic stress indicated by reduced growth
rate of the oysters.

ABSTRACT

EVALUATION OF CHESTER RIVER MORTALITY BIOTOXICITY

Three studies were performed to determine whether the recent
dieoffs of oysters in the Chester River can be correlated with
point sources of toxic substances. To this end two kinds of ex-
periments were performed: long-term experiments with oysters
placed in the Chester River; and 96-hour acute toxicity experi-
ments with golden shiners, Notemigonus chrysoleucas, and crayfish,
Procambarus acutus acutus.

In the long-term studies, 10 stations were established.
Three stations were below the areas of known oysterkills, four
were within the areas of recent oysterkills and three were in
areas where the kills occurred in 1974 and 1975. One or two trays
with 96 oyster specimens were placed at each station. Ten or
twelve oysters were removed at nine intervals during 4 months.
These were scored for condition. No significant mortality occur-
red during this period, but during July and August 1978 the five
stations most upriver had dieoff of the fouling organisms and
reduced growth rates of the oysters.

The 96-hour acute toxicity studies were performed by placing
cages of golden shiners and crayfish in streams receiving ef-
fluents from the Campbell's Soup plant, the Tenneco, Inc. plant
and the sewage treatment plant for the city of Chestertown. No
significant mortality occurred.

No point sources of toxicants were located, but since no
significant mortality occurred during the study these results are
not conclusive.

PHTHALATE ESTERS AND RELATED CHEMICALS IN THE CHESTER RIVER BASIN

The Tenneco factory on Morgan Creek is permitted to discharge
up to 2830 kg of organic extractables per year. These waste
chemicals empty into the adjacent Tenneco Pond with a residence
time of ca. 10J days. The pond discharge flows through Morgan
Creek into the Chester River. The residence time in the river is
a much shorter 130 days.

Model concepts based on available data allow plausible calcu-
lations of discharge organics in the contiguous downstream sedi-
ments. The possibilities range from 0.04 ppm based on simple

partition to 5 ppm for a "hot spot" model based on previous dye
discharge experiments.

New chemical methods based on gas chromatography/mass spec-
trometry analysis of DBP (dibutylphthalate) and DEHP (di(2-
ethylhexyl)phthalate) were developed. The relative standard de-
viation was demonstrated at + 20 percent of the 0.1 ppm level in
sediment. Accuracy is more aifficult to specify, but this may
be judged on the basis of split samples measurea by an independ-
ent laboratory (0.3 ppm) and our own data (0.1 ppm).

MICROBIAL TRANSFORMATION OF TIN

This work was undertaken to determine if tin or products re-
sulting from the biotransformation of tin may contribute to oys-
ter mortality in the Chester River, Maryland.

Data were collected at two sites in the River: Spaniard
Bar, which suffered extensive oyster mortality, and Buoy Rock,
which did not exhibit extensive oyster mortality. Three sites
associated with potential sources of tin in the River were also
studied: the Tenneco plant and the Campbell's Soup plant, both
near Chestertown, Maryland, and the Chestertown sewage treatment
plant. For comparison, some samples were taken in Baltimore
Harbor, a site known to be polluted with heavy metals, and in
Tangier Sound near Tilghman Island, a site regarded as relatively
free of pollution.

Water and sediment samples were examined for total viable
counts of microorganisms, for.counts of microorganisms resistant
to inorganic tin, and for counts of microorganisms resistant to
organic tin. Sediment from each site was used as inoculum for
cultures to determine if microorganisms at the site could trans-
form inorganic tin to volatile (organic) tin compound(s). Water
and sediment were assayed for tin content.

Among physiochemical parameters measured onsite, only low
dissolved oxygen is a potential contributor to oyster mortality.
Microorganisms resistant to inorganic tin were detected in all
samples and most samples contained microorganisms resistant to
organotin, although organotin was more toxic than inorganotin to
the microbial flora. Microorganisms capable of converting in-
organic tin to volatile tin compound(s) were present at every
site. Comparison of tin concentrations at the several sites
showed that it is not possible to attribute the oyster kill
solely to tin, although interaction with other stress factors is
possible.

This work was submitted in fulfillment of contract
R805976010 by the University of Maryland, Chesapeake Biological
Laboratory under contract to the Maryland Water Resources Admin-
istration and sponsored by the U.S. Environmental Protection

vii

Agency. This report covers the period January 1, 1978 to July
23, 1979, and work was completed as of August 6, 1979.

viii

CONTENTS

For eword . . . . . . . . . . . . . . . . . . . . . . . . .

Preface . . . . . . . . . . . . . . . . . . ... . . . . . iv

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . vi
Figures . . . . . . . . . . . . . . . . . . . . . . . . . xi
Tables . . . . . . o . . . . . . . . . . . . . . . . . . . xii
Acknowledgement . . . . . . . . . . . . . . . . . . . . . xiv

1. Introduction . . . . . . ... . . . . . . . . . . . 1
Evaluation of biotoxicity . . . . . . . . . . . . . 1
Phthalate esters and related chemicals . . . . . . 1
Microbial transformation of tin . . o . . o . . 2

2. Conclusions .: . . . . . . . . . . . . . o . . . 5
Evaluation of biotoxicity . . . . . . . . . . . . . 5
Phthalate esters and related chemicals 5

Microbial transformation of tin . . . . . . . . . . 6

3. Recommendations . . . . . . . . . . . . . . . 8
Evaluation of biotoxicity . . . . . . . . . . . . . 8
Phthalate esters and related chemicals . . . . . . 8
Microbial transformation of tin . . . . . . . . . . 9
4. Evaluation of Chester River Oyster Mortality
Biotoxicity ... . . . . . . . . . . . . . . . . 10*
Methods and Materials . . . o . . . . . . . . . . . 10
Procedures . . . . . . . . . . . . . . . . . . . . 17
Results . . . . . . . . . . . . . . . . . . . . . . 24
Discussion . . . . . . o . . . . . . . . 36

5. Phthalate Esters and Related Chemicals in
the Chester River . . . . . . . . . . . . . . . 39

Introduction . . . . . . . . . . . . . . . . . . . 39

Distribution Patterns . . . . . . . . . . . . . . 43

ProbleM Statement . . . . . . . . . . . . . . . . 43

Experimental Procedures . . . . . . . . . . . . . 44

Results and Discussion . . . . . . . . . . . . . . 56

6. Microbial Transformation of Tin . . . . . . . . . 87

Experimental Procedures . . . . . . . . . . . . . 87

Results and Discussion . . . . . . . . . . . . . . 91

Appendices

A. Chester River Oyster Mortality . . . . . . . . . . 103

B. Account of Interlaboratory Tests on Split Samples. 112

C. Methods for Water Quality Measurements in
Tin Study . . . . . . . . . . . . . . . . . . . 114

FIGURES

Number Page

1 Lower Chester River showing oyster tray stations
1 thru 6 . . . . . . . . . . . . . . . . . . . . . 12

2 Lower Chester River showing oyster tray stations
6 thru 10 . . . . . . . . . . . . . . . . . . . . . 13

3 mortality of golden shiners, by station, over the
96-hour observation period . . . . . . . . . . . . 31

4 Excess mortality of golden shiners, by station,
during 96 hours of observation . . . . . . . . . .. 33

5 Map of Chester River illustrating sample sites . . . 40

6 Downstream progession of oyster mortality . . . . . . 41

7 GCMS of Chester River mouth sediment extract . . . . 46

8 Liquid chromatography of Tenneco Products designated
by Tenneco as Alkyl Phthalates, or mixtures
(6-10P and 7-11P) . . . . . . . . . . . . . . . . . 57

9 Liquid chromatogram of DOP, DIDP, and DTDP
(synthetic mixture of Tenneco samples) . . . . . . 58

10 Liquid chromatogram of Tenneco pond water . . . . . . 59

11 Liquid chromatogram of Tenneco pond sediment . . . . 61

12 Liquid chromatogram of Morgan Creek sediment from
Frye Farm a few miles downstream from Tenneco . . . 63

13 Comparison liquid chromatogram of humic acid . . . . 63

14 Comparison of GCMS-SIP1 chromatograms . . . . . . . . . 70

15 Oyster tissue extract (spiked with 20 ppm DEP,
DBP, and DEHP) . . . . . . . . . . . . . . . . . .. 78

16 oyster tissue extract . . . . . . . . . . . .. . . . . 79

TABLES

Number Page

1 Physical Parameters Measured at Chester.,River
Oyster Study Sites . . . . . . . . . . . . . . . . 14

2 Hydrographic observations for the Possible Point
Source Studies . . . . . . . . . . . . . . . . . . 22

3 Mortality of Oysters During This Study . . . . . . . 24

4 Comparison of Mean New Growth in Oysters Planted
in Chester River . . . . . . . . . . . . . . . . . 25

5 Comparison of Mean observed Condition Scores in
Oysters Planted in Chester River . . . . . . . . . 27

6 Associated Organisms by Station . . . . . . . . . . 28

7 Relationship of Nonsurvival to Autopsy Data . . . . 34

8 Comparisons between Stations for Characters Related
to Nonsurvival . . . . . . . . . . . . . . . . . . 35

9 Field Sample Handling . . . . . . . . . . . . . . . 52

10 Methodology . . . . . . . . . . . . . . . . . . . . 53

11 Effect of Sonication on DBP Extraction by CH 2Cl 2
from Sample "R .. . . . . . . . . . . . . . . . . . 62

12 Comparison of Ultrasonic Extraction Efficiencies
of Varied Solvents Measured on Working Standard 64

13 DEHP Recovery Measurements Using Dichloromethane 65

14 Comparison of Methods for Extraction of Phthalate
Esters from Chester River, Mouth Sediment,
Test 1 . . . . . . . . . . . . . . . . . . . . . 68

15 Test of Varied Extraction and Re-extraction
Procedures . . . . . . . . . . . . . . . . . . . . 68

xii

Number Page

16 Interlaboratory Comparison of DEHP Measurements
of Split Samples . . . . . . . . . . . . . . . . . . 71

17 Determination of Sediment Composition . . . . . . . . 73

18 C oncentration of Phthalates in Oysters and
Related Sediments . . . . . . . . . . . . . . . . . 80

19 Determination of a Concentration of Tin Which Would
Select for Tin-Resistant Microorganisms . . . . . . 89

20 Physical Data for April 1978 Cruise/Excursion . . . . 92

21 Physical Data for July 1979 Cruise/Excursion . . . . . 92

22 Viable Counts of Bacteria from April 1978 Cruise . . . 93

23 Viable Counts of Bacteria from July 1979 Cruise . . . 94

24 Production of Volatile Tin in Media Inoculated
with Sediment . . . . . . . . . . . . . . . . ... . 96

25 Tin in Water and Sediment Samples . . . . . . . . . . 97

xiii

ACKNOWLEDGEMENTS

The Evaluation of Chester River Oyster Mortality Biotoxicity
study was performed by F. Douglas Martin and Willem H. Roosenburg
of the University of Maryland, Center for Environmental and Estu-
arine Studies, Chesapeake Biological Laboratory, Solomons, Mary-
land, 20688. They wish to thank Dr. George Krantz of UM/CEES
Horn Point Laboratories who advised on appropriate location of
stations; Robert Miller, Daniel Carver, Sharon Ambrose, Dorothy
Cutshaw, Wayland Owens, Rev. Mark Odell and others who assisted
in field work; and Captain William Keefe and several Kent County
residents for rescue services.

David H. Freeman, Department of Chemistry, University of
Maryland, College Park, Maryland, 20742, authored the study on
Phthalate Esters and Related Chemicals in the Chester River
Basin. The experimental work in this study was carried out by
James C. Peterson and Sally A. Gingras. The work depended
critically upon ultraclean glassware and John Trembly cooperated
by facilitating routine use of the Chemistry Department's an-
nealing ovens for this purpose. Mr. Arden Fox of Tenneco was
helpful in providing details and insights to the history and
present operations of the Tenneco plant and its bacterial waste
processing facility. Captain O'Berry and the crew of the
"Aquarius" ably assisted the collection of the Chester River
water and sediment samples. The research sailing vessel "Huckle-
berry Friend" and the facilities of the Podickory Sailing Assoc-
iation were used to gather the reference samples from the mouth
of the Chester River. Special thanks go to Dr. William Budde
who generously provided cooperative measurements on split sedi-
ment samples. Finally, we happily acknowledge the value of
Professor Ron Hites' intensive short course on environmental
applications of GCMS. Drs. Nelson Frew, Robert Gagosian and
John Farrington at Woods Hole Oceanographic Institution provided
helpful consultation during the course of this work.

Microb.ial Transformation of Tin was under the direction of
Joseph Cooney,.Chesapeake Biological Laboratory, Center for
Environment and Estuarine Studies, University of Maryland,
Solomons, Maryland, 20688. Experimental work was conducted by
L. E. Hallas, with the assistance of Marthe Cole-Jones and
Terri Ekelund.

xiv

SECTION 1

INTRODUCTION

EVALUATION OF BIOTOXICITY

The Chester River is a tributary of the Chesapeake Bay,
northeast of Kent Narrows. Historically, this estuarine part of
the river has been rich in natural oyster bars; but, in the
winter of 1974, a heavy mortality started upriver and moved over
a period of years to bars in the lower reaches. The symptoms of
this oyster mortality could not be attributed to any climatic or
other natural phenomena that cause occasional oyster mortalities.
A suspicion became plausible that something highly toxic to
oysters had entered the upper river and had gradually worked its
way downstream. Areview of possible point sources of waste dis-
charge into the Chester River system revealed the presence of the
Chestertown sewage plant on Radcliffe Creek, the Campbell's Soup
factory on Morgan'Creek and the Tenneco plant that dumps its ef-
fluent into a pond which eventually empties into Morgan Creek via
a small creek. The investigations reported here were aimed at
locating point sources of toxic materials, and stations were
chosen to maximize the ability to identify the roles of these
potential sources.

PHTHALATE ESTERS AND RELATED CHEMICALS

The State of Maryland has issued a permit to Tenneco, In-
corporated, in Chestertown, Maryland, to discharge 10 ppm of
total organic extractables into Morgan Creek which empties into
the Chester River and, then, into Chesapeake Bay. This permit
would appear to be reasonably conservative unless it conceals the
basis for long-range harm. Such threatening possibilities do
exist and will be considered in the present studies. The major
problem is related to the question of whether the discharge is
free to dilute itself in some innocuous way, or whether, on the
contrary, the dilution processes are blocked and the basis for a
toxic accumulation can be shown. In the latter case, the permit
would have to be viewed as nonconservative and the ecological
threats would have to be carefully considered and perhaps further
constrained.

The present study includes an initial investigation of
various models for the distribution of the chemicals discharged
from the Tenneco site. Various options are considered between
the two extremes--that the organics are fully trapped at the
Tenneco site, or that all the organics are distributed uniformly
into the Chester River sediment beds.

The chemical analysis of sediment and oyster tissue for alkyl
phthalates is prone to a myriad of error sources due to ubiqui-
tous presence of these plasticizer compounds in the laboratory.
As a result, there is a need to develop new technology that would
scrupulously suppress the opportunity for contamination. The
development succeeded because it relies on a well-established
statistical maxim--that any risk if repeated often enough must
lead to disaster. Since chemical methodology consists of a
series of steps, each with an assigned risk, the approach was to
develop a procedure that was designed to approach zero reliance
on chemical manipulations.

Quality assurance was not a part of the original work plan.
However, independent laboratory tests of split samples showed
that the developed chemical methodology was indeed adequate for
the purposes at hand.

The underlying goals, then, were to develop and test the new
methodology, to explore the possible causal link of industrially
discharged alkyl phthalates to the past oyster mortality, and to
determine whether the present discharge constitutes an ecological
threat.

MICROBIAL TRANSFORMATION OF TIN

Organotin compounds were first synthesized about 1850 (Van
der Kerk 1976), and they were first used as agents to control
biological activity around 1930, when they were used as moth-
proofing agents (Luijten 1972). Shortly thereafter, organotins
were used as stabilizers for vinyl resins, which continues to be
a major application (Subramanian 1978). In the last 10 years,
use of tin by industrial societies has more than doubled. Or-
ganotin compounds are used widely to control a variety of plants,
animals, and microorganisms (Deschiens and Floch 1962, Daum 1965,
Holden 1972, Luijten 1972, van der Kerk 1976). All such organo-
tin compounds are toxic, but the effect varies with the organic
group(s) present (Thayer 1974). In general, triorganotin com-
pounds are more toxic than di- or tetraorgano compounds. Dior-
ganotins behave like organomercurials, reacting with sulfhydryl
groups to inactivate enzymes. Trialklyl tin compounds interfere
with oxidative phosphorylation and with photosynthetic phos-
phorylation (Thayer 1974). Methyl tin compounds are poisonous
to the central nervous systems of higher organisms (Ridley,
Dizikes, and Wood 1977). Effects can be species-specific. For

example, at concentrations too low (15-30 ppb) to affect some
fish, triphenyltin acetate and several tributyltin compounds af-
fect snails, zooplankton, and small fish, while warm-blooded
species show no toxicity. In contrast, some soil microorganisms
are not affected by concentrations of tributyltin oxide at con-
centrations up to 100 ppm (Thayer 1974).

In comparison with other metals such as mercury, lead, and
cadmium, relatively little is known about biological transforma-
tions of tin. But sufficient information is available that a
biological cycle has been proposed (Ridley et al. 1977).

hv
(CH3)4Sn a.
@b
2e- (CH
SnX4-'@-+SnX2 3)3SnX
e- e- b bj@a
SnX3 + CH3 0 CH3SnX3 &a (CH3)2SnX2

b
(a) CH3SnX3 + e- + CH3 (CH3)2SnX2 + X_

in the diagram, X indicates a counteranion. The free radical
SnX3 can be methylated in a series of biological reactions (re-
action a) involving methylocobalamins (e.g., vitamin-B12) to
yield mono-, di-, tri-, and tetramethyl tin (Ridley et al. 1977).
A Pseudomonas species which is purportedly capable of methylating
tiii--has been isolated from Chesapeake Bay (Huey et al. 1974). In
the presence of ionic mercury, trimethyltin can react to yield
"2

the highly toxic methylmercury (Huey et al. 1974). Methyltins

could be cleaved oxidatively (reaction b) by mixed function
oxidases in the same way that trialkylyltin derivates are
cleaved by liver microsomes (Kimmel, Fish, and Casida 1977).
Alkyltins can also be degraded by photolysis (*) to yield free
radicals (Lloyd and Rogers 1973).

Thus, tin can be transformed biologically to toxic compounds
and these toxic compounds can be transformed chemically to other
toxic compounds. Since tin is used at the Tenneco plant near
Chestertown, it is a potential source of compounds toxic to
oysters.

The objective of the present study was to determine if tin
or products resulting from the microbial transformation of tin
could contribute to oyster mortality in the Chester River. The
investigation was not designed as a definitive study, but as a
preliminary, screening study to determine whether further studies
are warranted.

SECTION 2

CONCLUSIONS

EVALUATION OF BIOTOXICITY

Since there was no significant mortality in our experimental
oysters, there was no strong indication that the causative factor
for oysterkills in the Chester River was in operation during our
studies.

Fouling organisms died off in July and August at five sta-
tions upriver from Corsica Neck. Since this accompanied dieoff
of oysters in previous years, the phenomenon which is responsible
for oyster dieoff might have occurred but in a mild form.
Despite reasonable growth in the oysters planted in Chester River,
the growth rate during the period of fouling community dieoff was
significantly lower than that of controls placed in the Patuxent
River. The phenomenon this year may have been so mild that the
main effect in the oysters was a reduction in new growth.

The beginning of dieoff of the fouling community correlated
with a fishkill. The fishkill originated well above Morgan Creek.
The two phenomena may be unrelated as the fishkill may be bacte-
rial in origin and species-specific since only carp and catfishes
were noted dying.

Experiments with fish and crayfish in Radcliff Creek and
Morgan Creek do not find any indication that either creek con-
tains the sole source of the cause of oyster mortality within its
drainage; however, the lack of significant mortality during the
study period makes these results inconclusive.

PHTHALATE ESTERS AND RELATED CHEMICALS

The Chester River sediments taken from the vicinity of the
oyster mortality zone, as well as farther downstream, show no
evidence that Tenneco discharges are causally linked to the past
oyster mortality. Oysters are now grown with apparent health in
regions where the alkyl phthalates should be similar in concen-
tration to those presently measured in the vicinity of the oyster
mortality. Since the concentration history has not been measured,
the study cannot rule out the possibility of a past causal re-
lationship.

A massive buildup of alkyl @hthalates manufactured by Tenneco
was found in Tenneco Pond immediately adjacent to the factory
waste water discharge plant. The levels show clearly that the
pond sediments are serving as a sink, i.e., as a secondary waste
treatment facility.

Experimental measurements of Chester River sediments show no
significant differences between the mortality zone and the Ches-
ter River mouth where reasonably healthy oyster growth persists.
The results for DBP and DEHP are very nearly identical in these
regions in the Chester River: 0.02-0.85 ppm with average values
of DBP (0.5 ppm) and DEHP (0.05 ppm). The DEHP/DBP ratio is
0.1 + 0.07.

The Tenneco Pond results are quite different: DEHP (1.5 x
103 ppm) and DBP (0.2 ppm). Tenneco has rarely made DBP. The
difference in the DEHP/DBP ratio alone suggests that the alkyl
phthalates in the Chester River may originate from Tenneco as
well as other possible sources. Moreover, the estimated possible
accumulation in Tenneco Pond--103 kg of alkyl phthalates--sug-
gests that the pond functions in part as a waste treatment
facility.

The greatest threat seems to be that the Tenneco Pond may be
nearing the saturation state that it must reach eventually. In
that case, one can confidently forecast a serious accumulation
of the alkyl phthalates that will in time spread out into the
Morgan Creek area.

MICROBIAL TRANSFORMATION OF TIN

Examination of physiochemical data (pH, temperature,
salinity, dissolved oxygen) from four estuarine and three fresh-
water sites shows only low dissolved oxygen near the bottom as a
potential contributor to oyster mortality in the Chester River.

All sediment and water samples examined contain microorga-
nisms resistant to inorganic tin; resistant organisms comprised
as much as 55 percent of the total aerobic, heterotrophic popu-
lation detected. Most of the water and sediment samples con-
tained organisms resistant to the organotin compound,
dimethyltin chloride; such organisms comprised as much as 17
percent of the total aerobic, heterotrophic population detected.

Microbial populations are more sensitive to organotin than
to inorganic tin.

Microorganisms capable of converting inorganic tin to
volatile tin compound(s) are widely distributed in the Chesa-
peake ecosystem.

All sed iments associated with the Chester River--including
sediments from the Tenneco plant, the Campbell plant, and the
Chestertown sewage treatment plant--contained more tin than sedi-
ment from a site near Tilghman Island and less tin than a site in
Baltimore Harbor. Spaniard Bar, which suffered an oysterkill,
did not yield significantly more than Buoy Rock, which did suffer
such a kill. Thus, it is not possible to attribute the oyster-
kill in the Chester River solely to pollution by tin, although
interaction with other stress factors is possible. In addition
to sediment, water in the Chester River and water entering the
Chester River from the Tenneco plant, from the Campbell plant,
and from the Chestertown sewage treatment plant sometimes contain
significant quantities of tin.

