In the wake of the Argo Merchant proceedings of a symposium held January 11-13, 1978

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

r3oastal Zone
Information
Center.

In the Wake of the Argo Merchant

Proceedings of a Symposium
Held January 11 -135 1978, at the
Center for Ocean Management Studies,
University of Rhode Island

GC
1212
.m4
15
1978,

The Center for Ocear Management Studies was created in the fall of
1976 for the purpose of promoting effective coastal and ocean management,
COMS The Center identifies ocean management issues, holds workshops and
conferences to discuss these issues, and develops recommendations and
research programs to resolve them.

Cl OASTAL ZONE
INFORMATION CENTER

In the Wake of the Argo Merchant

Proceedings of a Symposium Held January 11-13, 1978

Center for Ocean Management Studies
University of Rhode Island
August, 1978

property of CSC Library

0 . S . DEPARTMENT OF COMMERCE NOAA
COASTAL SERVICES CENTER
2234 SOUTH HOBSON AVENUE
CHARLESTON SC 29405-2413

kz)

Preface

On December 15, 1976, the tanker Argo Acknowledgments
Merchant ran aground southeast of Nantucket Many people contributed ideas and information
Island. The subsequent discharge of its 7,700,000 to this effort. First and foremost, we would like to
gallons of No. 6 fuel oil resulted in one of the thank the members of the Symposium Planning
largest oil spills off the coast of the United States. In Committee, who developed the program and served
response to the spill, researchers from academic as session chairmen:
institutions and government agencies throughout Dr. Mason Wilson, Professor
the country conducted field studies and laboratory Department of Mechanical Engineepring and
experiments to assess the impact of the spill. This Applied Mechanics
document presents the results of those research University of Rhode Island
efforts and summarizes the discussions at a
symposium entitled "in the Wake of the Argo Dr. James Quinn, Professor
Merchant" held January 11 -13, 1978, at the Center Graduate School of Oceanography
for Ocean Management Studies, University of University of Rhode Island
Rhode Island. Mr. Kenneth Sherman, Director
The Center for Ocean Management Studies was Narragansett Laboratory
created in the fall of 1976 for the purpose of National Marine Fisheries Service
promoting effective coastal and ocean manage-
ment. It has provided a forum for interdisciplinary Dr. Paul Lefcourt, National Coordinator
communication, research, and education on marine Oil Spill Ecological Damage Assessment
management issues. We believe that the transpor- U.S. Environmental Protection Agency
tation of oil and the threat it poses to the environ- Special recognition must also be given to Dr. Eva J.
ment is a critical marine issue facing society today. Hoffman, URI marine research associate, who
Studies on the physical and chemical fate of oil initiated the idea for a symposium and helped the
spills, the biological effects, and the socio-economic Planning Committee identify the individuals who
impact resulting from people's. perceptions will had conducted research on the spill. We would also
enable us to respond more effectively to future spills like to thank the researchers, who made this effort
and will, we hope, mitigate their impact. In order to possible by pulling together their data -and pre--
achieve this, we must evaluate our research paring papers.
response to a spill such as that of theArgo Merchant Numerous individuals at the University pi@ovided
and discuss the implications of our research logistical and technical support in organizing the
results. This is why the Center sponsored the symposium and preparing the proceedings.
symposium, approximately one year after the spill. However, special recognition must be given to
The papers presented in the proceedings of this Carol Dryfoos, administrative secretary, Center for
symposium provide a comprehensive year's Ocean Management Studies, who handled the
documentation of the environmental and socio- logistics of the conference and the publication of
economic impacts of the spill. The results of the the proceedings. Vicki Desjardins, URI marine
physical and chemical studies, which were pre- editor, and Peter Brownell, of URI Printing Services,
sented on the first day of the symposium, comprise also deserve a special thanks for their combined
the first two sections of this report. Biological efforts in the editing and printing of this document.
impacts at the site and in surrounding areas were Funding for operational expenses was provided
discussed on the second day and are presented in by the National Oceanic and Atmospheric
the third section. Following the scientific presenta- Administration.
tions, disciplinary workshops were held to evaluate Virginia K. Tipple, Executive Director
laboratory and field techniques and recommend Center for Ocean Management Studies
appropriate scientific responses for future spills.
The discussions at the workshops are summarized
in the session chairman's introduction to each
scientific section. The socio-economic impacts of
the spill and response plans for future spills were
discussed on the last day and are presented in the
final sections of this report.
We hope that these proceedings will serve as a
historic document and assist various individuals
and agencies in responding to future spills.