Significant progress has been made toward developing a method
for separation and qualitative and quantitative measurement of
organotin species in environmental samples and in microbial cul-
tures.

SECTION 3

RECOMMENDATIONS

EVALUATION OF BIOTOXICITY

Possible links between the dieoff of associated organisms and
previous oysterkills should be further investigated. In addition
to monitoring fouling communities of oysters, transite sheets
should be placed on or near oyster bars and be sampled periodi-
cally. These sheets make quantification of the fouling organisms
simpler and quicker when the effects are not gross.

The slower oyster growth in Chester River may be related to
the oysterkill causal factor(s). Year-round sampling of oyster
bars in the Chester River to monitor new growth is recommended.
An area with a similar salinity regime and without such oyster-
kill phenomenon should be sampled simultaneously as a control.
The studies recommended above could be performed concurrently.

A study continuously monitoring dissolved oxygen during
periods when oysterkill may occur is recommended.

PHTHALATE ESTERS AND RELATED CHEMICALS

The levels of alkyl phthalates observed in the Chester River
do not appear to present an immediate threat. However, the oyster
lives a perilous existence. It is quite possible, and perhaps
likely, that the alkyl phthalate levels are changing rapidly
enough in the Chester River and the greater Chesapeake Bay to
constitute a serious threat at some future time.

It is recommended that the rate of buildup be monitored
through piston cores taken in well-stratified sediments. The
goal should be to forecast the date when the extrapolated levels
will be reached where healthy oyster production will be prevented
by the presence of these ubiquitous chemicals.

It is recommended that the Morgan Creek sediments be surveyed
for evidence of alkyl phthalate accumulation. The creek serves
as a conduit between the Tenneco pond and contiguous farmland
using the discharged water for irrigation purposes. The Tenneco
Pond sediments are likely to be weakening in their role as a sink
for these chemicals. Eventually, their sorptive capacity is

likely to become exhausted.

It is recommended that Tenneco Pond be surveyed to establish
the total phthalate plasticizer content. At the same time, co-
operative research by Tenneco should be elicited to see if a more
effective use of secondary waste water treatment by certain soils
could be established in order to furnish long-range protection to
the downstream ecology.

MICROBIAL TRANSFORMATION OF TIN

Tin should not be considered as the sole source of the exten-
sive oyster mortality observed in portions of the Chester River.

Tin should not be excluded as a partial cause of oyster mor-
tality in the Chester River, particularly when coupled with other
pollutants and with low dissolved oxygen.

When tin enters an aquatic ecosystem, it should be assumed
that microorganisms are present which can convert it to volatile
tin compounds.

Studies should be undertaken, using recently developed meth--
odology, to determine if oysters bioaccumulate tin and if the
oyster's gut flora can produce significant quantities of volatile
tin compounds.

SECTION 4

EVALUATION OF CHESTER RIVER OYSTER MORTALITY BIOTOXICITY

METHODS AND MATERIALS

Possible Sources of Toxicants

The Chester River has a typical estuarine portion with widely
varying temperature and salinity regimes. All of the possible
sources of toxicants that were examined have their waste mate-
rials entering the river in its tidal portion. These possible
sources dump materials of widely differing nature. The Chester-
town sewage plant discharges chlorinated and treated domestic
sewage. The Campbell's Soupfactory sprays its effluent, which
are wastewaters from preparing chicken, over percolation fields
and each leached water discharge is chlorinated before it enters
into Morgan Creek. Tenneco does not discharge any effluent that
can be made more acceptable by chlorination, as its effluent is
resultant from manufacture of plasticizers. Instead of chlori-
nating, the factory discharges into a leaching pond at one end
with a stand pipe overflow on the far end. The volume of ef-
fluent at the time of the study was a small volume compared to
both the sewage plant and Campbell's Soup plant.

The proposal called for two separate but related studies:
(1) a field study on mortality and growth of oysters in the
Chester River, and (2) a field study to detect possible toxic
effects on test animals by the effluents of the three plants
discharging into the Chester River basin.

Oyster Studies--
The oyster portion of the study was conducted in the portion
of the river from just below Chestertown to just above Kent
Narrows, which is near the mouth. There were 10 stations in the
Chester River from 15 May 1978 until 18 September 1978. Fifteen
trays, each containing 96 oysters, were monitored for mortality.
Qualitative information on oyster communities' associated orga-
nisms were also recorded.

Stations--See Figures 1 and 2 for orientation of stations.
Comments concerning these stations are given below.
Station 1. North of Kent Island, a single-tray station. This
tray was attached to parts of a sunken barge. At
the first sampling visit the tray had been tampered
with and when retrieved it came up upside down with
the lid open. No oysters could be recovered and the
station was abandoned.

Station 2. South of Cedar Point. This was a two-tray station
attached to clam buoy 'ISS." This station lasted
from 15 May until 27 June but could not be found
with the grapnel nor by diving on 12 July.

Station 3. Off Tilghman Creek. This single-tray station was
tied to a stake at 0.6 m depth. This station lasted
the entire study period.

Station 4. off Piney Point. A double-tray station tied to clam
buoy "C." This station served from 15 May - 4
August but could not be found on the 23 August
check and thereafter.

Station 5. At Ringgold Point, a single-tray station, tied to a
Coast Guard day marker by a nylon line to a ring on
this structure. The line was cut by vandals before
this station was visited on 27 June. The station
was observed from 15 May - 14 June.

Station 6. Off Corsica River. This station held two trays and
was fastened to clam buoy "A." The tieline was at-
tached to the buoy's anchor chain by means of a
chain ring that dropped to the bottom. This sta-
tion lasted the entire study period from 15 May to
18 September.

Station 7. Nichol's Point, a double-tray station. This station
was tied to a ring on a day marker. It was in
service for the entire 15 May - 18 September period.

Station 8. Off Cliffs Point, a single-tray station attached to
the remnants of a booby blind in about 1 m of water.
The station lasted the entire study period.

Station 9. Off the mouth of Shippen Creek. This single-tray
station was attached to a privately owned pier and
was not tampered with.

Station 10. Newman's Wharf just off Deep point. This two-tray
station was also suspended from a private pier and
lasted the entire study period.

Fj-
LQ

(D 6
r

F@ t-l -
0
rI :4
(D
I-i" 5

(D
En 4
rt
(D

En
NJ ;zr
0

F'-
10
LQ

0
11 2
Lei
rt CHESTER RIVER
(D

rt,
lkm
0)

pi
rt
FJ-
0
@:l
En

0
rt :@
@r m

C> (D

rt,
(D

LA)
0
Q7

D..p

rt 7 sp..,.,d PI
(D

(t
CORSICA RIVC, lkm

ul 6
rt,
0)
(t
FJ-
0

Hydrographic observations of all stations are contained in
Table 1.

TABLE 1. PHYSICAL PARAMETERS MEASURED AT CHESTER
RIVER OYSTER STUDY SITES

Depth Surface/Bottom Secci
W Salinity Temperature Dissolved 02 Depth
(ppt) (0c) (ppm) (m)

STATION 1

15 May 1.5 7.6/7.4 14.0 /14.1
23 May 5.5 18.0 /18.0 9.7 9.5 1.0
2June
14 June
27 June
12 July
4August
23 August
18 Sept.

STATION 2

15 May 3.3 7.1/7.1 14.3 /14.5
23 May 5.5 20.0 /19.0 9.9 9.2 0.8
2June 5
14 June
27 June 6.3 25.3 1.3
12 July 7.5/7.5 25.0 /22.5 6.9 6.5
4August
23 August
18 Sept.

STATION 3

15 May 1.5 6.9/7.1 14.1 /13.8
23 May 5.8 24.0 /24.0 9.3 / 9.8 0.5
2June 5
14 June 5 21.3 7.8 0.6
27 June 6.5 26.5 1.1
12 July 7.0 24.7
4August 8.0 25.8 5.5
23 August 10.0 27.2 6.8
18 Sept. 13.0

(continued)

TABLE 1. (continued)

Depth Surface/Bottom Secci
W ga'iinity Temperature Dissolved 0 Depth
(Ppt) (0c) (ppm) 2 W
STATION 4
15 May 3.6 6.6 /6. 8 15.1 /14.6
23 May 5.2 19.5 /19.0 11.4 / 8.4 0.5
2June 6
14 June 5 20.8 7.5 0.6
27 June 6.2 28.0 0.8
12 July 6.2 /6.2 24.7 /23.5
4August 7.0 26.1 /25.8 7.4 5.6
18 Sept.

STATION 5
15 May 3.0 6.0 /6.0 15.3 /15.3
23 May 5.8 21.0 /19.5 9.5 9.6
2June
14 June 5.0 21.0 8.4 0.9
27 June
12 July
4August
18 Sept.

STATION 6
15 May 3.6 5.8 /6.5 15.3 /14.8
23 May 5.9 19.0 /19.0 9.3 7.8 0.6
2June
14 June 5.5 21.0 8.0 0.7
27 June 6.0 28.2 0.6
12 July 7.2 A-8 25.2 /24.4
4August 6.2 26.2 /25.9 5.9 4.7
23 August 8.0 28.8 /26.6 7.0 5.4
18 Sept. 9.0

STATION 7
15 May 2.4 5.4 /5.4 15.3 /15.3
23 May 5.5 19.0 /18.6 8.2 7.7.
2June
14 June 5 22.1 7.2 0.7
27 June 6.0 28.3 0.7
12 July 6.0 /5.9 25.2 /25.8
4August 6.0 26.2 5.9
23 August 8.0 27.4 /27.0 6.8 5.8
18 Sept. 8.0

(continued)

TABLE 1. (continued)

Depth Surface/Bottom Secci
W Temperature Dissolved 0 2 Depth
(Ppt) (OC) (ppm) W

STATION 8
15 May 3.4 5.0/ 5.0 15.2/ 15.2
23 May 4.8 18.0/ 19.0 9.2/ 8.3
2 June 6
14 June 5 22.0 6.9 0.6
27 June 5.5 28.0 0.5
12 July 5.8 26.2
4 August 6.0 26.5 5.4
23 August 8.0 .27.8/ 27.0 6.5/ 5.4
18 Sept. 7.0

STATION 9
15 May 4.3 4.4 /4.2 15.3/ 15.1
23 May 4.5 18.5/ 181.5 8.2/ 8.0
2 June 6
14 June 4 23.0 6.8 0.4
27 June
12 July 5.0 26.0
4 August 5.8 26.5 5.8
23 August 7.0 27.8 27.0 5.8 /5.2
18 Sept. 9.0

STATION 10
15 May 14.3 3.3 /3.3 15.7 /15.7
23 May 3.0 19.0 /19.0 8.0 /7.5
2 June 2
.14 June 2 22.5 7.5
27 June 0.2
12 July 2.2 26.2
4 August 3.0 26.9 4.8
23 August 5.0 27.5 6.0 /5.6
18 Sept.

PROCEDURES

Oysters used in the field experiment had to be obtained from
a source that was free of phthalic esters, therefore, the natural
oysters in the Chester River could not be used. Cultchless
oysters with a height of approximately 3 cm were bought from
Frank Wilde's Hatchery in Shadyside, Maryland. The oysters were
distr,,ibuted randomly over 15 trays with 96 oysters per tray.
Trays were of stainless steel and measured 40 cm x 92 cm x 10 cm
with a mesh of 2 cm x 2 cm. The lids were hinged by two rings
through both lid and tray on one of the long sides and a stain-
less steel clip hasp lock on the opposite side. Each tray was
bridled by a 6.4 mm braided nylon line from each top corner that
were joined about 1 m above the tray with a loop. Another 6.4 mm.
nylon line was fastened to the loop on one end and to the station
on the other. Stations 8, 9 and 10 were suspended, but touching
the bott'pm, from permanent structures; all other trays were set
on the bottom with the line attached to either a buoy anchor
chain, a ring on a day beacon, or to private docks. All trays
had two bricks tied on edge to their bottom to hold the trays
above the muck.

Sampling Visits

A total of nine visits were made to the study area. Eight
visits were collecting and observation field trips after the sta-
ti.ons had been established on 15 May. These visits were made on
23 May, 2 June, 14 June, 27 June, 12 July, 4 August, 23 August
and 18 September. Not all stations were visited on each field
trip due to mechanical breakdown of the outboard motor. Dates
of visits and stations sampled are given below.

Visit 1, 15 May--Was done from R.V. AQUARIUS and established the
stations under the guidance of Dr. George Krantz. All sub-
sequent visits were made by outboard runabout.

Visit 2, 23 May_-Twelve oysters were collected from every sta-
tion except station 1 where the tray had been turned upside
down with the lid open. Qualitative notes were kept on
presence or absence of organisms associated with oyster com-
munities.. For hydrographic data see Table 1.

Visit 3, 2 June--Visited stations 2, 3, 4J, 8, 9 and 10 success-
fully. Station 1 had been abandoned. *Station 6 could not
be found. Stations 5 and 7 had gotten so badly tangled up
in the rusty metal of the day markers that the trays could
not be brought to the surface for inspection and sample
collection.

Visit 4, 14 June--Successfully collected samples from stations
3 - 10.

Visit 5, 27 June--Successfully collected oyster samples from
stations 2, 3, 4, 6, 7 and 8. Station 5 had been removed by
vandals. Outboard motor trouble prevented sampling of sta-
tions 9 and 10.

Visit 6, 12 July--Collections were made from stations 2, 3, 4, 6,
7, 8, 9 and 10.

Visit 7, 4 August--Collected from stations 3, 4, 6, 7, 8, 9 and
10. Station 2 could not be retrieved by either grapnel or
by diving. Efforts to locate this station's trays on
subsequent visits failed.

Visit 8, 23 August--Failed to bring up station 4 with either
grapnel or diving. Samples were taken at stations 3, 6, 7,
9 and 10.

Visit 9, 18 September--Ended the observation period. One tray
from each remaining two-tray station was taken out. The
remaining trays were left to be collected at a later date
so that oysters could be checked for phthalic esters after
a full growing season. However, those trays tied to the
clam buoys were taken up by D.N.R. and no identification
of trays could be made.

Controls--A control station was established on 15 May at the
pier at Solomons. Sampling of the controls coincided with
every visit to the Chester River stations.

Oyster Scoring Procedures

Oysters were brought into the lab for gross examination and
preservation. Oyster samples collected from the tray stations
were packed in wet paper towels and kept under refrigeration
until they could be examined. Oysters were shucked by severing
the aductor muscle at the left valve and displaying the oyster
in its right valve with the left valve removed. Features de-
scribed were shell height, new growth, color of the meats,
condition (which will be further explained) and possible imper-
fections. Most common imperfections were Polidora websteri
infections, shell ulcers, muscle ulcers, and mud bli-sters.
Tissue ulcers and bill obstructions were rare and no pea crabs
Pinnotheres ostreum or sponge penetration by Cliona celata were
found (salinity too low).

Length and new growth measurements were recorded in mm, but
imperfections and infestations were scored on a scale of 0 (not
present) to 5 (omnipresent). The scoring of observed oyster
condition (not to be confused with condition index (CI) which
is a totally different procedure) was also according to a scale
of 0 to 5 but deserves elucidation of its criteria:

0 = Dead.

I- = An oyster that is completely transparent is definitely un-
healthy, all organs clearly visible (no gonad or glycogen).
It does not fill its shell cavity but lies like a limp
piece of clear gelatin on its valve. This oyster is mori-
bund.

1 Same as above with slight nuances for improved scoring.
This oyster may live.

Oyster not transparent but liver (digestive gland) clearly
visible. Gonad or glycogen low and patchy. Does not fill
its shell cavity well. Not moribund.

3 Average looking. Reasonable amount of gonad and/or glyco-
gen. Liver barely visible. Mantle appears to be an active,
well-functioning organ. If examined immediately after col-
lection a crystalline style may be found.

4 Good looking. Enough "fat" to cover liver completely.
Plumper than 3, therefore, fills its shell cavity shell.
Mantle uniform rich color. Mantle edge active to stimu-
lae. Healthy heartbeat. Solid crystalline style upon
examination immediately after collection.

5 Excellent looking. The oyster appears cramped by its
shell cavity. It bulges over the half shell with thick
layers of gonad and/or glycogen. Clean uniform coloration
throughout. No deficiencies of any kind noted.

Upon completion of this gross examination, the oyster meats
were individually wrapped in cheesecloth together with a num-
bered tag for later identification and preserved in Davidson's
fixative.

Data for new growth and mean condition score were compared
using Student's t-test.

Possible Point Source Studies

A 96-hour toxicity study was performed at stations in the
vicinity of possible point sources of pollution. These sources
were the Chestertown Sewage Treatment Plant, the Tenneco chemi-
cal plant which manufactures plasticizers and the Campbell's
Soup Company where chicken carcasses are prepared and diced.
Tenneco and Campbell's discharge into Morgan Creek while the
Chestertown Sewage Plant discharges into Radcliffe Creek shortly
before it joins the Chester River at the southern edge of Ches-
tertown.

Stations--

There were a total of six stations. The control station was
located at least 1-1/2 km upstream from the Tenneco plant. This
station was in a small tributary to Morgan Creek but the same one
which carries the overflow from the discharge pond of the Tenneco
Pond to Morgan Creek.

Tenneco Pond--located in the Tenneco discharge pond about 6 m
away from the standpipe which is the point farthest away
from the plant's point of discharge. Depth at this station
was about 0.6 m.

Tenneco Downstream--in the receiving stream from the Tenneco
Pond about 350 m downstream from the pond. Depth at this
station was 0.5 m maximum. No tidal influence was noted.

Campbell Upstream--in Morgan Creek near the middle discharge
pipe of the Campbell leaching fields in 0.8 m depth at low
tide. Due to tidal amplitude, this station could only be
checked once a day as at high tide the depth became 1.6 m
or more.

Campbell Downstream--also in Morgan Creek about 500 m from
station upstream and located just upstream from Campbell's
most downstream discharge. Tidal difference here was also
more than 0.6 m, which often limited mortality checks to
once a day.

Sewage Plant--was located in Radcliffe Creek about 6 m down-
stream from the sewage outfall. Tidal difference varied
depth from 0.5 m to well over 1 m, which allowed only
one inspection per day.

Procedures---
Experimental species for this investigation were the golden
shiner, Notemigonas crysoleucas and the crayfish, Procambarus
acutis acutis. These species were chosen because they are
native to the area, are easily available, and easily acclimated
to new conditions and represent animals which are neither overly
delicate nor particularly pollution resistant forms. Golden
shiners were obtained from Green Valley Minnow Farms, Brogue,
Pennsylvania. Crayfish were obtained by seining various creeks,
ditches, and rivers on the Eastern Shore. All animals were held
for a minimum of 2 weeks in sand-filtered Solomons well water.
Temperatures were adjusted to those that were expected to be
encou,ntered in the field. Both species were fed Purina Trout
Chow(T daily. Transportation to and from the laboratory was in a
1400 1 tank aerated using tanks of compressed air.

Test animals were contained in cages of 30.5 cm x 30.5 cm x
91.4 cm with a 13 mm rebar frame covered with 3.2 mm mesh nylon
netting. 20

The experiment lasted 96 hours. one cage with 30 golden
shiners and one cage with 15 crayfish were put overboard at each
station. Where possible, the cages were checked twice daily and
dead animals were removed and preserved for future autopsy.
Once a day the animals were fed Furina Trout Chowo.

All survivors of the 96-hour experimental period were pre-
served in Davidson's preservative for later autopsy.

Hydrographic data, recorded at every visit, were temperature,
salinity, dissolved oxygen and residual chlorine (see Table 2 for
these data).

Autopsy of Specimens--

The autopsy of golden shiners was divided into two parts--
external and internal features. External features included
standard length, examination for damage such as broken, absent
of excessive mucus covering; missing or deformed scales, abra-
sions on body and/or fins; and afflictions such as fungus, dis-
colorations, cysts and parasites. Internal autopsy examined
general appearance, color, damage and foreign material on or in
the buccal cavity, gills and gill arches; texture, size, color,
content and abnormalities of stomach, internal and external
intestin4l lining, liver, gasbladder, gonads, spleen and adipose
tissue. Any observations that did not fit into the prepared
autopsy sheet were recorded under item "Other." All specimens
were individually wrapped in a bag of cheesecloth containing an
identifying tag and were placed in fresh Davidson's solution.

Cages containing crayfish were checked at the same time as
the cages with golden shiners. At that time dead crayfish were
removed and preserved in Davidson's solution. All surviving
crayfish were preserved at the end of the 96-hour experimental
period.

Autopsy of crayfish examined both external and internal
features. External features such as injury, regeneration to ap-
pendages and body, hardness of cephalothorax, color, affliction
with fungus, discoloration, and bacterial infection were recorded.
Internal features such as condition and foreign material in
the gills and gill chamber, texture and content of cardiac and
pyloric stomach, condition and appearance of heart, hepatopan-
creas and gonads were recorded.
After autopsy the specimens were wrapped individually in a
cheesecloth bag that contained an identification tag and placed
in fresh Davidson's preservative.
Autopsy data were examined for correlations using contingency
analysis with chi-square tests.

TABLE 2. HYDROGRAPHIC OBSERVATIONS FOR THE POSSIBLE POINT SOURCE STUDIES
Site Sept. Dissolved Total Residual
Date Time Temp rapure Saj@@it@ Ox Chlorine
,Zgen
VOC 00

Controls 25 1430
26 1400 15.2 0 6.9 0
1800 15.5 - 10.4 -
27 0645 12.0 6.7
1800 - -
28 0800 13.5 6.5
1500 17.2 4.8
29 0930 11.3 - 6.8

Tenneco Pond 25 1135
26 1300 21.0 2 13.8 <0.05
1900 21.0 2 17.8 -
27 0730 16.0 0 8.4 -
1845 25. @l 20+ 0.04
28 0845 17.5 <1 10.3 0
1545 21.1 2 16.2 0.01
29 1200 17.3 - 7.2 <0.02

Tenneco 25 1100
Downstream 1900 19.5 3.8 -
26 1230 15.0 0 6.8 0
1930 - - - -
27 0815 12.5 0 4.8 -
1930 18.3 - 5.2 -
28 0930 14.8 - 4.2 -
1630 18.6 1.7 4.5 -
29 1230 14.8 0 6.2 0

(continued)

TABLE 2. (continued)

Site Sept. Dissolved Total Residual
Date Time Temperature Sal nity Ot 0 n Chlorine
(OC) (ORO) (Ppm)
Campbell 25 1036 - 2 -
Upstream 26 1110 18.0 0 3.4 0
1630 21.0 1 12.5 <0.01
27 1015 19.0 <1 10.6 <0.02
1600 20.3 - 11.7 -
28 1015 18.9 0 9.9 <0.01
1330 19.8 <1 13.2 <0.02
29 1315 13.7 3 13.6 <0.01

Campbell 2 5 0945 - - - -
Downstream 26 1000 19.3 1+ 2.3 <0.05
1600 21.5 2 9.5 <0.02
27 1115 19.5 <1 9.8 <0.01
1700 21.0 2 8.4 <0.04
28 1100 18.6 <1 9.3 <0.01
1415 20.2 2 9.8 0
29 1330 19.2 3 12.5 0.01

Sewage Plant 25 1510 - 2 - -
1800 19.0 - 4 -
26 1500 19.3 2.5 6.5 0.45
27 1300 18.5 2 5.8 <0.02
28 1230 17.8 1 5.8 <0.01
29 1430 17.5 4 8.7 0.12

RESULTS

Oyster Experiments

Mortality over the period was low (see Table 3). The only
station showing mortality noticeably higher than that of the
control station is Station 10. This is the most upriver station
and subjectively seemed initially to have lower concentrations
of fouling organisms and more silt. The single event of six
dying, between 2 June and 14 June, may.be related to heavy runoff
and smothering by silt.

TABLE 3. MORTALITY OF OYSTERS DURING THIS STUDY

Station

2 3 4 5 6 7 8 9 10 Control

23 May 0 0 0 0 0 0 0 2

2 June 1 0 0 1 0 1 1

14 June 0 0 0 0 0 1 0 6 0

27 June 0 1 0 it 1 0 0 0

12 July 0 0 0 1 0@ 0 0 0

5 August 0 1 1 1 1 1 1 0

Total 1 1 1 2 2 3 1 8 3

Trays were not examined.

t Lines cut and tray missing.

Tampered with but appeared intact

In late spring and early summer, growth was mostly not
significantly different from that shown at the control station,
and there was no pattern of either increased.or decreased growth
(see Table 4). In c7uly, all stations showed significantly slower
growth than the controls while in August four out of seven sta-
tions showed significantly slower growth.

TABLE 4. COMPARISON OF MEAN NEW GROWTH IN OYSTERS PLANTED IN CHESTER RIVER*

Station
2 3 4 5 6 7 8 9 10 Control
23 May 6.25 5.88 5.75 3.10 4.29 5.08 5.82 6.10
(N.S.) (N.S.) (N.S.) (0.001) (0.01) (N.S.) (N.S.)

June 3.75 4.20 4.67 4.17 3.92 3.73 2.83
(N.S.) (N.S.) (0.02) (N.S.) (N.S.) (N.S.)

14 June 4.33 3.50 5.50 4.08 5.92 4.58 3.15 4.42 4.33
(N.S.) (N.S.) (N.S.) (N.S.) (0.01) (N.S.) (N.S.) (N.S.)

27 June 3.17 4.00 6.00 4.67 6.92 5.50 6.00
(N.S.) (N.S.) (N.S.) (N.S.) (N.S.) (N.S.)

@12 July 4.50 5.83 4.33 6.04 5.67 5.50 3.25 8.17
10.001) (0.01) (0.001) (0.01) (0.05) (0.001) (0-001)
Ln %

5August 4.20 9.08 7.92 6.17 6.83 7.42 1.42 8.44
(0.001) (N.S.) (N.S.) (0.01) (0.02) (N.S.) (0.001)

Means are expressed in mm. Numbers in parentheses are probabilities that these
means do not differ significantly from control means. N.S. indicates probability
values larger than 0.05.

There were few significant differences from the controls in
observed condition scores (see Table 5), however there is a pat-
tern that lower salinity stations accounted for most of these
differences. In all cases but two, where significant differences
occurred, the average condition scores were higher than that of
the controls. These aberrant stations are Stations 9 and 10
which are the two most upriver stations.

Condition scores are affected by the seasonal condition
changes that normally occur in oysters and only partially reflect
(except in extremes) the oyster's health. Condition scores are
based on the amount of solid color and how well the oyster fills
its shell. The buildup of gonad in the spring and the accumu-
lation of glycogen in the fall tend to elevate the score. Con-
versely, low scores should be expected at the end of summer when
oysters are completely spawned out and have suffered through
unfavorable high summer temperatures that adversely affected
their pumping and feeding. Also, oysters that have come through
a hard long winter and spring with low temperatures and low
available food have used up a great deal of their reserves (gly-
cogen) and therefore will score low. Still the health of these
oysters would not be as different as the two separate scores
seem to indicate. Salinity can also influence oyster condition.
oysters in low salinity upriver areas that often warm up early
will start their gonad development earlier than oysters in the
more saline downriver or bay areas that warm up slower. Freshets
often cause upriver salinities that are too low for spawning at
spawning temperature. These oysters will not spawn and in the
fall convert their gonad material immediately into glycogen with-
out going through a "summer slump." These oysters may score high
all year round. However, during the same period the oysters
downriver are likely to spawn because the salinity is favorable
at spawning temperatures and enter the fall in low condition.
After temperatures drop (usually later than upstream), they will
feed effectively again and increase their score until the water
temperature is 5'C when they cease feeding and rely on their re-
serve food which decreases their score. Thus, oyster scores
should be evaluated with season and location in mind using as
many individuals as possible. It is a subjective comparative
judgment and should be performed as much as possible by one
person to eliminate differences between workers.

Associated organisms and time of occurrence are given in
Table 6. When trays were inspected on 12 July, there were no
live commensal organisms present at Station 9, and at Station
10 the shells were almost clean except for a few live barnacles,
bryozoan colonies and amphipods. During inspection of oysters
on 5 August, it was noted that very few commensal organisms were
present at Stations 6, 7, 8, 9 and 10 and that the shells were
cleaner than before. On 23 August, it was noted that algae and
barnacles were dying and decomposing at Stations 6 and 7, and

TABLE 5. COMPARISON OF MEAN OBSERVED CONDITION SCORES IN OYSTERS PLANTED

IN CHESTER RIVER*

Station

2 3 4 6 8 9 10 Control
23 May .3.33 3.33 3.01 3.07 3.11 3.25 3.46 3.31
(N.S.) (N.S.) (N.S.) (N.S.) (N.S.) (N.S.) (N.S.)

2 June 3.18 3.20 3.68 3.38 3.23 3.18 3.28
(N.S.) (N.S.) (0.05) (N.S.) (N.S.) (N.S.)

14 June 3.79 3.68 3.65 4.34 4.33 3.85 +2.99 3.46
(N.S.) (N.S.) (N.S.) (0.001) (0.001) (N.S.) (0-01)

27 June 3.55 3.72 3.61 3.74 3.83 4.27 3.56
(N.S.) (N.S.) (N.S.) (N.S.) (N.S.) (0.01)

12 July 3.58 3.60 3.73 3.63 4.78 -t-3.21 3.95 3.74
(N.S.) (N.S.) (N.S.) (N.S.) (0.01) (0.05) (N.S.)

5 August 4.06 .3.42 3.67 4.74 4.75 4.83 4.68 3.49
(N.S.) (N.S.) (N.S.) (0.001) (0.001) (0.001) (0-001)

Numbers in parentheses are probabilities that these means do not differ significant-
ly from control means. N.S. indicates probability values larger than 0.05.

Values significantly lower than control value.

M M MON M Mao mom MIM M M M M

TABLE 6. ASSOCIATED ORGANISMS BY STATION

Station
Organisms 2 3 4 5 6
Filamentous green algae 1* 2,3,4 3,7
Sea lettuce 1,4 3,7

Brown algae
Bryozoans 3,6,8 3 4,5,6,8
Hydroids
Nereid polychaetes 6,7 6 6

Mussels
(Brachyodontis recurvis) 5,6 6,8 4,5,6,7
Gammarid amphipods 2 1,2 2,4,5 3 4,7
Corophid amphipods 4,5,6,8 3,6,8 5,6
CO Grass shrimp
(Palaeomonetes sp.) 3 4,5,6

Xanthid crab
(Rhithropanopeus harrisii) 5,6,7,8 4,5,6 3 4,5,6,8

Blue crab
(Callinectes sapidus) 5,6 4,5
Barnacles 4 3,4,5,6,7,8 3,4,6 3 3,6,7,8
Naked goby
(Gobiosoma bosci) 6

(continued)

TABLE 6. (continued)
Station
organisms 7 8 9 io,
Filamentous green algae

Sea lettuce
Brown algae 3,7 3,4
Bryozoans 3,5,8 4,5,6,8 3 3,5,6,8
Hydroids 4 4
Nereid polychaet.es 3,5 7

Mussels
(Brachyodontis recurvis) 6,7,8 6,7,8 7,8 7,8
Gammarid amphipods 4,5,7 3,4,7 2,3,7,8 2,3,5,8
Corophid amphipods 6 6 6
Grass shrimp
(Palaeomonetes sp.) 3,5

Xanthid crab
(Rhithropanopeus harrisii) 3,5,6,8 1,2,3,4,5,6,7 1,2,3,6,7,8 2,6,7,8

Blue crab
(Callinectes sapidus)
Barnacles 3,5,6,7,8 5,6,7,8 2,6,7,8 5,6,8
Naked goby
(Gobiosoma bosci) 6

(continued)

mow MMOOMMEOM M"OWWWOMURMM@

TABLE 6. (continued)

Comments
Station 2. Very few associated organisms throughout study; lost, not sampled
after 27 June.

Station 3. Very badly fouled with algae at inspections 1-4; inspected on all
dates.

Station 4. Inspected on all dates.

Station 5. Not inspected on trips 1 and 2 because of unfavorable tidal currents,
lost to vandals after inspection trip 3.

Station 6. Noticeably less fouled at inspection 6, barnacles and algae dying and
decomposing at inspection 7.

Station 7. Not inspected on trips 1 and 2 because of unfavorable tidal currents;
noticeably less fouled at inspections 6 and 7, at insepction 7 algae
dying and decomposing.

Station 8. Noticeably less fouled at inspections 6 and 7.

Station 9. At inspection 5 the shells were clean, few fouling organisms at in-
spections 6 and 7j recovered in density at inspection 8; not inspected
on trip 4.

Station 10. Almost clean shells at inspections 5, 6, 7; fouling community dense at
inspection 8; not inspected on trip 4.

Inspection trips
1 = 23 May 3 = 14 June 5 = 12 July 7 = 23 August
2 = 2 June 4 = 27 June 6 = 5 August 8 = 18 September

and fouling communities were still sparse at Stations 6, 7, 8, 9
and 10. By'18 September fouling communities seemed back to nor-
mal at all stations inspected; that is, all stations but 2 and 5.

Golden Shiner Experiment

Mortality was high at all stations including the controls.
Figure 3 is a graphic representation of mortality over time.
Excess mortality, that is mortality of controls minus mortality
of experiments, is shown in Figure 4. Ranked in descending order
of total survivorship the stations are Tenneco Downstream and
Control, with Tenneco Pond, Campbell Upstream, and Campbell Down-
stream all having the same survivorship and with The Sewage Plant
having the lowest survivorship. Sewage Plant station is unique
in the excess mortality curves in that all nonzero values are
negative, that is mortality was higher than control mortality at
all times.

30
CAMPBELL
DOWNSTREAM

2 4

DAYS

SEWAGE
PLANT

20-

DAYS

Figure 3. Mortality of golden shiners, by station, over the
96-hour observation period.

30. TENNECO 30 TENNECO
DOWNSTREAM POND

20 20

1D la

1 2 3 4 5 'one 1 2 3 4 5
DAYS > DAYS
cc >
c
30 3G CONTROL
CAMPBELL
UPSTREAM

2a 20

0,
1 2 3 4 5 1 2 3 4 5
DAYS DAYS
TENNEC
01 S
ON TOREAM

@AMIE N CONTROL
P' LL
UPSTREAM

Figure 3. (Continued)

TENNECO POND
5

-10

t 5 TENNECO DOWNSTREAM

-5

-10

5 CAMPBELL UPSTREAM

-5
-101
10

CAMPBELL
5 DOWNSTREAM

0
-101
IU.

SEWAGE PLANT

X
0

-5

-10
41

DAYS
Figure 4. Excess mortality of golden.shiners, by station, during
96 hours of observation. This is calculated by sub-
tracting mortality of experimentals from that of the
controls. Negative values indicate mortality more
than that of the control.

Autopsy data showed the following to be significantly cor-
related with death at one or more stations (see Table 7 for a
presentation of these data); fungus on body or fins; watery fat
in the mesentery and on the surface of the intestines; foreign
material on the gill surface; and hard, white crystalline nodules
attached to the body or fins. Comparing stations for frequency
of occurrence of each of these characteristics using nonsurvivors
only showed that no location had significantly higher or lower
frequency of watery mesenteric fat; Campbell Upstream and Camp-
bell Downstream had significantly higher frequency of occurrence
of fungus; the Control Station showed a significantly higher fre-
quency of foreign material (apparently silt) on the gills and
Sewage plant, Tenneco Pond and Campbell Downstream showed sig7
nificantly higher frequency of white nodules. See Table 8 for
levels of significance of these characteristics.

TABLE 7. RELATIONSHIP OF NONSURVIVAL TO AUTOPSY DATA*

Foreign Watery
White material mesenteric
Station Fungus nodules on gills fat

Control N.C.t N.C.-Il- 0.05 0.05

Campbell Upstream 0.005 N.C.-I- N.S. N.S.

Campbell Downstream 0.005 N.S. N.S. N.S.

Tenneco Pond N.S. N.S. 0.05 N.S.

Tenneco Downstream 0.05 N.S. N.S. N.S.

Sewage Plant N.S. 0.01 N.S. 0.01

Numbers are probability values calculated using contingency
analysis. N.S. indicates that probability was greater than
0.5.

Not calculable because of zero expected values.

TABLE 8. COMPARISONS BETWEEN STATIONS FOR CHARACTERS
RELATED TO NONSURVIVAL*

Campbell Campbell Tenneco Tenneco
Control Upstream Downstream Pond Downstream
Campbell Upstream 0.005 Fungus
Campbell Downstream 0.005 N.S.
Tenneco Pond N.S. 0.005 0.005

Tenneco Downstream 0.025 0.025 0.005 N.S.
Sewage Plant N.S. 0.005 0.005 N.S. N.S.

Foreign material on gills
Campbell Upstream 0.05
Campbell Downstream 0.005 N.S.
Tenneco Pond 0.005 N.S. N.S.

Tenneco Downstream 0.005 N.S. N.S. N.S.
Sewage Plant 0.005 0.05 N.S. N.S. N.S.

White nodules

Campbell Upstream N.S.
Campbell Downstream 0.025 0.025
Tenneco Pond 0.005 0.005 N.S.
Tenneco Downstream N.S. N.S. N.S. N.S. N.S.
Sewage Plant 0.005 0.005 0.005 0.025 0.005

Watery mesenteric fat
Campbell Upstream N.S.
Campbell Downstream N.S. N.S.
Tenneco Pond N.S. N.S. N.S.

Tenneco Downstream N.S. N.S. N.S. N.S.
Sewage Plant N.S. N.S. N.S. N.S. N.S.

Numbers are probability values calculated using contingency
analysis. N.S. indicates that probability was greater than
0.05.

Crayfish Experiments

No significant mortality was seen at any station, and
autopsies failed to disclose any obvious difference that could be
related to treatment between the Crayfish that died during the
experiments and those that survived.

DISCUSSION

Oyster Experiments

Since one of the principal objectives of these studies was to
identify possible causes of oyster mortality in the Chester River
since 1974, it is unfortunate for the study that no large-scale
mortality occurred at any of our stations. Observations by sci-
entists, watermen and representatives of the Department of
Natural Resources have noted that oysters that die in the "mor-
tality areas" of the Chester River have unusually clean shells
showing little or no fouling. We did note a dieoff of fouling
organisms. This dieoff was first noted at the two most upriver
stations on 12 July and coincided with a fishkill in the river.
Our personnel noted that the fishkill was a widespread phenomenon
with dead fish, mostly catfish and carp occurring from near the
mouth of the river to well above the mouth of Morgan Creek. It
might be speculated that the two phenomena are related in some
manner.

The dieoff of fouling organisms was never detected downstream
below Station 6 (clamline buoy A off Corsica Neck) and seeming
full recovery was noted by 18 September. An examination of Table
3 will show an interesting pattern of deaths. If we ignore the
unusual high mortality at Station 10 on 14 June, which may be re-
lated to high runoff and silt, 6 of the 13 remaining mortalities
occurred during the period of the fouling organism dieoff and at
the stations where commensal organism dieoff was noted. It may
be that the oyster dieoff did occur during the study, but for
some reason this was an unusually mild year for it.

An examination of Table 4 shows that summer growth rates of
our oysters placed in the Chester River were significantly lower
than that of the control oysters held in the Patuxent River.
This redu'ced growth rate coincides with the period of time when
the dieoff of fouling organisms occurred.

Table 5 shows that oysters in the four most upriver stations
had significantly higher observed conditions scores. These may
well relate to salinities being too low for spawning. These
oysters may simply have never lost gonadal material or glycogen.

one possible hypothesis for the dieoff has been suggested by
Donald Heinle (personal communication). He suggests that heavy
loading of the system with organic material will cause the water

3.6

to go anoxic or nearly anoxic at depth, and under such conditions
carbonic acid is produced which will dissolve the outer layer of
the oyster shell; thus, producing the "scrubbed clean" look of
the oysters. He has seen this occur in estuarine waters and in
the laboratory despite the heavy buffering of the salts present.
If in addition to producing carbonic acid, anoxic conditions hold
for a long period when the oysters were otherwise stressed, then
mortality would occur. This hypothesis is consistent with the
data and cannot be disproved as our sampling periods were far
enough apart that we might not have detected the event causing
the anoxia.

Golden Shiner Experiments

Table 7 indicates that different stations show different cor-
relations between nonsurvivorship and four factors noted during
autopsies. Both Campbell stations and Tenneco Downstream had
significantly higher infection percentages of fungus. The fungus
appeared to be Saprolegnia sp. but was not identified.
Saprolengia is not an obligate parasite. Usually it occurs as a
saprophyte. Its abundance at these stations indicates high con-
centrations of organic material.

The white nodules are problematic. Their identification is
unknown. They are white, translucent, hard, brittle and sub-
spherical to ovoidal to rounded irregular in shape. Largest
diameter is less than 1.5 mm. They occur in patches usually on
the ventral half of the body or on the fins. They are firmly
attached to the scales, fin rays or opercle but can be broken
loose by scraping. None were noted in specimens which had sur-
vived nor in any specimen from Campbell Upstream or the control
station. It was thought that these might be the results of an
interaction between the calcium of the boney scales and fin rays
and some ingredient in Davidson's fixative. Arguments against
this are numerous. The crayfish have as much or more calcium in
their exoskeletons than the fish scales, but no crayfish was ever.
found with one. The placement on the body in discrete patches
argues against any random process, and the fact that there were
significantly more or less of these at certain stations adds
doubt to any simple explanation.

Foreign matter on the gills irritate the gills, cause stress
symptoms and synergistically increase toxicity of other sub-
stances.* It is therefore conceivable that the foreign material
noted on the gills of fishes that died at the control station
was a contributing factor to their death. In addition to silt
particles on the gills, fibers resembling those from toilet paper
were noted on one'fish's gills. This indicates that untreated
domestic sewage may find its way into the stream above our con-
trol station.
Wilber, Charles G. 1976. The biological aspects of water pollu-
tion. Charles C. Thomas, Publisher, Springfield, IL, 296 p.

What we have referred to as "watery fat" is a type of fatty
tissue that has a flabby look and a semiliquid consistency. It
appears under low magnification to be translucent and to have
inclusions of water or other liquid free in it. This fat was
found in both surviving and dead specimens but a significant cor-
relation between it and the nonsurvivorship was noted at the con-
trol station and at the sewage plant station. No mention of this
kind of fat in the literature was found, but it was speculated
that the presence of watery fat may be an early symptom of
stress. In prolonged periods of high energy expenditure and
during starvation, body fat is mobilized to act as an energy
source. If lipids are used up, fat cells may lose their turgor
producing this soft,gelatinous texture.

The fact that dead fish from different stations have very
different patterns of correlations with these factors indicates
that there are probably multiple causes of mortality and that
each station has its own unique combination of factors. In
short, no single cause of mortality was found nor was it thought
that the mortality noted in these experiments is necessarily re-
lated to the cause of oyster mortality downstream.

Crayfish Experiments

Mortality was too low and autopsy data failed to find any
changes which could be related to treatments. It seems that this
species of crayfish is hardier than the golden shiners. The mor-
talities occurred mostly with crayfish that had moulted and that
were killed then by other crayfish.

SECTION 5

PHTHALATE ESTERS AND RELATED CHEMICALS IN THE CHESTER RIVER

INTRODUCTION

The immediate purpose of this work is to assess the potential
for bioaccumulation of alkyl phthalates as a way to explain the
recent past mortality of oysters in the Chester River. Our
starting approach involved dual plans to use both gas and liquid
chromatography. Careful technique is needed to avoid contamina-
tion by the ubiquitous presence of the phthalate esters in the
environment, including their reported presence in the chemical
laboratory. Our own facilities proved to be no exception.

Two main problems needed to be solved: positive errors due
to contamination during the measurement process, and negative
errors due to incomplete extraction. Hopefully, both problems
have been solved. Several models were considered for influx of
alkyl phthalate waste from the Tenneco plant. These considera-
tions were used to guide the sample selection and interpretation
process. In brief, our overall approach emphasized measurement
reliability, it narrowed the number of samples for analysis, and
it gave a conceptual base to guide future work. Alkyl phthalate
contamination in the Chester River clearly includes di(2-
ethylhexyl)phthalate (DEHP). Our finding of massive contamina-
tion in Tenneco Pond is consistent with the pollution conse-
quences that are consistent with the existing discharge permit
which has been approved by the State of Maryland.

Geographic and Background History

.The Tenneco factory is sited on a tributary of Morgan Creek
approximately 7 miles (10.6 km) upstream from where the creek
empties into the Chester River. The location is shown in Figure
5. The oyster mortality starts 12 miles (19.2 km) downstream
from Morgan Creek toward the mouth of the Chester River. The
mortality involved a downstream progression. The progressive
movement is evident in Figure 6, which shows the status of
mortalitities in oyster beds monitored in 1970-1975 (see
Appendix A). The progression of the total mortality zone (0
percent live oysters) is seen by comparing the 1973 and 1975
results where the mortality front is seen to advance 5.2 km in
the downstream direction.

Tenneco Plant

Morgan Creek

Chestertown hester River Brklge at Chestertowr

Figure 5. Map of Chester River illustrating sample sites.
Site coordinates are given in Table 18.

UPPER CHESTER RIVER

P1
3
10

P1.
-1973
29 7 cots ica R;.*,
35
38

39
1975
41 fow
Oyster Bed Lines
43 4A of TotW Mortaft
1973-1975

5
(data from Appendix A)
4e9
0 , M;ter ..s5000
Figure 6. Downstream progression of oyster mortality.
1. Northwest 17. Bailey 33. Town Point
2. Melton Point 18. King's Creek 34. Holton Point
3. Booker Wharf 19. Wilson's Point 35. oldfield
4. Hollyday 20. Eagle Point 36. Robin's Cove
5. Haddaway 21. Island Point 37. Chester River Mid-
6. Shippen Creek 22. Davis Creek . dleground
7. Mummy's Cove 23. Drum Point 38. Bluff Point
8. Deep Point 24. Boathouse 39. Hell's Delight
9. Sheep 25. Sand Thistle 40. Bay Bush Point
10. Commegy's Bight 26. Hudson 41. Piney Point
11. Einmory Hollow 27. Nichols 42. Belts
12. Spaniard Point 28. Limekiln 43. Durdin
13. Cliff 29. Willow Bottom 44. Horse Race
14. Ebb Point 30. Possum Point 45. Carpenter Island
15. Ware 31. Ship Point 46. Black Buoy
16. Phillip 32. Emory Wharf 47. Hail Creek
48. Hail Point
49. Poplar
(From: Meritt, D. W. 1977, Univ. Md. Ctr. Environ. Estuarine
Studies Spec. Rpt. 7.)
41

The discharge of any toxic chemical might account for the
mortality pattern. The Tenneco plant waste discharge is one
possibility based on circumstances that will be considered here.

The Tenneco plant became operational in 1959. Esterification
equipment was acquired in 1965. A process for easte treatment by
bacterial digestion was installed in 1968. At this time, a dam
was located across the water flowing from the Tenneco plant. As 2
a result, a small pond (Tenneco Pond) was formed with 40 x 103 m
(10 acres or 4 hectares) of surface area.

Hurricane Agnes occurred in May of 1972 bringing 16.0 cm of
rainfall in one day (Palmer 1972) to the plant area which was
flooded. The contents of a waste collection tank were washed
out. It was estimated that 50-100 gallons (190-380 L) of organic
waste liquid were washed out into Morgan Creek and then into the
Chester River. It is possible that the washed-our chemicals
could be causally related to the oyster mortality. (It is also
possible that the effect might be the result of one or more other
factors, including the continual discharge of unidentified organ-
ic compounds which reside after waste treatment, agricultural
runoff, a "natural phenomenon," etc.). Various pathways for the
transport of phthalate esters from the Tenneco plant to the oys-
ter mortality area are now discussed.

The Upper Bay Survey (Palmer, Schubel, and Cronin 1975)
makes it clear that sparingly soluble organic compounds are speci-
fically associated with the finely has been established with ob-
servations that the presence of such organic compounds is
precisely correlated by the octanol-water partitioning model, and
the adsorption of neutral organic compounds by sediment has been
strongly correlated to the weight fraction of organic carbon
(Karicknoff et al. 1979). This finding is consistent with the
known tendency of humic acid to adsorb such compounds rather
tenaciously. Schnitzer's work (1972) suggests possibly irrevers-
adsorption of phthalate esters by clay soils, although there
is some controversy on the significance of that finding. Regard-
less, the known distribution of neutral organic compounds in the
marine ecosystem includes biomagnification by life in the marine
environment, and a corresponding adsorptive magnification effect
into the natural clay-organic complexes present in marine sedi-
ments. Since dioctylphthalate (DOP) was observed early in the
study as a conspicuous and major component in Tenneco Pond sedi-
ment, it is important to show whether or how that material would
be transported from the Tenneco factory to the clay-rich sedi-
ments in the Chester River. These latter sediments according to
the Chester River Report (Palmer 1972) occupy the deeper terrace
portions of the river floor and the transitional channels between
nearshore and the main river channel. Taken in perspective, these
are the most critical samples where investigation can be focused.
Such sediments are widely recognized for their ability to adsorb
compounds for long periods of time, geological epoch in many

instances. As a result, the clay-humic structure contains a
historical reservoir of many compounds deposited through natural
or anthropogenic influx. In the present study where no biologi-
cal specimens were preserved, the key to the retrospective anal-
ysis of the past oyster mortality would have to told by the
sediment composition.

DISTRIBUTION PATTERNS

Three factors which may modify the distribution patterns of
organic distribution in the Chester River as as follows.

a. The downstream dilution effect. The width of the
Chester River grows from roughly 0.5 km at Chestertown to 1.6 km
near the midway region close to Spaniard Point to 5 km at the
mouth. This suggests that the sediments, behaving as a pollu-
tant sink, may become more dilute in the pollutant as one proceeds
downstream. In the initial pollution history of a tidal river,
this would would be more likely than later on when the accumula-
tive pollution may build toward a uniform pollutant concentration.

b. Uniform recirculation model. The Chester River, as a
tributary estuary, is moderately stratified into two layers as
described by Pritchard (1967). The net advective surface water
flow is in the upstream direction while that in deeper water is
downstream. Suspended or dissolved matter is dispersed in both
directions as a result of two processes: (1) diffusion of the
dissolved solute across the thermocline so that it is spread into
layers whose net motions are in opposite directions. It is likely
that the nonpolar plasticizers will quickly (2) sorb onto su-
spended sediment and then settle down to the leptopellic layer on
the sediment surface, and then either deposit or become resuspend-
ed into a circulation cycle. These circula-processes may cause
wide mixing, motably upstream as well as downstream.

C. "Hot spots." As noted in Han's dye tracer experiments
(1972), a pollution plume is apt to remain intact during the first
few tidal cycles. That pattern was shown to exhibit selective
localization effects. This would produce the opportunity for hot
spots where an elebated concentration of pollutant may deviate
significantly from the model based on complete mixing.

PROBLEM STATEMENT

It seems fundamental that environmental influx of toxic
pollutants is feasible for study when the pollutant is identified,
or at least limited to be among a group of candidates. In the
present study, this line of first approach was blocked from the
beginning because none of the affected oyster tissue had been
saved. The circumstances are such that the toxic agent (if there
was any) may have been introduced, reached toxic levels, and re-
turned afterwards to pretoxic conditions. Although alkyl

phthalates are readily degraded by bacteria, photolysis, and
hydrolysis, it is our belief that the known persistence of alkyl
phthalates in the marine environment is the result of rapid sorp-
tion by suspended clay-rich matter which eventually settles
through sedimentation. Once sorbed, the rate of degradation is
indefinite, but it is slow enough that the phthalates are con-
sidered "persistent." In view of the rapid bioturbation process,
surficial sediments (upper 10 cm) are quite likely to carry
memory effects due to accumulative pollution during the previous
decade.

The difficult part in the present study was to develop
reliable methods of measurement since it was felt that none of
the initially available measurement technology was reliable
enough or accurate enough for the purposes of this study.

The principal focus in the work to be presented next is
based on the significance of the clay-rich sediment regions be-
low a 2-8 m water column. These should contain the residual
compounds that may help to understand whether threatening or
tpxoc plasticizer concentrations exist now, or were likely within
the past decade. Since sedimentation occurs at an estimated rate
of 1 cm per year, our efforts were focused on values in the top
10 cm.

Based on preliminary studies we initially picked sampling
sites included apex of the mortality. The sites were chosen as
likely to show whether or not the alkyl phthalates could have
caused the oyster mortality. The first region is Tenneco Pond
where the sediment composition was found to be highly polluted.
The second region involves a series of five samples taken at the
apex of the oyster mortality. That is where the downstream
mortality movement was not clearly in evidence. The third region
is at the mouth of the Chester River. That region is likely to
contain a representative mixture of pollutants from the upper
Chesapeake Bay. Since this region is a reference point for pos-
sible upstream river pollution, we used a homogenized sample of
the river mouth sediment "R" for reference purposes and for in-
dependent analysis.

Efforts were also made toward development of methods for
analysis of oyster tissue. We underestimated the problems here
so that aspect of the study was the last to be completed. These
results will be presented later in this report.

EXPERIMENTAL PROCEDURES

Trace Analysis Techniques

The key goals in trace organic analysis of sediment are to
achieve complete extraction, to avoid loss, to prevent contami-
and to avoid interferences. In preliminary experiments we found

that ultrasonic extraction of dry Tenneco Pond sediment with
methanol gave easily measured 'quantities of DOP. Later, tetra-
hydrofuran was found to give a higher yield. Then, a series of
extracting solvents was tested comparatively and dichloromethane
was found to give the highest extractive efficiency and relative-
ly short times were required. The duration of the ultrasonic
agitation was varied until it became certain that longer extrac-
tion periods did not give greater extractive yield.

The true extractive efficiency is not subject to direct
measurement. Spiked addition of analyte produces a different
sorbed state than the analyte in the sample matrix. To assess
the apparent accuracy, two samples were selected for independent
measurement as will be described. The approach was to do what
could be done to show that re-extraction by baried methods did
not lead to a higher yield.

A major problem of contamination arises when measuring
traces of plasticizers. The environment is widely contaminated
by these substances, and the chemical laboratory is no exception.
Their presence has been reported in bottle cap liners, solvents,
extraction thimbles, preconcentrating resins and adsorbents,
aluminum foil, glass, wood, air and pipet fillers (de Zeeuw
1975, Singmaster et al. 1976, Webster and Nickless 1976).

It is obvious that the reduced contamination can be achieved
by deliberately minimizing the overall exposure to the sources
of contamination by minimizing solvent volume, the number of
manipulations, avoiding contact with any plastic materials, and
ultracleaning of apparatus. This can be summarized in two
statements: the sampling and processing apparatus must be un-
contaminated, and the amount of methodology must.be kept to an
essential minimum.

Although these aspects led us away from available methodo-
logy (Giam et al. 1976), the end of the study showed that our
method did come close to similar techniques by Dr. William
Budde (EPA-Cincinnati). The procedure used by Budde (ultra-
sonication, GCMS-selected ion monitoring) differs mainly in the
technique used for drying the sample--direct addition of Na2SO4
to partly dried sediment, Our method uses the same desiccant
but the drying offici@ncy im qrodtar. Roqard1-PPo4# th@ two 4p-
proaches are appropriate for interiaboratory comparigon.

The measurement approach is based on a reported (Watson 1976)
application of GCMS where the MS section is used in the selected
ion monitoring mode (SIM). This provides an impressive gain in
selectivity since only the predominant fragment ion masses are
monitored. As a result there is a corresponding boost in
instrumental sensitivity. The technique uses instrumental de-
tection selectivity to replace the lesser certainty of multiple
extractive and separation conditions (Giam et al. 1976).

The basis for the analytical methodology can be seen in the
following way. The upper curve in Figure 7 is a gas chromatogram
of a dichloromethane extract of a dried sediment sample from the
Chester River mouth. This is the working standard, sample "R,"
located in Figure 5. The omission of any cleanup-procedure
creates an unresolved bunching of chromatographic peaks due to
the many thousands of compounds that are present in the extract.
However, when the mass spectrometer is used as GC analyzer, an
extraordinary boost to the selectivity of the combined instru-
mentation is realized. This is shown for mass 149 as the lower
curve in Figure 7. The integration of m/z 149 intensity to
measure alkyl phthalates was confirmed by comparing integrated
area ratios for confirmational fragment ion masses for DBP
(m/z 2-5,223) and for DEHP (m/z 167,279). This wa8 done for con-
firmational purposes to show that no interferences were present
at the levels being investigated.

It has been concluded that there is no ambiguity in the pre-
sent dual use of multiple ion monitoring to identify and measure
DBP and DEHP in a single GCMS-SEM experiment. This is a result
of the extreme molecular selectivity, and correspondingly justi-
fied methodological simplication. The generic term DOP (dioctyl-
phthalate) will be used to refer to one or both of the isomers,
DNOP (di-n-octylphthalate) and DEHP (di-2-ethylhexylphthalte).
Distinction between these substances was obtained later in the
study (see Figure 14)..

Total Ion Current

DOP

DBP Current of
m/z 149 Ion

3 6 9 12 15 18 21 24 27 30
Time (minutes)
LCurrIent of
m/z 4 Ion

Figure 7. GCMS of Chester River mouth sediment extract.

Cleaning Procedure

Scrupulous cleaning procedures were used to minimize organic
contamination of the material used in this work. The details
are given in the following.

Glassware (Sample bottles, vials, lab glassware, all
glass materials)

1. Wash with Aquanox detergent.

2. Rinse with tap water followed by filtered,
distilled water.

3. Bring to annealing temperature of the glass
(4500C) for at least 1 hour.

4. Slowly cool (>4 hours).

5. Cover exposed areas with baked aluminum foil.

6. Cover with clean caps or lids.

Tools (Forceps, trowels, spatulas, etc.)

1. Clean with Aquanox.

2. Rinse with charcoal filtered water.

3. Dry in air.

4. Wrap in aluminum foil.

Piston Core Liner (Brass)

1. Scrub with Aquanox detergent using brush
fixed on ram-rod.

2. Rinse with tap water.

3. Add chromic acid/sulfuric acid cleaning
solution to etch surfaces: tube is in-
verted several times.

4. Rinse with tap water.

5. Inspect for bright shiny inner surface.

6. In field, tube was rinsed by brush and
rinsed between cores using the on-board
water supply.

Bottle cap liners (Teflon)

1. These liners were cut from Teflon sheet using
cork bores. They were then extracted for
24 hours using Soxhlet extraction.

2. Liners were stored in an air-tight, cleaned
glass bottle with an aluminum foil liner.

Solvents

1. Water was purified by passing in succession
through activated carbon filter, XAD-2 column
(Chesler et al. 1976) and final ultrafilter.

2. All other solvents--dichloromethane, MeOH,
.etc.--were distilled in all-glass distilla-
tion apparatus. Tetrahydrofuran was di-
stilled from freshly out sodium, an important
recommended procedure that prevents buildup
of explosive peroxides.

Septa (GC-Injection Port)

1. Septa with high thermal stability were used:
Thermogreen (TM) LB-1 Septum, SUPELCO, INC.

2. The septa were stored in capped organic-free
containers until use. They were inserted
into the injection port using cleaned tools.

This procedure made it unnecessary to precondition the
septa after installation in the GC or GCMS.

Field Sampling Procedures

Water Extraction using Sep Pak Cartridge

1. Luer-type syringe was fitted with a C18 9ep Pak
(Waters Associates).

2. A clean glass jar was immersed to obtain a
subsurface water sample.

3. This water was transferred into the barrel of
the syringe. The plunger was inserted and the
water was pushed through the C18 cartridge.

4. Step 3 was repeated until the back pressure
buildup, due to filtered particulates plugging
the cartridge, prevented further concentration

of organics on the C18 material. The total water
volume was noted, usually 120-150 ml.

5. The cartridge was returned to its original foil
packet. The foil was bent to give partial seal-
ing protection. The collected samples were
stored in a clean glass bottle that was sealed
until analysis was ready to be performed.

6. Trials with varied amounts of isopropanol,
methanol and tetrahyd'rofuran showed that 75-80
percent recovery was obtained following desorp-
tion with at least 1.5 ml of tetrahydrofuran;
2.5 ml was used in the actual procedure.

7. 1.7 ml of water was added to the extract from
(6) to match the liquid composition to the
initial carrier composition used in liquid
chromatography.

Sediment---
1. Tenneco Pond and contiguous creek sediments.
A small grab sampler was used to take surface
samples. The sampler was opened and the sam-
ple was discharged into a scrubbed galvanized
metal bucket. A clean metal scoop was used
to transfer the moist sediment to clean glass
jars.

2. Chester River. Sampling sites were selected
near the main channel in order to obtain clay-
rich specimens. A Van Veen grab sampler was
used to bring up surface sediment samples.
The sampler was fitted with a sliding panel
to permit insertion of the coring tube. The
clean brass coring tube was inserted to get
a 10 cm vertical core of the uppermost sedi-
ment. The brass tube was sealed by placing
a hand at the top. Then the cylindrical core
was lifted out of the grab sampler. The core
was released onto baked aluminum foil, and
then stored in the foil in a clean glass con-
tainer. The latter was stored for library
purposes.

3. All sediment samples were transferred within
12 hours to storage at 40C until ready for
drying.

4. All sediment samples used in this study were
subjected to the following drying procedure.

Sediment Drying by Isopiestic Dehydration

a. The collected sediment is homogenized in the sample
container using a clean metal spatula.

b. A 10 g sample is spread evently onto the surface of a
clean 12 cm watch glass using the same spatula.

c. An organic-free desiccator is charged with 0.25 kg
Drierite in the desiccator at 230C for 48 hours.

d. To complete the drying, the sediment is removed to
repeat the drying process. The desiccant is re-
dried by baking in the-desiccator at 2500C for
12 hours. The partly dried sediment is then re-
turned to the cooled freshly baked desiccant for
another 48 hours.

e. The dry sediment is scraped into a clean glass
mortar. It is then pulverized using a clean glass
pestle, transferred to labeled vials, and stored
at 4 OC.

Two solid substances will undergo a gas phase transfer of
water until a final equilibrium vapor pressure is reached. The
repeated exposure of moist sediment to repeated fresh charges
of the desiccant Na2 so4causes isopiestic dehydration. The
advantages of this approach are as follows:

1. Water content in clay sediment is reduced from about
60 percent to about 1 to 2 percent on a dry weight
basis.

2. Unlike the procedure of mixing the desiccant in with
the sediment (Bulla, personal communication), the
water is physically removed and there is less oppor-
tunity for organic contamination from the desiccant.

3. Unlike lyophilization, isopiestic dehydration is
carried out at atmospheric pressure with 103-fold
shorter gas phase collision distances and a corre-
sponding less likelihood for sublimative loss of
the measured analytes (DBP or DEHP).

Sediment Extraction

From Karrickhoff's work (1979), it is known that organic
pollutants in sediment are associated with the organic carbon
content, due principally to humic and other associated polymers.
Based on this, the sorptive structure should be modeled effec-
tively by n-octanol. Accordingly, a volative extraction-salvent,
similar to polarity to octanol, was chosen as more likely to

give efficient extraction. since water polarity is much higher
than that of octanol, the partitioning model clearly infers the
need to remove water before attempting the extraction of neutral
organic compounds.

This approach differs considerably from the conventional
Soxhlet extraction procedure applied to wet sediment. The lat-
ter also requires large solvent to sample ratios so an extra
burden is implied in terms of needed solvent purity. We were
advised by Dr. George Boughman (personal communication) that
Soxhlet extraction tends to be incomplete due to slow diffusion
of solvent through the sediment sample. For that reason, a thin
coating of sediment is advised to line the extraction thimble.
Unfortunately, this procedure boosts the solvent-to-sample ratio
even further.

Sediment Extraction with Methylene Chloride

1. Two grams of sediment is weighed to the nearest mg
and placed in 10 ml capacity vials with Teflon
lined caps. The caps are screwed on tightly to
prevent evaporation.

2. Five ml of methylene chloride, containing 0.40 ppm
dimethoxyethyl phthalate (DMEP) as internal standard,
is introduced using a Repipet apparatus. (A different
amount of DMEP was used for the Tenneco Pong sample.
This involved 4.0 ppm DMEP in CH2Cl2, but there was no
subsequent evaporation.)

3. The vials are first agitated in a Vari-whirl mixer to
suspend the particles.

4. The vials are placed in an ultrasonic bath. Circu-
lating water is used to hold a temperature of ap-
proximately 30'C for 2 minutes.

5. The vials are centrifuged at 2,500 G for 15 min.
This is done to remove suspended particles.

6. The supernatant liquid is drawn off and placed in
a 5 ml capacity vial using a Pasteur pipet.

7. The extract is evaporated under a stream of purified
nitrogen to a volume of approximately 0.5 ml. The
nitrogen is purified by passing through a column of
carefully extracted washed (Chester et al. 1976)
XAD-2 resin.

8. Two hundred pl of isooctane is added to the vials to
prevent evaporation to dryness. The extract is blown
down to 0.20 pl. This gives an overall 25 x concen-
tration (5 ml to 0.2 ml).

9. A procedural blank was prepared in triplicate for
each experiment. Each contains the extracting
solvent and internal standard but no sediment
samples. Isooctane is added after the blanks
have been sonicated and blown down. For LC
analysis, tetrahydrofuran was substituted for
the isooctane. These procedures are summarized
in Tables 9 and 10.

TABLE 9. FIELD SAMPLE HANDLING

Procedure Precleaning

Surface Grab on-board
Sampler High Pressure
Water Hose

Obtain Core within Use brass tube precleaned
Grab Sampler by acid etch, soap, water.
Swabbed with abrasive
between samples.

Storage Use baked glass bottles,
aluminum cap liners.
Hold for 14 days at 4*C.

Isopiestic Two 48-hour periods.
Dehydration Closed system

Sample Sample grinding with
Homogeneity baked glass mortar and
pestle.

Storage Dried sample stored at
40c.

TABLE 10. METHODOLOGY

Step Procedure Reagents Added Conditions
1 Isopiestic dehydration Na2SO dried Closed system
of sediment at 1880C 220c, 1 atm.

2 Ultrasonic extraction CH2C12 contain- Closed system
with CH 2C12 ing 0.4 or 4.0 t = 30*C
ppm DMEP internal
standard

3 Centrifugation None Closed system
t = 100C
2500 x G

4 Evaporate solvent XAD-2 filtered Open system
N
2'
(a) Isooctane
added for
GC or GCMS
(b) THF added,
CH2C12 re-
moved, for
HPLC

PreliminaryLiquid Chromatographic Analysis

High performance liquid chromatography has been reported
for analysis of alkyl phthalate esters (Hellman 1978). This
technique was adapted to confirm GC measurements on Tenneco
Pond sediment, as a screen and to provide upper limit data on
samples containing small amounts of alkyl phthalates. The
Hellman method is based on adsorption chromatography. A more
reliable method was developed based on partition chromatography
using water-tetrahydrofuran as the carrier with bonded C18 as
the stationary phase.

A surface sample of water from the environment was trans-
ferred onsite to the barrel of a baked glass Luer-type syringe.
A Sep Pak C18 cartridge was already fixed in place. The water
was then passed through the cartridge using hand pressure after
inserting the plunger into the barrel of the syringe. This was
continued until the back pressure became too great. Then the
cartridge was returned to its original protective envelope,
sealed, and labeled.

Preliminary laboratory tests had shown that THF (tetrahydro-
furan) gave sharper peaks than either methanol or isopropanol
which are widely used. The recovery of DEHP standard was about

74 percent by desorbing with 1 ml of THF, or 75-80 percent if
> 1.5 ml were used. The final procedure called for 2.5 ml of
THF. Then 1.7 ml of purified water was added so that the
liquid sample composition matched the initial HPLC carrier
make up. The liquid chromatograph 254 nm UV monitor (Waters
Model 440) remained on scale using 25 pl samples of extract.

Similar tests were carried out using DTDP (ditridecyl-
phthalate) and the results were quite similar except that a
lower recovery, approximately 60 percent, was obtained.

The precision of the HPLC results were found to be linear
with the amount of added standard to within 5 percent variabil-
ity. This was observed during Sep Pak adsorption of 10 ppm
DEHP dissolved in water samples ranging from 30 to 150 ml in
volume.

Water analysis in Tenneco Pond and several sampling sites in
the Chester River was attempted using C18-Sep Pak cartridges to
concentrate the analyte. These cartridges were soon clogged by
suspended particles so the sample volume was limited to 150 ml
or less. GCMS showed that the liquid chromatographic (LC)
technique alone in one instance gave an apparent but false
identification of ditridecylphthalate. At trace concentrations,
LC was considered valid for setting upper limits.

The final conditions for HPLC analysis were obtained using
a linear carrier gradient proceeding from 60 percent THF/40
percent H20 to 90 Percent/10 percent over a 5.0 minute period.
A flow rate of 2.0 ml/min was used.

Sediment extracts in dichloromethane were twice treated by
adding THF and using purified nitrogen gas blowdown to remove
the dichloromethane. Then makeup water was added, as before,
to give the correct liquid solvent rates.

Quantitative Analysis of Sediment Extract using GC/MS-SIM

The Hewlett Packard 5992A GC/MS system was used for the
analysis of sediment extracts. The microprocessor-controlled
5992A uses a jet separator to interface the gas chromatograph
to a hyperbolic quadrupole mass filter. A 3-foot (90 cm)
silanized glass column (packed with 3 percent SE-30 on Chromo-
sorb W-AW-DMCS) was used. The column was temperature program-
med from 140*C-2500C at 50/min. The system was later converted
to a SE-52 glass capillary column programmed from 160*C-275*C
at 7.5*/min. Up to 6 ions can be monitoring during a chromato-
graphic run in the selected ion monitoring mode. The base peak
of DEHP, m/z 149 was monitored. Mass 59 was also monitored
which is characteristic of the internal standard, dimethoxyethyl
phthalate (DMEP). DMEP has a low mass 149 abundance which de-
creases its chromatographic interference with other possible

phthalates. It is also likely to exhibit physical and chemical
properties similar to other phthalates, and it is not produced
commercially as a plasticizer. By adding DMEP to the extracting
solvent, methylene chloride, it is susceptible to the same sys-
tematic errors during the analytical scheme as other phthalates
being determined.

GC/MS Autotune Procedure

Each week "autotune" was run to adjust and to assess the
condition of the GC/MS instrument. This procedure tunes the
ion source and mass filter to produce a mass spectrum of per-
fluorotributyl amine (PFTBA) to meet certain minimum specifica-
tions. If these specifications, as recommended by the
manufacturer, were not met, the problem was diagnosed and
corrected before continuing the work.

GC/MS-SIM Calculation

The mass 149 chromatogram of all sediments taken from the
Chester River show only two well-defined peaks at GC retention
times that correspond to dibutylphthalate DBP, and dioctyl-
phthalate. Quantitative analysis for each of these two
compounds is based on monitoring mass 148 peak areas for DBP and
DEHP, and mass 59 for the internal standard. Within experimen-
tal error, we verified the linearity typical in GC/MS-SIM.
analysis. Defining the following terms
C = concentration of solute in CH 2 Cl2 (mg/ml CH2C1 2)
S - area under mass-chromatographic peak (area units)
a = analyte (DEHP or DBP)
i - internal standard (DMEP)
k = response factor

we have

S Ka and S i = Ki ci*

The ratio ka/k. was found to be constant for total analyte in-
jections less ihan 0.3 pg providing C a/Ciwere within a factor
of 100 of unity. Then, it follows that

ka Sa Ci S C.
R ki constant C S. C@!-S
a 1 a

The latter term was measured giving k a/ki = 11.0 + 0.4. it
follows that
C. S
Ca = Si Ra

In order to calculate the weight fraction f of analyte in
the sediment sample, the following relationship awas used:

fa( pg of a C a( Pg of a Ve(ml CH2C1 2)
g sediment ml CH 2 Cl2 Ws(g sediment)

This uses the volume Ve of the final concentrate extract, and
the initial weight WS of dry sediment to provide the weight of
DBP or DEHP. The units, pg/g, are the same as parts per million
(ppm).

RESULTS AND DISCUSSION

Preliminary Results with HPLC

A detailed study (Gingras 1979) was made of the basis for
applying modern liquid chromatography to analysis of field
sediment and water samples for alkyl phthalates. Previously,
Hellman (1978) had developed the use of adsorption liquid chro-
matography but this is prone to irreversible adsorption effects
and lesser reproducibility. Amundson (1978) used reversed phase
Bondapak C-18 as the stationary phase with methanol plus
LC wi
1 percent acetic acid as the carrier. More sharply defined
peaks were obtained more rapidly with tetrahydrofuran-water
mixtures, 60:40 linear programmed in a 5-minute period to 90:10
at 2.0 cm3/ml flow rate.

Analysis was made by LC bases on preliminary test samples of
phthalates as manufactured, named and provided to us by Tenneco.
These are shown in Figures B-12. The LC technique has notable
stability@but qualitative identifications can only be shown to
be consistent or inconsistent with the LC retention volumes. The
time of analysis was only 7 minutes, or less than half required
by Anumdson. The resolution of a synthetic mixture of the three
reference materials DOP, DIDP (diieodecylphthalte) and DTDP is
illustrated in Figure 9. Later on in the study we established
that Tenneco's "DOP" was actually DEHP.

Analysis of Tenneco Pond water was carried out using the Sep
Pak technique and the results are shown in Figure 10. The pre-
sence of DOP 0.25 + 0.15 ppm and DTDP (ditridecylphthalate,
1.5 + 0.2 ppm) were estimated by LC alone. The presence of
phth@Tlate esters in downstream water samples could not be de-
tected by LC so these will not be reported here. The first peak

DOP 6-10 P

ject Inject

DIDP 7-11 P
C
0
CL

cc
0

Inject Inject

0 2 4 6
Time (minutes)
DTDP

Inject

2 4 6 8
Time (minutes)

Figure 8. Liquid chromatography of Tenneco Products de-
signated by Tenneco as Alkyl Phthalates, or mixtures (6-10P
and 7-11P). L Linject
57

UV Detector Response

75*
0

0 F1
0
;5 ;5 CD
F'- 0)
x rt
rt 0
r@

(D 2)
0 @l CD
0

F-3
(D 0
0 --------

(D
0 01@
Ou

(D :J

@-3
0

LQ
UV Detector Response

rlr
Fj-0
0
rt
0
LQ
En
m DOP
0151 & DIDP
0 0 C>. DTDP

00 CO

0 rt

En
I
F-I

in Figure 10 is likely to be due to the fulvic acid component
which occurs in natural water and sediment extracts. Although a
number of Morgan Creek and Chester River water samples were run
by the same technique, the method was frustrated by early plug-
ging of the Sep Paks. This prevented the obtaining of large
enough samples to give the needed sensitivity.

Analysis was made by LC of Tenneco Pond sediment. The re-
sult is shown in Figure 11. The concentrations are so high that
peak verification by GC/MS was readily demonstrated and quanti-
tative interpretations are thus possible. The retention volumes
are consistent with the presence of humic acid, of DEHP
(1.8 + 0.1) x 103 ppm and a secondoietermination of (1.3 x + 0.2)
x 103-ppm, of DIDP (1.4 + 0.2) x 1 ppm and of DTDP
(1.9 + 0.2) x 103 ppm. The qualitative identifications by LC
were each confirmed by GC/MS-SIM retention times of m/z 149.

A few miles downstream from Tenneco on Morgan Creek sediment
samples were taken for LC analysis. Comparison was made to
humic acid under the same conditions. These results are shown
in Figures 12 and 13, respectively. It is clear that the natural
background, the multiple peaks due to fulvic and humic acid con-
stituents, seriously interfere with and prevent quantitative use
of these LC results. Sediment samples taken from the
Chestertown Bridge and from the Chester River (site 5, Figure 4)
showed no evidence for DEHP, apparently less than 1.0 ppm.
Artifact peaks were observed and these suggested exaggerated
levels of DIDP and DTDP. The GC/MS technique gave qualitative
but not quantitative confirmations of DEHP and DBP so these were
concluded as present at both sites. The DBP peak position was
found to be consistently obscured in the LC by the humic com-
ponents. Significantly, this was the only alkyl phthalate which
Tenneco did not make, but which is otherwise massively produced
by U.S. industry. Second, certain peaks evident at the equivalent
of <10 ppm levels by LC were found by GC/MS not to be phthalates.
Attention was then directed to the GC/MS technique which was con-
sidered necessary for identification and measurement of DBP and
DEHP at levels below 100 ppm. The LC technique offered prelimi-
nary qualitative utility and it was the more reliable for meas-
uring DIDP and DTDP.
Preliminary Qualitative Results Using GC and GC/MS

A preliminary grab sample of Tenneco Pond sediment was
obtained. The sediment was desiccated and then extracted by
methanol using ultrasonic agitation. Centrifugation gave a
clear filtrate which was then analyzed by gas chromatography.
The result showed a single peak which accounted for >95 percent
of all volatiles apparent by the FID detector.

Gas chromatographic retention times and mass fragmentation
patterns showed clearly that the very conspicuous main organic
volatile component in Tenneco Pond sediment is the same as the

DEHP reference sample which, in turn, matches the samples of DEHP
from Chem Service, Inc. The mass fragmentation patterns showed
closely similar abundance profiles at m/z 61, 70, 71, 83, 104,
112, 113, 149 (base), 150, 167, 168, and 279.

(L
(D

0
CL

cc
L-
0 CL
4-0
Q 0
0) a

inject

2 4
Time (minutes)

Figure 11. Liquid chromatogram of Tenneco pond sediment.

@61

Reference data from the NIH/EPA library, recently published by
the National Bureau of Standards, strengthened the assignment
of DEHP.

DEHP was subsequently distinguished from DNOP by their g.c.
retention times and mass spectra. There was no DNOP detected in
the Chester River sediment. The use of glass capillary GCMS
permitted a clear distinction (see P. 69).

Tests of the Methodology Based on GC/MS

A series of experiments were carried out in order to
validate the method. Preliminary experiments showed that de-
sorption of DEHP or DBP occurred more efficiently from dry
rather than wet sediment. Very few wet extractions were
carried out thereafter.

Rate of Extraction---
A series of extractions by dichloromethane of the dry work-
ing standard "R" (see Figure 7) were carried out. The DBP
extraction was independent of the time given to the sonication
step which was varied from 30 seconds to 4 hours. The results
are shown in Table 11. The results show no apparent trend with
time of sonication. Other experiments confirmed the finding

TABLE 11. EFFECT OF SONICATION ON DBP EXTRACTION BY CH 2C12
FROM SAMPLE "R"

Sonication DBP Measurement
Time (min) (ppm )*

0.5 0.78!@

5.0 0.72 0.66

30.0 0.75

120.0 0.79, 0.715

Average Value 0.74 + 0.02 (+ SDM, n 6)
Standard Deviation 0.05

Standard Deviation of Mean 0.02

Internal Standard: d-10-Anthracene

These preliminary results are to be used on the basis of their
relative accuracy. Subsequent determinations of absolute DBP
levels were found to be more accurate.

0.
humic peak
a.

0-0 CD
0 CO
C
C 0
0 OL
CL

cc
0 .0

Inject Inject

0 2 4 6 0 2 4 8
Time (minutes) Time (minutes)

Figure 12. Liquid chromatogram. of Figure 13. Comparison liquiA_ chromatogram
Morgan Creek sediment from of humic acid.
Frye Farm a few miles down-
stream from Tenneco.
u

that the desorption rate in CH2C12 was too rapid to be observed
kinetically. The extraction time was fixed at 2.0 minutes. It
should be noted here that reproducible time course studies were
not demonstrated until we discovered the necessity of using
cooling water (300C) to prevent a temperature rise and sub-
sequent loss problems during the extraction period.

Effect of Solvent--
Various solvents were tested for their extractive power.
The conditions were identical to those used in Table 12. The
results are shown in Table 12. It is clear that hexane gives
decidedly poor extractive efficiency, while benzene, dichloro-
methane and methanol give quite similar extractive efficiency.

The choice of solvent depends upon more than the rate of
extraction. The variation in total amounts of extracted organic
matter was not measured. A simple color comparison of extracts,
often clarification by centrifugation, showed that hexane exhi-
bited least color, dichloromethane gave a visible light yellow
color, and methanol showed a substantially deeper color.
Analyses were performed on the extracts immediately after ex-
traction. After a period of several weeks a gum would form in
the dichloromethane extract and inconclusive tests suggested
that dissolved organics might have been lowered in concentration
as a result. Clearly, methanol would lend to a distinctly
higher concentration of nonvolatile substances and these would
be likely to foul the chromatograph. Hexane gave the cleanest
extract but, of course, it has a poor yield. Benzene is toxic
so this was ruled out. Therefore, dichloromethane was chosen
for its high extractability, low toxicity, and ease of
volatilization.

TABLE 12. COMPARISON OF ULTRASONIC EXTRACTION EFFICIENCIES
OF VARIED SOLVENTS MEASURED ON WORKING STANDARD*

Solvent Trials Measured DBP
(ppm)

Hexane 3 0.34 + 0.07t

Benzene 3 0.65 + 0.02

Dichloromethane 3 0.70 + 0.01

Methanol 3 0.67 + 0.03

Based on a fixed ultrasonic extraction period of 2.0 minutes
(Chester River Sample "R").
t Tolerance is expressed as the standard deviation.
Internal Standard: D-10 anthracene.

Recovery Tests--
Several experiments were carried out to show whether DEHP
could be added to the measurement system and then recovered
without serious loss. The results are summarized in Table 13,
which is largely self-explanatory. The most serious loss seemed
to occur when dichloromethane containing a DEHP spike was de-
liberately evaported to remove all of the solvent. When this
was done a 7 percent DEHP loss was observed. To prevent this
in the actual procedure, a small amount of isooctane, a higher
boiling solvent, was added to prevent the volatilization and
no other measurements in the report involved total solvent
volatilization exhibited here for test purposes.

TABLE 13. DEHP RECOVERY MEASUREMENTS USING DICHLOROMETHANE
Sample Procedure Trials, Percent
Recovery
Spiked Solvent* Remove 95% of solvent. 3 100 + 2
Add fresh solvent to
bring to initial volume.

Spiked Solvent* Remove 100% of solvent. 3 93 + 1
Add fresh solvent to
bring to initial volume.

Spiked Blankt Remove 98% of water by 98 +.l
(attapulgite) desiccation.

Spiked Blank Ultrasonic extration by
(attapulgite) dichloromethane

5 ml of CH 2Cl2 containing 1 mg of added DEHP.
t 1 g of Attapulgite containing 0.2 mg of added DEHP.
N.B.(l) Internal standard = d-10-anthracene
N.B. Reagent blanks are as follows:
CH2Cl 2 = 20 + 10 ppb of DEHP

Attapulgite = 10 ppb of DEHP

Effect of Varied Methodology and Combinations---
. The establishment of a 2.0 minute sonication procedure for
.extracting dry sediment with dichloromethane was subjected to
a series of tests to measure the completeness of the extractive
process. The same working standard "R" was used. The sonica
tion procedure is the same as that previously described. The

Soxhlet extraction procedures are as follows:

Wet Soxhlet Procedure

1. The extraction thimble 43 x 123 mm was extracted
with methylene chloride for 24 hours before use.

2. A clean 2000 ml flask was charged with 500 ml of
methanol containing approximately 5 pg of the
internal standard, d-10 anthracene (C 14D 10).
3. The thimble was loaded with approximately 90 g
of wet sediment (60 percent moisture content).
The sediment was smeared onto the walls of the
thimble cylinder to provide more intimate con-
tact with the solvent.

4. After 24 hours the methanol was emptied into a
storage container and replaced by 500 ml methylene
chloride and an additional 5 pg of d-10 anthracene.

5. The extraction was continued for at least 48 hours,
when the extract solution in the thimble chamber
became completely clear.

6. The solvents were distilled off using Buchi Roto-
vapor R. until approximately 30 ml of aqueous
extract remained.

7. This aqueous liquid was extracted with 5 volumes
of methylene chloride. This extract was concen-
trated to 5 ml using the Rotovap. Further
evaporation was done using a stream of nitrogen
(XAD-2 filtered).

8. A blank was prepared by duplicating the procedure
described above, but without the sediment samples.

Dry Soxhlet Extraction--This procedure is the same as the
wet Soxhlet extraction procedure except for the following modi-
fications.

1. The sediment was first dried by desiccation to a
moisture content of 1-2 percent.

2. Approximately 40 g of dried sediment was added to
the previously extracted thimble.

3. Only methylene chloride was used as extracting
solvent and the extraction duration was 72 hours.

4. The methylene chloride extract wa s concentrated
directly to 5 ml using the Rotovap followed by a
stream of nitrogen. No solvent extraction step
was needed.

The results of the various extraction tests are as
follows:
TestA1comparison of the efficiency of ultrasonication
and Soxhlet extraction is shown in Table 14. Ultrasonication
extracted greater amounts of phthalates than Soxhlet extraction
from the same sediment collected at the mouth of the Chester
River. The incompleteness of the Soxhlet method is believed to
be the result of the lower eddy diffusion of solvent in the com-
pacted solid sediment. With particular regard to the time course
study, it is felt that the present experiments give a sound
basis to reject the Soxhlet technique on the basis of its
greater proneness to contamination, its incompleteness and the
high cost associated with slow rate and need for laborious re-
petition.

Test 2

A second and more stringent test of the sonication
technique was carried out. Sonicated material was rinsed free
of retained extracting liquid, and the samples were then
subjected to re-extraction. The results are shown in Table 15.

The repeated use in test 26 of sonication drew a
blank--no evidence for further extraction was observed. Es-
sentially the same finding was observed when the sonicated
sample was washed free and then subjected to Soxhlet extraction.

Test 3
A comparison was made of wet Soxhlet and dry sonica-
tive extraction of Tenneco Pond sediment. In general, these
more heavily polluted samples of sediment seemed to be more
easily extracted so that is not a stringent test.

Analysis of the Internal Standards--The use of dimethoxyethyl
phthalate (DMEP) seemed ideal in the sense that its gas chromato-
graphic elution time fit into a window that caused no inter-
ference with the DBP or DEHP measurements. Further, its
chemistry was parallel to that of the other phthalate esters.
However, we eventually became aware that our procedural blank
levels were significant: 0.10 + 0.04 (n=6) ppm for DBP and
0.28 + 0.22 (n=3) ppm for DEHP. Direct analysis of the DMEP
stahdard revealed the cause since it contained 4 percent DBP and
11 percent DEHP! These are not serious interferences since the

TABLE 14. COMPARISON OF METHODS FOR EXTRACTION OF PHTHALATE
ESTERS FROM CHESTER RIVER, MOUTH SEDIMENT. TEST 1.
Extraction Method Amount Extracted, ppm, and
Standard Deviation

DEP DBP DEHP

Ultrasonication 0.19 + 0.03 0.36 + 0.07 0.40 + 0.06

Soxhlet, dry 0.10 + 0.03 0.33 + 0.14 0.34 + 0.06

Soxhlet, wet 0.05 + 0.10 0.26 + 0.07 0.21 + 0.12

Re-extraction by ultrasonication of sediment frorn all three
methods yielded less than 1 percent of the first extraction value
for-all three phthalates.

TABLE It. TEST OF VARIED EXTRACTION AND RE-EXTRACTION PROCEDURES
Sample Test Method Sediment Trials Result (ppm)
State
Chester River 1 Soxhlet and Dry and 3 See Table
Sample "R" sonication wet

(same) 2a Soxhelt Dry 3 0.9+0.1(DBP)
Re-Sonication 3 0.04T0.02(DBP)

(same) 2b Soxhlet 4 0.7+0.1(DBP)
Re-Sonication 3 0.62 (DBP)t
Re-Soxhlet 1 0.02 (DBP)t
Tenneco Pond 3 Soxhlet Wet 1 1.0x103(DEHP)
Sonication Dry 5 (1.2+0.1)xlO3
(DEHP)

+ Uncertaintie s are expressed as standard deviations of the mean
values (rounded up).

i.e., none detected.

Internal Standard: D-10-anthracene

calculations used a conventional blank subtraction step. How-
ever, the use of the contaminated DMEP was not considered to be
desirable and another standard was introduced.

Since the method verification and environmental measurements
were made at different times, we have been careful to note the
internal standard for each experimental series.

The more recent work involved the use of an excellent in-
ternal standard, D-10-anthracene, Cl4DlO_ A small amount of DBP
and a negligible amount of DEHP were observed with the anthracene
standard. A correction for the DBP in the average blank was used
in the caluculations. The use of C14D,o standard allowed more
reliable measurements in the sub-ppm region. The basis for these
statements, for the GCMS capability for discriminating between
DEHP and DNOP, for the virtual absence of DNOP from sediment, and
for the purity of the D-10-anthracene internal standard, i.e.
illustrated in Figure 14.

Interlaboratory Comparisons of Split Sediment Samples

Two samples were carefully homogenized and sent to Dr.
William Budde (EPA - Cincinnati) for interlaboratory comparison.
The samples are the Chester River working standard "R" and a
Tenneco Pond sample. The results are shown in Table 16. A much
closer agreement exists between the Tenneco Pond than the poor
results obtained on the initial "R" sample. Since this involved
two essentially independent experiments on the two unrelated
samples and standards, it was decided to accept the results as
validative for Tenneco Pond and to reject the results on sample
"R" as nonvalidative. .. I

The interlaboratory comparison was repeated using a fresh
sample from our working standard "R." Samples were sent with and
without use of our drying procedure. In addition, a spiked blank
sample containing 200 ppm DEHP was sent for comparison measure-
ments. Measurements on the split portions that were saved for
that purpose were repeated. The results are given in Table 16.
The results show reasonable agreement between the two labora-
tories. It should be noted that the results from EPA-
Cincinnati seem to be higher than our results, and this remaining
problem is still under study. However, the main purpose of
demonstrating comparability has been achieved. A chronological
account of interactions with Dr. William Budde (EPA-
Cincinnati) is presented in Appendix B.

FI
0 0 0 0, 0
1 9
> DEP
Ln H- V1 Fl-

Lo 0 ko 0
n rt
0 0,@
t-bo (nmo

rt
ti t-h (D (n tt Diisobutylphthalate
0 w @3 0
>
0 DBP m fQ DBP
0) C:) 0
ro C) @I 0 o

0 F'
t4
rt, (t (D rt >
(D 0
t- U) 0
rt J@ rt ri-
En @I (D SI)
rt, ID W "
m 91 P, rt
:@ el W (D
91 P. X f-- X
0)10 x rt rf-
F-I 5' NJ En t1i A)
c 09 a 0 %00
0) w Dihex ylphtha late
0
rt rt
o 'Ei o
1 0

0 co 0
(D
co w U)
:@ 10 DEHP
rt 0
0 01 @r :5
LQ z oj DEHP
Lp M 0
P) m rf-
@g @j0
cn 9u m DNOP

(n
rt,

Ln
ihex
D@blpht
ISO Uty

DBP

Ylphtl

DEHP

op
D @N

TABLE 16. INTERLABORA TORY COMPARISON OF DEHP MEASUREMENTS OF
SPLIT SAMPLES* This Laboratory EPA-Cincinnati
Samples Method DFHP (ppm) see DFHP (ppm) (n)
note d
Chester River, "R" GCMS 0.097+0.03 (5) 58 see note a.
0.3 see note b.
0.4 see note c.

Spiked Attapulgite GC 200+10 (5) 121
(dry) GCMS

Tenneco Pond Wes t GC 1200+100 (5) 1700
(dry)

Notes:

a. Initial determination. This errant result necessitated
the subsequent repeating the entire analysis using
freshly split samples.
b. Dried by this Laboratory.
C. Dried by EPA-Cincinnati.
d. Reagent blanks measured DEP 0.0005 ppm, DBP 0.001 ppm,
DEHP 0.003 ppm.

Further details are given in Appendix B.

Tenneco Pond and Chester River Sediments

Sediment analyses for DBP and DEHP were measured using the
GC (Tenneco Pond) and GCMS (Chester River) techniques. The
results are presented-in Table 1-7. The samples in all cases were
surficial--top 10 cm--that were thoroughly homogenized prior to
drying and subsampling. The results are persented along with
percentage of organic carbon [HC1 (0.1M) treatment was used to
remove carbonate] and percentage of water. All results are pre-
sented on a dry-weight basis.

Two samples were taken from Tenneco Pond which is roughly
oval in shape. These separate samples were obtained near the
center axis of the pond about one third of the distance from
either end: these are labeled East and West.

It is clear that the pond sediment is quite substantially
polluted with 0.15 percent DEHP. The previously cited LC re-
sults which seem to be much better for the larger alkyl phtha-
lates show similar amounts of other compounds: 0.14 percent DIDP

and 0.19 percent DTDP. The sum is nearly 0.5 percent on a dry-
weight basis.

The total content of organic phthalate esters in Tenneco
Pond can only be guessed at since the vertical concentration
distribution was not determined in this work. For a dnth of
0.04 m, the Tenneco Pond sediment volume for 40 x 103 m of sur-
face area is 1.6 x 103 m3, or roughly 6.4 x 105 kg on a weight
basis. The measurements are thus compatible with the presence
of 960 kg of DEHP, 900 kg of DIDP, and 1200 kg of DTDP, or a
total of about 3,000 kg of these three organics.

As discussed earlier, the State discharge permit allows
Tenneco to release total organic extractables at a rate of
2000 kg/year. It would appear that Tenneco discharge may be
operating rather near the permit, depending on where the
measurement is made: plant discharge or outfall from the dam.
The pond is clearly acting as a secondary waste treatment facil-
ity, and it is obvious that less organics have been flowing out
of the pond than those that enter. The DBP content is low be-
cause Tenneco has rarely manufactured it.

The fact that the pond levels are now high should be inter-
preted as a warning. As the pond sediment becomes increasingly
saturated with these organics, they may eventually move out into
the Morgan Creek conduit. Since the linear flow rates are much
higher and the creek bed is rather narrow, it is important to
consider the eventual saturation of the present sorptive capacity
of the pond sediment. At a future time the allowed Tenneco dis-
charge, 2000 kg/year, may become more likely to make a direct
transit from the factory site to the Chester River, and to pos-
sibly high localized concentrations in the river sediments.. If
this situation is allowed to continue without any further re-
straint, one can not help but visualize a more pessimistic
future for the sediment beds in the Chester River.

The Chester River sediments exhibit ranges of 0.020 to 0.064
ppm for DEHP and 0.23 to 0.85 ppm for DBP. Stations 3, 4, 5 and
6 were deliberately localized at the apex of the oyster mortality.
The data are not marked by major apparent differences from the
site "R" at the river mouth. Since healthy oyster bed's have been
maintained downstream of sited 4-6, there is no framework pro-
vided by the present data to assign alkyl phthalates as causally
related to the 1973-75 oyster mortality. of course, this is
purely circumstantial reasoning, and the results can not be used
to rule out the possibility. However, no oyster samples were
saved.

TABLE 17@ DETERMINATION OF SEDIMENT COMPOSITION

Site Lat/Long %H 20 % C Method Alkyl Phthalate Concentrations* DEHP/DEP

DEP (n) DBP (n) DEHP (n)
ppmt ppmt ppmt
,Tenneco 39012115" 52 4.0 GC 1.5xlO 3(4)
Pend 76004145" LC (1.3+0.2)x
East 163(4)
LC (1.8+0.3)x
15-3(4)
Tenneco of 52 2.6 GC 0.2+0.1(5)
West

Tenneco Pond average value 8000+4000

Frye Farm 1 1.6 GCMS 0.027(2) 0.013+ 3.9+1.4(3)
(Morgan Creek) 0.007(4)
at Rt. 314

Morgan 3.8 GCMS 0.014(2) 0.076+ 0.050+ 0.7+0.4
Creek 0.057(6) 0.01T(6)
at
Rt. 291

Chester- 1.3 GCMS 0.016+ 0.049+ 0.020+
town 0.00f(3) O.Olf(6) 0.00-9(5) 0.4+0.2
Bridge

Site 7 39009'13" 52 2.8 GCMS 0.073+ 0.23+ 0.064+
76004113" 0.021(4) 0.14(4)

Site 6 39005'44" 57 2.8 GCMS .0.040(2) 0.60(2) 0.053(l)
76009'1511

(continued)

TLABLE 17. (continued)
Site Lat/Long %H20 % C Method Alkyl Phthalate Concentrations* DEHP/DEP
DEP (n) DBP (n) DEHP (n)
ppmt ppmt ppmt

Site 5 39005'52" 51 2.3 GC MS 0.042(2) 0.51+ 0.039(l)
76008149" 0.1'T(3)

Site 4 3900511511 53 2.4 GCMS 0.075(1) 0.85(2) 0.050(l)
76011'03"

Site 3 39004112" 49 2.4 GCMS 0.025+ 0.46+ 0.044(2)
76009155" 0.002(3) 0.23(3)
Average value in mortality zone (Site 3-7)0.05+ 0.5+ 0.05+ 0.1+0.07
O.OT(12) 0.-:T(14) 0.0T(q)

Site 2 3900211811 58 3.4 GCMS 0.065+ 0.57+ 0.071(l)
76009'55" 0.001(3) 0.09(3)

Site 1 38059'31" 57 2.9 GCMS 0.029(2) 0.47(l) 0.11+
76012'50" 0.02(3)

Site R 3902154" 65 2.9 GCMS 0.014(2) 0.033+ 0.097+ 3+1
Chester 0.017(5) 0.03-f(5)
.River
Mouth

t error values are only given for n>2
+ DnOP (Di-n-octyl phathalate) was not found in Chester River sediment (<0.001 ppm)
Reagent blanks measured: DEP<0.0005 ppm, DBP<0.001 ppm, DEHP<0.0003 ppm
Other alkyl phthalates found in Tenneco Pond (East) DIDP(l.4+0.2xlo3(4);
DTDP(l.9+0.2)x,03(4)

ANALYSIS OF PHTHALATE ESTERS IN OYSTER TISSUE

Procedure

1. Whole oysters were collected from the Chester River and
pooled according to sampling site. The number ranged
from 4 to 8 individual oysters per site.

2. The pooled tissue was homogenized with a Virtis "23"
homogenizer for one minute at low speed followed by
two minutes at medium speed.

3. The homogenized tissue was then disrupted with a
Branson W185 ultrasonic probe for 10 minutes.
Light microscopy showed that this caused complete cell
disruption.

4. The ultrasonic probe -Yas cleane d by running the device
twice in distilled water and once in methylene chloride.
The homogenizer flask and blades were then rinsed with
water, methanol, and methylene chloride. Blanks of
water were run between tissue preparation to check for
contamination.

5. Twelve to fifteen g of tissue from each site plus two
controls prepared from commercial oysters were placed
in 50-ml Erlenmeyer flasks, shell frozen, and placed
in a vacuum desiccator.

6. The desiccator was attached to a Virtis lyophylizer and
the dry tissue was dried in 24 hours.

7. The dried tissue was removed from each flask, pulverized
with a mortar and pestle, and stored at 30C.

8. Two hundred mg of dried tissue was placed in 10-ml vials.
A 10 pg aliquot of anthracene d-10 was added with a
Corning disposable micro pipette (+ 0.5 percent ac-
.qpracy). Five ml of methylene chl-oride was added.

9. The tissue was extracted using ultrasonic agitation
(Bransonic 220 water bath) for 5 min.

10. The vials were centrifuged at 4,000 rpm for 15 min. to
sediment the remaining tissue residue. The supernatant
was drawn off. The sample was concentrated by solvent
evaporation to a volume of approximately one milliliter.

11. The concentrated extract was injected directly in the
GC-MS with a 25M capillary column coated with SE-52.

Samples

Oyster samples from the Chester River were collected July 2,
1979 by personnel of the Chesapeake Biological Laboratory. The
samples were frozen for subsequent delivery. They arrived in
jars which had cracked presumably during the freezing process.
The extent of possible contamination, however, was considered
minimal.

Oyster sample CBL-1 was collected at Buoy Rock correspond-
ing to sediment site 1. Sample CBL-2 was from Spaniard Point
where sediment sample 6 was collected. There is no equivalent
sediment site for the CBL-3 sample collected at Ferry Bar. A
fourth sample from Love Point contained too few oysters to
analyze. A homogenate of seven commercial oysters from northern
Chesapeake Bay was used for reference purposes. Unspiked sam-
ples as well as samples spiked with DEP, DBP and DEHP were pre-
pared from this.

Results

Oysters contain approximately 80 percent water and 2 percent
lipid. Both of these tend to interfere with the analytical de-
termination. The water must first be removed before considering
extraction of the phthalates with methylene chloride. The lipid
content is readily soluble in methylene chloride, as are the
alkyl phthalates. An attempt was made to dry the oyster tissue
using the same method successfully used for wet sediment. The
tissue homogenate was spread as thin layer on a watch glass and
placed in a desiccator containing sodium sulfate. This method
was discarded because a bacterial bloom tended to form on sam-
ples prepared in this way. However, the controls from commer-
cial oysters were free from this mold-like growth. It is
suspected that these oysters are thoroughly rinsed and an
anti-bacterial preservative is added during processing.

A more rapid drying method was needed. The preferred alter-
native was lyophilization or "freeze-drying." Spiked and un-
spiked controls were first freeze-dried to evaluate the method.
The spiked samples contained 20 ppm each of DEP, DBP and DEHP.
The extracts of these samples were injected directly into a
glass capillary gas chromatograph with an FID detection.
Numerous, partially resolved peaks resulted but the spiked peaks
were not evident. GC-MS was used to identify the four major
sets of peaks. The first large peak was identified by matching
its mass spectrum to reference spectra as palmitoleic acid and
closely related compounds that elute at 200'-203*C. The second
large set of peaks eluting at 215'-2170C was identified oleic
acid. The third (2280-232*C) and fourth (244*-248*C) groups are
linoleic acid and linolenic acid, respectively.

The extract were analyzed for phthalate ester using the
selected ion monitoring mode. Chromatograms of the spiked and
unspiked oysters are shown in Figures 14 and 15. Even though
percent abundance of m/e 149 is linolenic acid is 3 percent, its
high concentration in the injected extract makes it clearly
evident in the m/e 149 selected ion monitoring team. Indeed,
all'four sets of peaks similarly appear in the m/e 149 mass
chromatogram. While the DEP and DBP are resolved from inter-
fering peaks, DEHP co-elutes with the linolenic acid group.
The DEHP concentration must be greater than 2 ppm in order to
make the interfering background insignificant and obtain mean-
ingful quantitative results.

The recoveries of DEP, DBP and DEHP from spiked (20 ppm)
oyster tissue takers through the entire drying and extraction
procedure were 60 + 6, 96 + 5, and 100 + 15 percent, respective-
ly. The smaller recovery for DEP is assumed to be due to voli-
tization during the freeze drying process.

The measured levels of alkyl phthalates in oyster tissue are
reported in Table 11 along with values for sediment samples
taken from the same zone.

As discussed earlier, the octanol-water partitioning model
predicts that the hydrophobic phthalates will be concentrated
in the total lipid portions of the oysters relative to the sur-
rounding water. We did not measure lipid content per se, but
the oyster and sediment values were compared on the basis of
the measured total organic carbon content which was measured.
The oyster contained a 16-fold higher concentration of organic
carbon than the sediment. Accordingly, the alkyl phthalate
concentration is predicted by the partition model to be approx-
imately 16-fold higher than that in the sediment. This pre-
diction@does not consider any mechanism otherthan simple
partitioning. Table 17 shows that the oyster-to-sediment ratio
of total phthalate ester residue is consistent with this model.
At the two sites where contiguous oyster and sediment samples
were obtained, the ratio is 16:1 (Buoy Rock) and 23:1 (Spaniard
Point). This suggests that the surficial analysis of stratified
(anoxic, nonbioturbated) taken nearby oyster beds may provide a
basis for predicting the concentration of phthalates in the
nearby oysters. It is also apparent that the level of phtha-
lates in the oysters from the Chester River do not signifi-
cantly differ from the reference sample of commercial oysters,
also taken from the northern Chesapeake Bay.

Table 18 shows that the measured levels of alkyl phthalates
in oysters range from 0.45 to 1.5 ppm (wet basis). The U.S.
Bureau of Sport Fisheries and Wildlife reported recently (Mayer
et al. 1972) a range of 0.2 to 3.2 ppm. of alkyl phthalates in
channel catfish and walleyes from various parts of North
America. Also, a survey of 1455 catfish farms (Haudet 1970)

Ui
Ion 149.30 SE-52 glass capillary column. 20 meters
Hewlett Packard 5992 GC-MS, selected ion mode
column temperature: 160-2750C 7.50C/min
injection port temperature: 2300C
splitless injection.3 ul

CD
0 CL
0
z
-JL-
--J LL LL
30

16:1-Paimitoleic Acid
Sm of Monitored Ion Abundances CM !0- 18:1-0leic Acid
66 18:2-Linoleic Acid
18:3-1-inolervic Acid

0
0
0

d
Q)
cc
U

LL
5 10

TIME (minutes)

Figure 15. Oyster tissue extract (spiked with 20 ppm DEP, DBP, and DEHP).

o- Ion 149.30 a. SE-52 glass capillary column, 20 meters
W Hewlett Packard 5992 GC-MS, selected ion moniter mode
column temperature: 160-2750C 7.50C/min
injection port temperature:2300C
splitless injection,3 ul 0.
W CL
0
z

CD
N

Sum of Monitored Ion Abundances 16:1 Palmitoleic Acid
18: 1 -01eic Acid
18:2-Linoleic Acid
OD 18:3-Linolenic Acid

0
0
0 Co
66

U-
5 10 15

TIME (n**jtes)

Figure 16. Oyster tissue extract.

M@ =201m

no mm mm mm mm man mm M mm=

TABLE 18. CONCENTRATION OF PHTHALATES.IN OYSTERS AND RELATED SEDIMENTS*
Site Lat/Long Sample %H20 % C DEP, ppm (n) DBP, ppm (n)
dryt wet T - dryt wet
CBL-1 38059131" Oyster 82 44 4.1+ 0.74+ 3+1(3) 0.5+0.2(3)
Buoy Rock 7601215411 O.'T(3) 0.0-5(3)
(sediment 3805911311 Sediment 57 2.9 0.029(2) 0.012(2) 0.47(l) 0.20(l)
site 1) 76012150"-
CEL-2 39105153" Oyster 62 44 3+3(3) 0.5+ 5+2(3) 0.9+0.4(3)
Spaniard 76009115" 0.!@(3)
Point
(sediment 39005144" Sediment 51 2.8 0.040(2) 0.017(2) 0.60(2) 0.026(2)
site 6 76009115"

CBL-3 39100108" Oyster 82 44 <0.05(3) <0.01(3) <0.05(3) <0.01(3)
Ferry Bar 76014152"
no equi-
valent
sediment
site

Control Oyster 82 44 2+2(3) 0.4+ 7+4(3) 1.3+0.7(3)
Commercial 0.4(3)
Oyster
N. Chesa-
peake Bay

(continued)

TABLE 18. (continued)

Site Lat/Long Sample DEHP" ppm (n) Total C T CT oyster % C oyster
dryt wet dry CT sediie-nt % C sedime

CBL-1 38*59'31" oyster 3+6(3) 0.5+ 10+6(9) 16+9 16
Buoy Rock 7601215P-1 1.5(3)
(sediment 38059113" Sediment 0.11+ 0.047+
site 1) 76012150" 0.0f(3) 0.007(3)

CBL-2 39*05'53" Oyster 8+5(3) 1.4+ 16+9(9) 23+13 16
Spaniard 0.9(3)
Point

(sediment 39005'44" Sediment 0.053(l) 0.023(l) 0.7
s
'te 6) 16009'15"
A-

CBL-3 39*00'08" oyster <2(3) <0.4(3) <3(9)
co Ferry Bar 7601415211
1-4
no equi-
valent
sediment
site

Control Oyster <2(3) <0.4(3) <11(9)
Commercial
Oyster
N. Chesa-
peake Bay
Refer to page 76 for discussion of alkyl phthalate toxicity to aquatic organisms.
Based on use of D-10 anthracene as the internal standard.
t Based on whole, dry oyster tissue.
Based on whole, wet.oyster tissue.
Reagent blanks measured: DEP<0.0005 ppm, DBP<0.001 ppm, DFHP<0.0003 opm.

MM =,owns MMONSM MMM

revealed that 95 percent of the fish analyzed contained DEHP
residues with an average DEHP concentration of 3.15 ppm.

The alkyl phthalate measurements reported here can be com-
pared, at least indirectly, to available toxicity data. Chem-
ical toxicity results for aquatic organisms are usually
presented as LC50 values (LC50 is the estimated concentration
where the toxic substance in the surrounding water will kill
half the population of exposed organisms within a certain time
period, usually 96 hours.) Since the present study presents
values for sediment concentrations, not water, an estimate of
the water concentration using the octanol/water partitioning
model was made. The partitioning Vnstant (Kow) of phthalate
esters is estimated to be 103 - 10 (TSCA 1978). The average
sediment values on an organic carbon basis (Coc) for the mor-
tality zone are 2 ppm, 20 ppm. and 2 ppm (pg/g organic acid) for
DEP, DBP, and DEHP, respectively. The estimated ester concen-
trations in water are, therefore, 0.2 ppb, 2 ppb, and 0.2 ppb
(pg/L), respectively, for DEP, DBP and DEHP. Mayer and Sanders
(1973) reported that acute 96 hours LC50 values for DBP with
fathead minnor, channel catfish, rainbow trout, scud and cray-
fish fell between 730 and 10,000 ppb (pg/L). The LC 50 concen-
trations for DEHP were estimated to be above 10,000 ppb. The
estimated water concentrations in the Chester River are cur-
rently significantly less than these values. In addition, the
average alkyl phthalate concentrations in oyster tissue in the
Chester River range from 3 to 8 ppm. This is 1000-fold lower
than the acute toxicity of alkyl phthalates for rats (intra-
peritonel) which ranges from 3-14 g/kg (percent). In comparison,
the-LD50's for organochlorine pesticides in rats are three
orders of magnitude lower (20 - 300 mg/kg) than alkyl phthalates
(TSCA 1978).

Chronic toxic effects of alkyl phthalates on aquatic orga-
nisms appear at much lower levels than acute toxicity. Mayer
.and Sanders (1973) reported that a concentration of only 3 ppb
of DEHP in the water was sufficient to significantly decrease
growth and reproduction of the crustacean Daphnia magna. Zebra
fish and guppy reproduction was also impaired when err food
was spiked with 50 and 100 ppm of DEHP. Various effects of
alkyl phthalate have been reported for brine shrimp, goldfish
and.,ring doves at concentrations varying from 3 to 10,000 ppb.

No chronic or acute toxicological data are available on the
effects of phthalates on oysters. Martin and Roosenburg (1979)
studied oyster mortality at 10 stations in the Chester River.
No significant mortality occurred during the four-month period
of observation. However, during July and August five stations
furthest upstream had dieoff of the fouling organisms and re-
duced growth rates of the oysters. Ninety-six hour acute
toxicity studies were performed on golden shiners and crayfish
in streams receiving effluents from the Campbell's Soup plant,

the Tenneco, Inc. plant and the Chestertown sewage treatment
plant. No significant oyster mortality was observed.

It appears that a comparison of our measurements with the
known toxic effects of phthalate esters on oysters and other
aquatic organisms is unable to explain the Chester River oyster
mortality during 1973-1975. However, there is enough evidence
to cause concern over long-term effects on aquatic organisms
especially with expected future increases production of alkyl
phthalates.

BIBLIOGRAPHY

Amundson, S. C. 1978. Determination of di(2-ethylhexyl)
phthalate, non(2-ethylehexyl) phthalate and phthalic acid by
high pressure liquid chromatography.
J. Chromatogr. Sci. 16:170.

Anonymous. 1976. Outlook for PVC plasticizers in bright.
Chem. Eng. News (Nov..22, 1976):12.

Chesler, S. N., B. H. Gump, H. S. Hertz, W. E. May, S. N. Dyszel,
and D. P. Enagonio. 1976. Trace hydrocarbon analysis: The
National Bureau of Standards William Sound/Northeastern Gulf
of Alasks Baseline Study. NBS Tech. Note 889.

Gaudet, J. [Ed.] 1970. Report of the 1970 workshop on Fish
Feed Technology and Nutrition. Bureau of Sport Fisheries
and Wildlife, Fish and Wildlife Service, U. S. Department of
Interior.

Giam, C. S., H. S. Chan, and G. S. Neff. 1975. Sensitive
method for determination of phthalate ester plasticizers in
open ocean biota samples. Anal. Chem. 47:2225-2229.

Giam, C. S., H. S. Chan, T. F. Hammergren, and G. S. Neff. 1976.
Sensitive method for determination of phthalate ester plas-
ticizers in open ocean biota samples.
Anal. Chem. 48:78-80.

Gingras, S. A. 1979. Analysis of alhyl phthalates and other
organic substances in the environment. M.S. thesis, Univ.
of Maryland, Department of Chemistry, College Park, MD.

Han, G. 1972. The delination of an exclusion area around the
Chestertown outfall on the Chester River and the Back River
sewage treatment plant outfall. Chesapeake Bay Institute
(The Johns Hopkins Univ.) Special Rpt. 26, Baltimore, MD.

Hellman, M. Y. 1978. Analysis of phthalate plasticizers for
PVC by liquid chromatography. J. Liq. Chromatogr. 1:491-505.

Kahn, S. U., and M. Schnitzer. The retention of hydrophobic
organic compounds by humic acid.
Geochim. Cosmochim. Acta 36:745-754.

Karickhoff, S. W., D. S. Brown, and T. A. Scott. 1979. Sorp-
tion of hydrophobic pollutants on natural sediments.
Water Res. 13:241-248.

Martin, F. D., and W. Roosenburg. 1979. Evaluation of Chester
River oyster mortality biotoxicity. EPA Rpt. R805976010.
Chesapeake Biological Laboratory, Solomons, MD., Ref. No.
79-103 CBL (unpublished).

Mayer, F. L., Jr., and H. 0. Sanders. 1973. Toxicology of
phthalic acid esters in aquatic organisms.
Environ. Health Perspect. 3:153-157.

Mayer, F. L., Jr., D. L. Stalling, and J. L. Johnson. 1972.
Phthalate esters as environmental contaminants.
Nature 238 (Aug. 18, 1972):411-412.

Meritt, D. W. 1977. Oyster spat set on natural cultch in the
Maryland portion of the Chesapeake Bay (1939-1975).
Univ. Md. Ctr. Environ. Estuarine Stud. Spec. Rpt. 7.

Neely, W. B., D. R. Benson, and G. E. Blau. 1974. Partition
coefficient to measure bioconcentration potential of organic
chemicals in fish.
Environ. Sci. Technol. 8:1113-1115.
Palmer, H. D. 1972. Geological investigations, pp. 75-137.
In: W. D. Clark [Ed.] Chester River study. State of
Maryland, Dept. of Natural Resources, and Westinghouse Elec.
Corp., Vol. II.

Palmer, H. D., J. R. Schubel, and W. B. Cronin. 1975. Estuar-
ine sedimentology, pp. 4/29-4/30. In: Munson, T. 0., D..,K.
Ela, and C..Rutledge, Jr. Upper Bay survey and final report
to Maryland Department of Natural Resources, Annapolis, MD
21401 [Nov. 30, 1975].

Pritchard, D. W. 1967. Observations of circulation in coastal
plain estuaries, pp. 37-44. In: G. H. Lauff [Ed.]
Estuaries. Publ. 83, Amer. Agsoc. Advancement Sci.,
Washington, D. C.

Singmaster, J. A., and D. G. Crosby. 1976. Plasticizers as
interferences in pollutant analysis.
Bull. Environ. Contam. Toxicol. 16:291-300.

U. S. Department of Agriculture. 1973. Composition of foods.
Agricultural Res. Serv. Handbook No. 8, Washington, D.- C.

U. S. Environmental Protection Agency. 1978. Initial report of
the TSCA Interagency Testing Committee to the Administrator,
Environmental Protection Agency, pp. II-i.-II-55.
U. S. EPA 560-10-78/001.

Watson, J. T. 1976. Introduction to mass spectrometry:
Biomedical environmental and forensic applications.
Chpt. 3. Raven Press, N. Y.

Webster,' R. D. J., and J. Nickless. 1976. Problems in the
environmental analysis of phthalate esters.
Proc. Anal. Div. Chem. Soc. 13:333-335.

de Zeeuw, R. A. 1975. Plasticizers as contaminants in high-
purity solvents:A potential source of interference in bio-
logical analysis. Anal. Biochem. 67:339-341.

SECTION 6

MICROBIAL TRANSFORMATION OF TIN

EXPERIMENTAL PROCEDURES

Sampling

Samples were taken in the Chester River at Buoy Rock (lati-
tude 3859'33"N, longitude 76012'27"W), which is a productive
bar; at Spaniard Bar (latitude 30*05'57"N, longitude 76*08'55"W),
which has experienced extensive oyster mortality starting in
1974; at the outfall from the Tenneco plant near Chestertown and
in the Tenneco holding pond near the pond's outlet (latitude
39*15'00"N, longitude 76"02'30"W); at the outfall from the
Chestertown sewage treatment plant at the edge of Morgan Creek
(latitude 39*12'00"N, longitude 76'04'15"W); and at the Campbell
factory near Chestertown (latitude 39*14'03"N, longitude
76"12121"W). Three sampling excursions were taken in the period
September 1977 through July 1979.

For comparison, some samples were taken in Baltimore Harbor
(latitude 39*13'57"N, longitude 76*30'16"W), which was expected
to be polluted with heavy metals, and other samples were taken
near Tilghman Island (latitude 39*40'64"N, longitude 76123'16"W),
an area expected to be relatively free of pollution by heavy
metals. Samples from the two Chester River sites, from Baltimore
Harbor, and from the site near Tilghman Island were considered
estuarine samples; samples from other sites were considered fresh-
water samples.

At each site, water temperature, pH, salinity, and dissolved
oxygen were determined. Methods are given in Appendix C.

Water samples were taken with a Kemmerer bottle. Concentra-
ted HC1 (0.5 ml) was added to each 200-ml water sample to keep
metals-in solution. This brought the pH to 1.8 to 1.9. Prior
to cheminal analysis, the pH of each sample was adjusted to 2.0
with 1 N NaOH. Sediment samples were collected with a Van Veen
dredge. A plastic corer was used to obtain material which had
not touched the metal walls of the dredge. Samples were taken
from the top centimeter of the core, and the remainder of the
core was sliced into 1-cm bands which were frozen and will be
maintained frozen for possible future use.

Samples from the sewage treatment plant, from the Campbell
factory, and from the Tenneco plant were iced in the field and
stored in ice until they were used in the laboratory. Samples
taken on board ship (Buoy Rock, Spaniard Bar, Baltimore Harbor,
Tilghman Island) were used for microbiological analysis within
15 min of sampling; the remainder of each sample was stored on
ice until it was used in the laboratory.

microbiological Samples

Water samples were used to prepare appropriate dilutions for
?lating. For sediment samples, 1.0 g (wet weight) was suspended
in 9.0 ml of sterile estuarine salts; further dilutions were pre-
pared from this suspension.

Total viable counts of aerobic, heterotrophic bacteria were
made using the spread plating technique, plating on Nelson's
medium (Nelson et al. 1973). Nelson's medium contains casamino
acids; 5.0 g; yeast extract, 1.0 g; glucose, 2.0 g; agar, 15.0 g;
and salt solution, 1 liter. For sediment samples from estuarine
sites, the estuarine salts solution contained NaCl, 10.0 g;
MgCl2-6HjO, 2.3 g; KC1, 0.3 g; and distilled water, 1 liter.
For samp es from freshwater sites, the salts solution was used
at one-tenth strength.

For viable counts of tin-resistant organisms, appropriate
dilutions were spread on the surface of Nelson's medium prepared
as above, supplemented with SnCl4 to yield 75 ppm tin suspended
in the medium. Extensive preliminary testing (Table 19) indi-
cated that this concentration of tin was appropriate to select
for tin-resistant organisms. Addition of SnC14 to the medium
resulted in a fine precipitate of Sn02 which was uniformly sus-
pended in the medium by agitation. In one series of experiments
the organisms resistant to organotin were estimated by plating
on Nelson's medium containing 15 ppm tin as (CH3)2SnCl2-

All platings were prepared in triplicate. Plates were incu-
bated at 25 + 2 C for 3 days prior to counting time.

Two systems were employed to determine if the microbial flora
in samples could transform tin to volatile organotin compounds:

i) Bioflask s. 250-ml flasks containing 20 ml of Nelson's
medium supplemenE-ed with 75. ppm tin as SnC14, were inoculated
with 1.0 ml of sediment suspended in estuarine salts. Each
flask contained a 10-ml beaker embedded in the agar. The beaker
held 5.0 ml of a solution of 8 percent (wt/vol) citric acid in
10 percent HC1. The flask was sealed with a rubber stopper.
After 14 days incubation at 27 + 2 C, material in the beaker was
examined for tin. Thus, if organisms growing on the medium pro-
duced volatile tin compounds, they would be detected in the acid
solution. Sterile controls to check for nonbiological production

TABLE 19. DETERMINATION OF A CONCENTRATION OF TIN WHICH
WOULD SELECT FOR TIN-RESISTANT MICROORGANISMS

Type of Tin added Viable count (mean + standard error)*
Sample (ppmSn, as
SnC14)
Water 0 8.7 x 102 + 1.8 x 102 a
50 5.4 x 102 + 6.8 x 101 b
Sediment 0 8.1 x 105 + 2.6 x 104 a
50 9.4 x 104 + 8.5 x 103 b
100 4.0 x 104 + 3.9 x 103 c
150 2.6 x 104 + 7.1 x 103 c
200 1.7 x 104 + 3.4 x 103 c

Means with the same superscript are not significantly different
at the 5 percent level as determined by a one-way analysis of
variance (ANOVA).

of volatile tin compounds were included in each experiment.
Each experiment also included positive controls in which the
medium contained 75 ppm tin as dimethyltin chloride. The use of
a solid medium containing a suspension of tin renders this method
qualitative, not quantitative.

ii) Hungate tubes. Each tube (Hungate 1969), containing
5.0 ml of liquid Nelson's medium (Nelson's medium minus agar),
was inoculated with 1.0 ml of sediment suspended in estuarine
salts. Sterile controls and positive controls were included in
each experiment. Tubes were incubated for 16 days at 27 + 2 C
on a rotary shaker operating at 80 rpm; each tube was then sam-
pled for the presence of organotin compounds.

For each water or sediment sample, one set of bioflasks and
Hungate tubes was inoculated from a sample which had received
500 )ag of sodium azide per ml. This additional set of controls
was used as a demonstration that effects observed in cultures
were due to biological activity.

Analyses of Tin

Inorganic tin--
For sediment samples, 1.0 g (wet weight) was transferred to
an acid-cleaned, screw-capped tube. Then, 2.0 ml of a solution
containing 50 percent (vol/vol) HC1 and 50 percent (Vol/Vol)
concentrated HN03 were added. The tube was shaken vigorously
for 1 hr. It was then centrifuged and 1.0 ml of the supernatant

fluid was removed for analysis. The method of additions (Perkin-
Elmer 1977) was used to minimize matrix interference.

Water samples (100 ml, not filtered) were adjusted to pH 2.
A 15.0-ml quantity was then extracted with 10 ml of 15 percent
(wt/vol) ammonium pyrrolidine dithiocarbamic acid (APDC) in dis-
tilled water. The APDC solution was prepared by dissolving APDC
in distilled deionized water which had been adjusted to pH 7
with NaOH. The APDC solution was then extracted three times with
CC14 to remove impurities. The water sample, containing APDC,
@qas stirred for 10 min. It was then extracted with a minimum
volume of methyl isobutyl ketone (MIBK). The mixture was agita-
ted for 1 min and then allowed to separate for 10 min. The MIBK
phase was analyzed for tin.

organic tin--
The organic acid solution from bioflasks was analyzed di-
rectly for tin content.

Medium from Hungate tubes was extracted in such a way that
inorganic tin compounds remained behind while organotin com-
pounds were extracted. Extensive preliminary experiments showed
that this extraction was effective. Medium was centrifuged at
3000 x g for 10 min to remove cells and sediment. The super-
natant medium was then extracted with 2.0 ml of dichloromethane:
chloroform (9:1, vol/vol). The extraction was conducted over a
period of 1 hr, with periodic agitation. The mixture was then
centrifuged at 1,000 x g for 10 min to sediment remaining in-
organic tin precipitate. The lower, organic phase was removed
and evaporated by dryness under a stream of nitrogen. The re-
sidue was dissolved in MIBK and analyzed for tin. Evaporation of
the organic phase under nitrogen may have removed some organotin
material, since organotins are volatile. Thus, results from Hun-
gate tubes are qualitative, not quantitative.

Atomic absorption spectrophotometry--
Tin was analyzed with the aid of an HC-2200 graphite furnace
with ramp accessory on a Perkin-Elmer model 503 AA unit using D2
background correction (to minimize interference from "smoke" and
nonatomic absorption) and an Sn electrodeless discharge lamp
(EDL). Argon was the sheath gas.

Mineral acid solutions from sediment samples and organic
acid solutions from bioflasks were analyzed with the EDL set at
286,3 nm. Samples were dried at 105 C for 40 sec (10 sec ramp)f
charred at 800 C for 40 sec (10 sec ramp), followed by atomiza-
tion using maximum power at 2700 C for 8 sec with a 3-sec stop
of gas.

MIBK from water samples and from biotubes was analyzed with
the EDL set at 224.5 nm. Samples were dried at 110C for 40 sec

(10 sec ramp), charred at 650 C for 40 sec (no ramp), followed by
atomization at 2500 C for 8 sec with a 3-sec stop of gas.

RESULTS AND DISCUSSION

Partially Developed Methods

One of the objectives of this work was to apply quantitative
methods developed for other metals and organometallic compounds
to tin and orgAnotin compounds. The approach was to separate
organotin compounds from one another via gas liquid chromato-
graphy (GLC) and to allow the effluent from the GLC to pass
through a heated transfer tube to an atomic absorption spectro-
photometer (AA) which would detect tin. Thus, those compounds
which contained tin would be detected by the AA. The method is
in use in several laboratories for other metals (Brinckman et
al. 1976, Parris et al. 1977, Trachman et al. 1977) and contact
was maintained with workers at the National Bureau of Standards
throughout the project. Due to time pressures, further develop-
ment of the method was stopped before it was ready for use on
this project. Success was achieved in separating mono-, di-,
tri-, and tetra-methyltin via GLC, although the system is not
yet sensitive enough for direct application to environmental
samples. The AA unit is close to being used as a detector for
organotin compounds eluted from the GLC. Not the least of the
difficulties was an 11-month delay in receiving the heated
transfer line from the sole manufacturer. Work on this aspect
of the project is continuing and*support of this contract will
be acknowledged in all publications.

Physical and Chemical Data on Samples

Physical and chemical data for samples taken on cruises and
excursions in April 1978 and July 1979 are summarized in
Tables 20 and 21. Temperatures and salinities were as expected
for these sites in spring and summer seasons, respectively. pH
values, which were,taken only for the summer 1979 samples, were
significantly higher at the freshwater sites than at the estu-
arine sites; they were highest at the two Tenneco sites. Values
for dissolved oxygen were higher in April 1978 than in July 1979,
as expected for spring and summer seasons, respectively. It is
noteworthy that in summer 1979, dissolved oxygen was dangerously
low near the bottom at the Buoy Rock site, which did not suffer
extensive oyster mortality; while dissolved oxygen was only
slightly higher at Spaniard Bar. Both oyster bars gave lower
summer readings for dissolved oxygen than were found in Balti-
more Harbor. A very high reading was obtained at the Tenneco
pond in July 1979. A comparison of physiochemical data for the
healthy bar, Buoy Rock, and the data for Spaniard Bar, show only
dangerously low dissolved oxygen near the bottom as a potential-
ly lethal condition. None of the physiochemical data from the
freshwater sites gives a direct clue to the extensive kill of

TABLE 20. PHYSICAL DATA FOR APRIL 1978 CRUISE/EXCURSION
Station/Site Depth* Temperature Salinity Dissolved
(meters) (0 C) Oxygen
(parts per
thousand)
Buoy Rock 1 10.8 4.4 10.6
6 11.7 5.2 10.7
12 7.7 8.1 10.2

Spaniard Bar 1 12.3 4.6 10.3
5 12.0 5.5 10.3
10 12.1 5.4 10.2

Tenneco - effluent - 20 3 9.1
Tenneco - pond 0.2 17 2 10.1
Campbell plant 0.2 18 1 7.0

Chestertown sewage
treatment plant 0.2 19 .1 9.4

TABLE 21. PHYSICAL DATA FOR JULY 1979 CRUISE/EXCURSION
Station/Site Depth* T@@!mperature pH Salinity Dissolved
(meters) (0 C) Oxygen
(parts per
thousand)
Buoy Rock 1 22.6 6.5 6.2
6 22.5 6.6 6.0
12 22.0 6.6 6.8 4.5
Spaniard Bar 1 24.0 5.6 7.4
5 23.9 5.6 7.4
10 23.9 6.1 5.6 7.3

Baltimore Harbor 1 23.2 3.5 7.9
2.5 22.8 3.5 8.0
5 22.7 5.2 3.8 7.6
Tilghman Island .1 22.5 8.7 9.4
5 22.6 8.8 9.6
10 22.6 5.5 9.1 9.3

Tenneco - effluent - 30.0 9.6 2.0 8.4
Tenneco - pond 0.2 24.9 8.4 2.0 ')20
Campbell plant 0.2 21.0 7.2 2.0 9.5

Chestertown sewage
treatment plant 0.2 24.0 7.6 2.0 9.4

The deepest reading at each point was taken at approximately
1 m from the bottom.

benthic fauna in the Chester River. It is interesting that dis-
solved oxygen values for Baltimore Harbor are lower than for the
site at Tilghman Island and only slightly higher than the values
at Spaniard Bar.

Data for Microbial Populations

Data for enumeration of bacteria are shown in Tables 22 and
23. Counts in sediment are higher than counts in water, as ex-
pected. In general, estuarine sites showed higher bacterial
counts in summer than in spring; this is expected as a function
of temperature and of increased production of organic materials
in the estuary in the summer.

TABLE 22. VIABLE COUNTS OF BACTERIA FROM APRIL 1978 CRUISE
Station/Site Type of Total Reg-11-stan-F-t-0 %
Sample Viable Inorganic-Sn Resistant
Sample
Buoy Rock Water 1.2xlo2 1.6xlo2 133
Sediment 2.OxlO5 1.3xlo4 6
Spaniard Bar Water <1.2xlo2 <1.2xlo2
Sediment 2.4xlO5 1.2xlo5 50

Tenneco - Water 2.8x,04 2.ox,04 10
effluent

Tenneco Water l.lxlo4 2.2xlO3 20
pond Sediment 2.4xlO6 3.7xlO5 16
Campbell Water 6.4xlO3 3.2xlO3 50
plant Sediment 4.lxlO7 2.lxlO7 51

Chestertown Water 1.4xlO2 <1.2xlO2
sewage Sediment 5.8xlO5 3.2xlO5 55
treatment
plant

Data not accurate enough to yield a useful percentage.

mmmm MMMMMMAMM Momm

TABLE 23. VIABLE COUNTS OF BACTERIA FROM JULY 1979 CRUISE
Station/Site Type of Total Number Resistant to % Resistant to
Sample Viable Inorganic-Sn Organic-S.n. !-@rganic-Sn Organic-Sn
Count
Buoy Rock Water 5.OxlO 1 <1.5xlO1 0 0
Sediment 2.3xlO 6 3.4xlO5 1.6X104 15 1
Spaniard Bar Water 7.3xlO 2 <1.5xlO1 0 0
Sediment 7.8xlO 5 2.8xlO5 1.4xlO4 36 2
Baltimore Water 3.3xlO 4 9.5xlO3 5.8xlO2 29 17
Harbor Sediment 6.OxlO 5 2.3xlO5 1.3xlO4 40 2
Tilghman Water 3.lxlO 3 1.5xlO1 0 .40.5 0
Island Sediment 3.2xlO 5 6.OxlO4 0 19 0

5 4 4
Tenneco - Water 1.2xlO 4.OxlO l.lxlO 33 9
effluent
Tenneco - Water 1.5xlO 4 7.2xlO3 6.6xlO3 50 4
pond Sediment 9.2xlO 5 l.lxlO5 9.5xlO4 12 10
Campbell Water 1.3xlO 2 <1.5xlO1 0 <8 0
plant SIediment 8.lxlO 6 9.31x1o5 5.2xlO5 12 6
Chestertown Sediment 5.6xlO 6 3.lxlO6 5.5xlO5 55 10
sewage treat-
memt plant

In most cases, a significant fraction of the microbial popu-
lation was resistant to inorganic tin and is, therefore, poten-
tially capable of metabolizing tin to more toxiccompounds. A
much smaller fraction of the population was resistant to the
organotin compound, dimethyltin chloride, attesting to the anti-
bacterial properties of organic tin compounds. The data are con-
sistent with the hypothesis that aquatic microorganisms can
protect themselves against toxic tin compounds by transforming
tin to organotins which, although toxic in themselves, are vola-
tile and leave the immediate vicinity of the cell which formed
them.

Sediment from Spaniard Bar did not contain higher numbers of
tin-resistant organisms than sediments from Buoy Rock, although
a higher percentage of the population was resistant to tin at
Spaniard Bar than at Buoy Rock. Sediment samples taken at the
Chestertown sewage treatment plant contained a high percentage
of organisms resistant to inorganic tin. There is little indi-
cation of a higher level of tin-resistant microflora at Spaniard
Bar than at Buoy Rock. The data available do not suggest that
the freshwater sites contain higher numbers of tin-resistant or-
ganisms than the estuarine sites.

Data for production of volatile tin compounds are summarized
in Table 24- The method used was effective in detecting organic
tin compounds produced in bioflasks, as indicated by values ob-
served from sterile medium which contained dimethyltin chloride.
Variation among replicates indicates that the method is qualita-
tive. Flasks which received inoculum containing the metabolic
poison sodium azide yielded no volatile tin, indicating that
volatile tin detected was the result of biological activity.
Results from bioflasks and Hungate tubes demonstrate that each
site contains microorganisms capable of converting inorganic tin
to volatile organotin(s). The species of organotin produced were
not identified.

Tin in Water and Sediments

The lower limit of sensitivity for tin analysis was 2 ppb.
A recovery value of 77.8 percent was obtained when sediment from
Tilghman Island was spiked with SnC14- When salt water or
Nelson's liquid medium was spiked with SnC14, recoveries of
94-96 percent were obtained.

Sediment samples contained more tin than water samples
(Table 25), as expected. Sediment from Baltimore Harbor, known
as a polluted site, contained over 200 ppm tin (;>0.02 percent on
a wet weight basis)'. In contrast, sediments from the Tilghman
Island site contained less that 1 ppm (<0.0001 percent on a wet
weight basis). All sediments associated with the Chester River,
including sediments from the three freshwater sites, yielded more
tin than sediments from the Tilghman Island site (Table 25).

TABLE 24. PRODUCTION OF VOLATILE TIN IN MEDIA INOCULATED WITH
SEDIMENT

Result from
source of Bioflask Hungate tube
inoculum. Replicate 09 Sn in
No. organic
acid
solution*

None, sterile control 1 148 t
with medium contain- 2 45
ing dimethyltin
chloride

Buoy Rock 1 21 +
2 0

Spaniard Bar 1 262 ++
2 57

Baltimore Harbor 1 344 +
2 145

Tilghman Island 1 658 +
2 0

Tenneco pond 1 361 ++
2 314

Campbell plant 1 218 +
2 144

Chestertown sewage 1 176 ++
treatment plant 2 201

Corrected for values obtained from sterile flasks containing
inorganic tin.

t The notation used represents peak height from the atomic
absorption unit: -, 0.0; +, 0.1 to 0.2; ++, 0.3 to 0.6;
+++, 0.6 to 1.0; ++++, 1.0 and above.

TABLE 25. TIN IN WATER AND SEDIMENT SAMPLES
Dg/ml or pg/q for samples taken in
Station/Site Type of April 1978 July 1979
Sample Mean Standard Mean Standard
error error

Buoy Roc k Water <0.002 - <0.002 -
Sediment 1.574 0.180 7.882t 0.088

Spaniard Bar Water 0.240 0.007 <0.002 -
Sediment 9.686 0.420 3.061T 0.187

Baltimore Water <0.002
Harbor Sediment 239.633* 14.073

Tilghman Water <0.002 -
Island. Sediment 0.861� 0.052

Tenneco Water 0.034 0.002 0.052 0.002
effluent

Tenneco - Water 0.103 0.005 0.023 0.001
pond Sediment 0.043 0.081 3.523T 0.046

Campbell water 0.032 0.001 <0.002 -
plant Sediment 1.583 0.131 3.876T 0.028

Chestertown Water 0.002 - 0.152 0.028
sewage Sediment 2.857 0.060 5.178t,T 0.028
treatment
plant

* t, �- Sediment samples values with identical superscripts
w@re Jt' significantly different at the 5 percent confidence
level.

Smith and Bur ton (1972) reported values of <0.05 to 16 ppm in
various marine sediments. Surface sediments from Narragansett
Bay contain 20 ppm tin (Hodge, Seidel, and Goldberg 1979). Furr
et al. (1976) reported values as high as 492 ppm tin in sewage
sludge. Thus, values obtained for sediments in the present study
are in the same range as reported by other workers. Sediments in
the Chester River, in the Tenneco holding plant, near the Camp-
bell plant, and in the Chestertown sewage treatment plant all
contain significant quantities of tin (Table 25). Based on the
two samples taken, Spaniard Bar, which suffered an oysterkill,
did not yield significantly more tin than Buoy Rock, which did
not suffer such a kill (Table 25). The null hypothesis that
mean tin concentration (Spaniard Point) did not differ from the
mean tin concentration (Buoy Rock) was supported. Thus, it is
not possible to attribute the oysterkill at Spaniard Bar solely
to pollution by tin, although interaction of tin with other fac-
tors cannot be excluded. It is also possible that different tin
species are present at the two sites, e.g. inorganic tin vs. one
or more organotin species.

Smith and Burton (1972) reported inorganic tin concentrations
of 0.02 to 0.04 pg/kg (0.002 to 0.04 parts per trillion) in
estuarine and continental shelf waters. Hodge et al. (1979) re-
ported 2 to 38 ng inorganic tin/liter(O.002 to 0.038 @ig/kg, 0.002
to 0.038 parts per trillion) in water from San Francisco Bay and
San Diego Bay, and 84 to 490 pg/liter (0.084 to 0.490 pg/kg,
0.084 to 0.490 parts per trillion) in water from Lake Michigan.
In Lake Michigan water, the concentrations of,organic tin com-
pounds were two to eight times greater than the concentrations
of inorganic tin. The methods used in these two studies were
several orders of magnitude more sensitive than the method
employed in the present work. Some of the values obtained for
water samples in the present work (Table 25) are significantly
higher than values reported by Smith and Burton (1972) and by
Hodge et al. (1979). Variations from one sample time to another
(e.g., values for water from Spaniard Bar and water from the Ten-
neco pond) can be expected due to variations in input and in flow
rate, which can also account for the low value in the water sam-
ple from Baltimore Harbor. Our data suggest that water in the
Chester River and water entering the Chester River from the Ten-
neco plant, from the Campbell plant, and from the Chestertown
sewage treatment plant sometimes contains significant quantities
of tin. The chemical species of tin were not identified in the
present study.

The data obtained in the present study yielded reasonable
recovery from spiked samples. The method of additions was used
successfully to alleviate matrix interference. Thus, the values
obtained for tin are regarded with confidence, even though they
are high.

Statistical Analyses

Data for bacterial enumerations, and data for tin levels were
analyzed using a one-way analysis of variance (ANOVA).. Sample
means were compared using the studentized (SNK) multiple range
test.

Correlation coefficiencts (r) were analyzed for significance
of association between tin concentration and counts of tin-re-
sistant microorganisms; and for significance of association
between tin concentration and percent of the microbial popula-
tion resistant to tin. All associations were tested at the
5 percent level of significance. No significant positive or
negative correlations were noted. Thus, the data available do
not suggest that tin (organic and/or inorganic) selects for a
population of tin-resistant microorganisms.

General Discussion

The data indicate that surface sediments from oyster bars in
the Chester River contain significantly more.tin than is con-
tained in sediments from a site near Tilghman Island, but they
contain significantly.less tin than sediment from Baltimore
Harbor. The three freshwater sites examined--the Tenneco plant,
the Campbell factory and the Chestertown sewage treatment plant--
all contribute tin to the Chester River ecosystem at an undeter-
mined rate. There may be other contributors., It is clear that
microorganisms capable of converting inorganic tin to more toxic
organotin compounds are ubiquitous in the Chesapeake ecosystem
(Table 24) .

Zuckerman et al. (1978) reviewed the known toxic effects of
organotin compounds on a variety of living organisms. Relative-
ly little is known of effects on estuarine organisms, but the
following toxic levels have been reported for organisms relevant
to the present study: guppies, less than 1 ppm of bis(tributyl-
tin)oxide and 0.1 ppm triphenltin hydroxide; molluscs, 1.0 ppm
of several trialkyltin and 0.05 - 0.10 bis(tributyltin)oxide;
algae, barnacles, shrimp and tubeworms, 0.1 to 1.0 ppm of tri-
butyl and triphenyltin compounds. Very little is known of the
levels of organotin compounds in estuarine systems. But sedi-
ments in the Chester River ecosystem contain between 3'and
8 ppm of tin (Table 25). Data from the present study do not in-
dicate the chemical speciation of the tin compound(s) detected.
Although the chemical species of tin compound(s) was not estab-
lisi@ad, if even one-third of the tin is present as organotin
compounds, the estuarine biota could be at risk, particularly
if stress from toxic tin(s) is coupled with other stresses such
as low dissolved oxygen.

Comparisons of tin in surface sediments from Spaniard Bar
with data from Buoy Rock and with data from other marine sedi-
ments (Smith and Burton 1972, Hodge et al. 1979) indicate clearly
that tin is not the sole source of the oysterkill at Spaniard
Bar. However, if oysters and other benthic organisms were stres-
sed by other factors, e.g. low dissolved oxygen, tin could con-
tribute additional stre ss leading to death.

Little is known about bioaccumulation of tin by benthic
invertebrates. Until recently, analytical methods were not
sensitive enough to permit one to deal with tin in the same
quantitative manner as with other inorganic pollutants such as
copper, cadmium, mercury, and zinc. These include aspects of
levels of tin and organotins in the environment, accumulation of
tin compounds in the food chain, chemical and biological forma-
tion and transformation of inorganic tins and organotins, trans-
port of tin compounds in soils and in aquatic ecosystems, and
toxicity of tin compounds to the biota. The appropriate ques-
tions should be asked regarding tin. Moreover, it is possible
that the gut flora of oysters and other invertebrates can con-
vert inorganic tin to more toxic organotins; these questions
should be addressed.

100

BIBLIOGRAPHY

Brinckman, F. E., W. P. Iverson, and W. Blair. 1976. Approaches
to the study of microbial transformations of metals, pp. 919-
936. In: J. M. Sharpley and A. M. Kaplan (Eds.) Proceedings
of the Third International Biodegradation Symposium. Applied
Science Publishers, London.

Daum, R. J. 1965. Agricultural and biocidal applications of
organometallics. Ann. N. Y. Acad. Sci. 125:129.

Deschiens, R., and H. Floch. 1962. Comparative study of the
molluscidial action of 5, 2'-dichloro-4'-n itro-salicyl-
anilide and salts of triphenyl tin. Bull. Soc. Pathol.
Exot. 55:816.

Furr, A. K., A. W. Lawrence, S. S. C. Teny, M. C. Grandolfe,
R. A. Hofstader, C. A. Bache, W. H. Gutenman, and D. J. Lisk.
1976. Multielement and chlorinated hydrocarbon analysis of
municipal sewage sludges of American cities. Environ. Sci.
Technol. 10:683-687.

Hodge, V. F., S. L. Seidel, and E. D. Goldberg. 1979. Deter-
mination of tin (IV) and organotin compounds in natural
waters, coastal sediments and macro algae by atomic spectro-
metry. Anal. Chem. 51(8):1256-1259.

Holden, A. W. 1972. The effects of pesticide s on life in fresh
waters. Proc. R. Soc. Lond. 180:383.

Huey, C., F. E. Brinckman, S. Grim, and W. P. Iverson. 1974.
The role of tin in bacterial methylation of mercury. Proc.
Int. Conf. on Transport of Persistent Chemicals in Aquatic
Ecosystems II, pp. 7.3-78. Natl. Res. Council, Ottawa,
Canada.

Hungate, R. E. 1969. A roll tube method for cultivation of
strict anaerobes, pp. 117-132. In: J. R. Norris and D. W.
Ribbons (Eds.) Methods in Microb@iology, Vol..3B. Academic
Press, London.

Kimmel, E. C., R.-H. Fish, and J. E. Casida. 1977. Bioorganotin
chemistry, metabolism of organotin compounds in microsomal
monooxygenase systems and in mammals.
J. Agr. Food Chem. 25(l):1-9.

101

Lloyd, R. V., and M. T. Rogers. 1973. Electron spin resonance
study of some silicon-# germanium-, and tin-centered radi-
cals. J. Am. Chem. Soc. 95(8):2459-2464.

Luijten, J. G. A. 1972. Applications and biological effects of
organotin compounds, pp. 931-932. In: A. K. Sawer (Ed.)
Organotin Compounds. Marcel Dekker-,New York.

Nelson, Jr., J. D., W. Blair, F. E. Brinckman, R. R. Colwell, and
W. P. Iverson. 1973. Biodegradation of phenylmercuric ace-
tate by mercury resistant bacteria.
Appl. Microbial. 26:321-326.

Parris, G. E., W. R. Blair, and F. E. Brinckman. 1977. Chemical
and physical considerations in the use of atomic absorption
coupled with a gas chromatograph for determination of trace
organometallic gases. Anal. Chem. 49:378-384.

Perkin-Elmer. 1977. Analytical methods for atomic absorption
spectoscopy using the HGA graphite furnace.
Perkin-Elmer Corp., Norwalk, CT.

Ridley, W. P., L. J. Dizikes, and J. M. Wood. 1977. Biomethyla-
tion of toxic elements in the environment.
Science 197(4301):329-332.

Smith, J. D., and J. D. Burton. 1972. The occurrence and dis-
tribution of tin with particular reference to marine environ-
ments. Geochim. Cosmochim. Acta 36:621-629.

Subramanian, R. V. 1978. Recent advances in organotin polymers.
Poly.-Plast. Technol. Eng. 11(l):81-116.

Thayer, J. S. 1974. Organometallic compounds and living
organisms. J. Organometal. Chem. 76:265-2.95.

Trachman, H. L., A. J. Tyberg, and P. D. Branigan. 1977.
Atomic absorption spectrometric determination of sub-part-
per million quantities of tin in extracts and biological
materials with a graphite furnace. Anal. Chem. 49:1090-1093.

Van der Kerk, G. J. M. 1976. Organotin chemistry:past, present,
and future, p. 2. In: J. J. Zuckerman (Ed.) Organotin Com-
pounds: New Chemisti-y and Applications. American Chemical
Society, Washington, D. C.

Zuckerman, J. J., R. P. Reisdorf, H. V. Ellis III, and R. R.
Wilkinson. 1978. Organotins in biology and the environment,
pp. 388-424. In: F. E. Brinckman and J. M. Bellama (Eds.)
Organometals ai@_d organometalloids, occurence and fate in the
environment. American Chemical Society, Washington, D. C.

102

APPENDIX A

CHESTER RIVER OYSTER MORTALITY

A review of recent Fisheries Administration catch records
and oyster bar surveys of the upper Chester River indicates a
significant decline in the populations of both oysters and
associated organisms.
A major oyster mortality occurred during the spring and
summer of 1973 in Langford Creek and the Chester River above
Oldfield oyster bar. The 1974 fall survey showed the mor-
tality was continuing as Oldfield had died-off and some mor-
tality had occurred on Piney Point bar. No live oysters were
found above Hells Delight bar during the 1975 oyster survey.
These mortalities caused oyster landings from the area above
Piney Point bar to decline from 50,000 bushels in 1972-1973
to 650 bushels for the 1975-1976 season.

The following is the text of a 8 June 1978 letter sent to
us by George E. Krantz of the Horn Point Environmental Labora-
tories:

"Please excuse the long delay in sending you some data on
the Chester River mortality. Unfortunately my files were loaned
to other investigators who removed many of the original documents
that I described to you during our phone conversation last month.
I think I have found copies of most of the information but I was
unable to find a complete briefing document for you. Perhaps
some of the data, especially from the Dept. of Natural Resources
will be helpful in your study. At a later date we may be able
to more thoroughly discuss the observations that may have ex-
isted in this historical phenomenon."
Data on pages 103 thru 110 were compiled by Roy Scott, DNR,
from field data sheets reflecting Fall oyster bar survey results.
Time of survey of specific bars varied from October thru March of
a given year.

103

1970. Continuation of ltr. from G. E. Krantz of June 8, 1978
Bar Markets* Smalls* Spat* Boxes*

Ferry 216/75 90/8 2 4

Ferry 174/65 236/30 2 8

Side Shoal 42/20 762/80 2 2

Buoy Rock 198/70 74/15 28 2

Buoy Rock 188/60 43/15 4 0

Buoy Rock 110/55 40/20 0 6

Blunt 50/9 0 2 58

Blunt 180/65 244/30 0 0

Blunt 266/90 12/4 4 .0

Hail Point 114/50 208/30 6 2

Hail Point 170/70 26/5 0 .0

Hail Point 182/85 12/4 0

Hail Point 98/60 12/2 2 2

Carpenters Island 124/40 530/55 4 4

Carpenters Island 236/75 28/3 4 4

Durdin 98/30 406/45 4 10

Durdin 124/40 152/50 8 10

Durdin 36/40 22/1 2 2

Horserace 182/95 6 2 0

Horserace 110/45 414/40 4 0

Piney Point 202/70 38/10 0 4

Bay Bush Point 160/70 56/10 6 12

Hells Delight 158/60 4 0 0

Hells Delight 282/90 4 0 4
(continued)
104

1970. Continuation of ltr. from G. E. Krantz of June 8, 1978.

Bar Markets* Smalls* Spat* Boxes*

Bluff Point 212/80 28/2 0 6

Middlegrounds 230/85 16/2 0 10

Oldfield 4 686/85 0 16

Oldfield 216/75 16/3 0 8

Oldfield 156/50 22 6 0

Willow Bottom 116/55 10/2 0 10

Hudson 186/50 30/6 4 8

Hudson 174/30 494/60 0 8

Boathouse 78/25 234/35 0 6

Drum Point 55/35 20/3 0 0

Davis Creek 142/60 30/3 0 4

Ebb Point 100/45 306/38 4 2

Ebb Point 132/40 117/12 2 2

Spaniard Point 148/30 730/60 0 4

Spaniard Point 4 908/100 0 0

Cliff 72/30 400/60 0 6

Emory Hollow 20/15 2 0 .2

Mummys Cove 116/65 4 2 4

Shippen Creek 132/70 0 0 4

Shippen Creek 170/50 18/3 0 16

Number/Percentage of 1 oyster bushel.

105

1973. Continuation of ltr. from G. E. Krantz of June 8, 1978.
Bar Markets* Smalls* Spat* Boxes* Associated
organisms
Present t

Buoy Rock 152/70 80/10 0 0 0

Hail Point 162/90 22/4 2 2 X

North of Hail Pt. 108/50 36/7 0 4 X

Carpenters Island 114/50 16/2 0 4 X

Durdin 188/65 80/7 0 8 X

Horserace 118/50 58/10 0 8 0

Piney Point 54/30 0 0 12/6 X

Bay Bush Point 112/55 8/<l 0 18/8 X

Bluff Point 54/40 2 0 20/25 X

Oldfield 12/8 2/<l 0 52/40 X

Oldfield 104/70 2 0 10/4 X

Oldfield 128/60 12/<l 0 14/6 X

Oldfield 58/40 8/<l 0 20/23 X

Oldfield 58/40 38/10 0 46/12 0

Oldfield 16/5 4 0 60/40 X

Willow Bottom 0 0 0 122/18-5 X

Nichols 0 0 0 126/75 X

Holton Point 0 0 0 82/55 X

Ebb Point 2 0 0 64/45 X

Spaniard Point 0 2 0 142/45 X

Spaniard Point 0 0 0 154/80 X

Cliff 0 2 0 130/50 X

Cliff 0 0 0 112/45 X

(continued)

106

1973. Continuation of 1tr. from G. F. Krantz of June 8, 1978.

Bar Markets* Smalls* Spat* Boxes* Associated
organisms
Present t

Sheep 0 0 0 26/20 X

Mummy Cove 0 0 0 50/25 X

Shippen Creek 0 0 0 96/25 X

Number/Percentage of 1 oyster bushel.
t 0 no organisms detected; X organisms present.

107

1974. Continuation of ltr. from G. E. Krantz of June 8, 1978.

Bar Markets* Smalls* Spat* Boxes* Associated
Organisms
Present t

Buoy Rock 184/80 102/16 0 2 0

Blunt 134/55 4/1 0 12/3 X

Hail Point 90/40 6 0 20/6 X

Carpenters Island 80/30 6/1 0 14/2 X

Durdin 108/53 12/1 0 26 X

Horserace 72/40 4/<1 0 10/10 X

Piney Pt. (Lower) 90/40 38/8 0 22/5 X

Piney Pt. (Upper) 84/30 10/1 0 46/15 X

Bay Bush Point 56/20 72/15 0 22/5 X

Hells Delight 10/5 0 0 16/10 X

Bluff Point 54/40 2 0 20/25 X

Oldfield 0 0 0 68/25 X

Willow Bottom 0 0 0 7 4/25 X

Nichols 0 0 0 30/20 X

sand Thistle 0 0 0 60/15 X

Boat House 0 0 0 88/60 X

Drum Point 0 0 0 28/12 X

Holton Point 0 0 0 82/55 X

Ebb Point 0 0 0 68/25 X

Spaniard Point 0 0 0 132/50 X

Cliff 0 0 0 60/25 X

Emory Hollow 0 0 0 0 X
(continued)

108

1974. Continuation of ltr. from G. E. Krantz of June 8, 1978.

Bar Markets* Smalls* Spat* Boxes* Associated
organisms
Present t

Emory Hollow 0 0 0 0 X

Mummys Cove 2 0 0 30 0

Mummys Cove 0 0 0 42/25 X

Shippen Creek 2 0 0 22/10 X

Shippen Creek 0 0 0 56/40 X

Number/Percentage of one oyster bushel.

t 0 no organisms detected; x organisms present.

109

1975. Continuation of ltr. from G. E. Krantz of June 8, 1978.

Bar Markets* Smalls* Spat* Boxes* Associated
organisms
Present +

Strong Bay 162/60 4/<l 0 8/4 X

Ferry 170/80 6/<l 0 16/8 X

Ferry 84/15 0 0 20/6 X

Ferry 184/90 6 0 10/6 X

Side Shoal 196/90 12/4 0 8/4 0

Buoy Rock 150/65 70/12 0 2/<l X

Blunt 204/85 0 0 16/10 X

Blunt 128/50 4/1 0 20/12 X

Blunt 108/55 4/<l 0 12/6 X

Hail Point 188/80 14/2 0 8/4 X

North of Hail Pt. 148/45 16/1 0 10/8 X

Carpenters Island 122/45 20/5 0 30/12 X

Carpenters Island 160/50 22/4 0 10 X

Carpenters Island 118/75 4/1 0 4/2 X

Carpenters Island 140/45 54/11 0 14/12 X

Carpenters Island 238/50 21/6 0 0 X

Horserace 138/60 24/3 0 6/3 X

Horserace 138/50 60/12 0 19 X

Piney Point 70/25 102/18 0 8 X

Piney Point 118/50 54/10 0 30/25 X

Bay Bush Point 8 52 0 14 0

Oldfield 0 0 0 5 0

Oldfield 0 0 0 15 X
(continued)

110

1975. Continuation of ltr. from G. F. Krantz of June 8, 1978.
Bar Markets* Smalls* Spat* Boxes* Associated
organisms
Present t

Oldfield 0 0 0 15 X

Willow Bottom 0 0 0 3 X

Nichols 0 0 0 10 X

Drum Point 0 0 0 5 0

Holton Point 0 0 0 48/20 X

Spaniard Point 0 0 0 94/22 X

Cliff 0 0 0 18/6 0

Number/Percentage of 1 oyster bushel.
t 0 indicates no organisms detected; X organisms present.

APPENDIX B

ACCOUNT OF INTERLABORATORY TESTS ON SPLIT SAMPLES

This is a chronological account of interactions with Environ-
mental Monitoring and Support Laboratory (EMSL), Office of Re-
search and Development, U.S. Environmental Protection Agency,
Cincinnati, Ohio. Contact: William L. Budde.

October 1978

Two dried sediment samples were each split and respective
subsamples were sent to EMSL in Cincinnati. They were labeled:
(1) Chester River, mouth, top sediment; (2) Tenneco Pond sediment
(>100 ppm dioctyl phthalate). The EMSL facility was selected due
to its fine reputation in the scientific community.

January 1979

EMSL advised this laboratory of their results on the two
samples: (1) 61 ppm. of DEHP in the Chester River, mouth, and
(2) 1700 ppm. DEHP in the Tenneco Pond. No error values nor blank
values were given. In a telephone conversation prior to this re-
port, EMSL communicated a preliminary verbal report indicating
6 ppm of DEHP in the Chester River, mouth sediment. (After dis-
covering a computational error, the corrected value was described
as 58 ppm.) The value this laboratory had determined--prior to
receiving the EMSL report--was 0.6 ppm DEHP in Chester River,
mouth sediment and 1200 + 100 ppm DEHP in Tenneco Pond sediment.

Due to the large discrepancy (2 orders of magnitude) this
laboratory undertook an extensive battery of experiments to ex-
plore previously undetected systematic errors, and to find a
possible reason for its low value for DEHP in Chester River,
mouth sediment. After 2 months of experimentation, we continued
to obtain values less than 1 ppm. In fact, improvement of our
methodology produced a value lower than determined previously
0.1 ppm.

April 1979

Agreement for retrial was secured and a second set of split
samples was sent to EMSL Cincinnati. This time three samples
were sent: (1) Dry Chester River mouth sediment, (2) Wet Chester
River mouth sediment, which EPA dried themselves, and (3) "or-
ganic free" clay which we had carefully spiked with DEHP.

112

June 1979

EMSL returned the second set of results and reported a
drastically lowered result for Chester River, mouth.

EPA-Cincinnati Univ. Maryland
(1) Chester River, mouth 0.3 mg/kg 0.097+0.03 ppm,
dried by U. of Maryland DEHP DffH_P

(2) Chester River, mouth 0.4 mg/kg
dried by EPA DEHP

(3) Attapulgite, spiked 121 mg/kg 200+10 ppm,DEHP
DEHP
In a telephone conversation prior to this report, EMSL commented
that during the time of the first exchange, EPA Cincinnati had
been experiencing some contamination problems with their homo-
genizer used in sediment extraction. This is a probable explana-
tion for the excessively high values found by EMSL in the origin-
al exchange. In the report of June 1979, signed by William
Budde, it was mentioned that EMSL procedures were the same as
before, except that they had used a new Tissumizer (R) by Tekmar.

While there is still somedifference between the values de-
termined by each laboratory in this study, we believe they are
within acceptable limits especially considering the history of
trace organic interlaboratory comparisons (Hilpert et al., 1978,
Hertz et al., 1979).

Hertz, H. S., L. R. Hilpert, W. E. May, S. A. Wise, J. M Brown,
S. N. Chesler, and F. R. Guenther. Special Technical Publication
686, American Society for Testing and Materials, 1979.

Hilpert, L. R., W. E. May, S. A. Wise, S. N. Chesler, and H. S.
Hertz. interlaboratory comparison of determinations of trace
level petroleum hydrocarbons in marine sediment. Anal. Chem.
50:458-463, 1978.

113

APPENDIX C

METHODS FOR WATER QUALITY MEASUREMENTS IN TIN STUDY

Water temperature was measured with the temperature probe on
the YSI oxygen meter used to determine dissolved oxygen.
Measurement of pH was with a Coleman meter, and salinity was
measured using a Wheatstone conductivity bridge.

114

I
I
I
I
I
I
I
I
I
I
i
i
i
I
i
I
I
I
- 3 6668 14109 5044
I - r
I