contents

Physical Studies Histopathological Analyses of Benthic
Summary, Mason P. Wilson 3 Organisms From the Vicinity of the Argo
Merchant Wreck, Robert S. Brown and Keith
The Role of Physical Studies Before, During, R. Cooper 96
and After Oil Spills, Jerome H. Milgram 5 Some Physiological Effects of the Argo
Chronology of Events and Oil Slicks From the Merchant Oil Spill on Several Marine
Argo Merchant, James S. Mattson 15 Teleosts and Bivalve Molluscs, Frederick P.
The Behavior of Floating Oil From the Argo Thurberg, Edith Gould, and Margaret A.
Merchant, P. L. Grose 19 Dawson 103
Can Oil Spill Movement Be Predicted?, Ivan Observations on Argo Merchant Oil in
Lissauer and Pat Welsh 22 Zooplankton of Nantucket Shoals, R. Polak,
A. Filion, S. Fortier, J. Lanier, and
Risk Forecasting for the Argo Merchant Spill, K. Cooper 109
Timothy Wyant and Richard A. Smith 28 Field and Laboratory Measurements of Stress
Near-Bottom Transport in the Vicinity of the Responses at the Chromosome and Cell
Argo Merchant A Seabed Drifter Study, Levels in Planktonic Fish Eggs and the Oil
Barclay P. Collins, Clement A. Griscom, Problem, A. Crosby Longwell 116
and Eva J. Hoffman 34 Eff ects of the Water Soluble Fraction of a
Surface and Subsurface Spill Trajectory Venezuelan Heavy Fuel Oil (No. 6) on Cod
Forecasting: Application to the Argo Merchant Eggs and Larvae, W. W. Kuhnhold 126
Spill, Malcolm L. Spaulding 37 Interactions Between Petroleum and-Benthic
Oil Droplet Measurements Made in the Wake of Fauna at the Argo Merchant Spill Site,
the Argo Merchant, Peter Cornillon 43 Sheldon D. Pratt 131
Fish Predation on Oil-Contaminated Prey From
Chemical Studies the Region of the Argo Merchant Oil Spill,
Summary, James G. Quinn 51 Ray E. Bowman and Richard W. Langton 137
Eff ect of the Argo Merchant Oil Spill on Bird
Water Soluble Fraction of Argo Merchant Populations off the New England Coast,
Cargo, J. Richard Jadamec 53 15 December 1976-January 1977, Kevin D.
Hydrocarbon Chemistry of the Water Column of Powers and William Timothy Rumage 142
Georges Bank and Nantucket Shoals, The Argo Merchant Oi I Spi I I and the Fisheries,
February- Nove m ber 1977, Paul D. Boehm, Kenneth Sherman and Donna Busch 149
George Perry, and David Fiest 58
Where the Argo Merchant Oil Didn't Go, Socio-Economic Study
Chris W. Brown, Patricia F. Lynch, and Mark
Ahmadjian 65 Public Knowledge and Perceptions of the
Hydrocarbon Patterns in Some Marine Biota Effects of the Argo Merchant Oil Spill, Peter
and Sediments Following the Argo Merchant Fricke and John Maiolo 169
Spill, William D. MacLeod, Jr., Marianne Y.
Uyeda, Lawrence C. Thomas, and Donald W.
Brown 72 Future Response Plans
A Comparison of Argo Merchant Oil and Ecological Damage Assessment of Oil Spills-
Sediment Hydrocarbons From Nantucket the Federal Government's Response, Paul
Shoals, Eva J. Hoffman and James G. Quinn 80 Lefcourt 179

Biological Studies
Summary, Kenneth Sherman 91
Microscopic Observations on Vertebrates and
Invertebrates Collected Near the Argo
Merchant Oil Spill, Thomas K. Sawyer 93

Physical Studies

Mason P. Wilson, Chairman

Summary of Physical Studies

It is the physical and chemical physical processes that Surface Current Measurement Radar might be a
that control the dispersion of spilt oil into the marine useful tool for such work.
environment. An understanding of these processes is There is a great need to obtain more visual data on
fundamental to almost every aspect of control, cleanup, slick dynamics in the open ocean such as was obtained
and environmental assessment of oil spills. by the Navy-NOAA team on pancakes from the Argo
A discussion of these processes and how they Merchant spill. Better instrumentation is needed to obtain
affect the nature of the spill and the cleanup operation is slick thickness measurements to develop prediction
presented in the first paper of this session. The second models. The overflights were good which aided in making
paper lists the chronology of events and describes the the Argo Merchant spill probably the best documented
character and extent of the oil spill while the third paper major spil I to date, but there was a consensus that better
presents the results of a series of experiments concerned documentation would be needed in the future to obtain an
with the relative velocity of the slick with respect to the accurate budget of the fate of the oil.
surface waters. There are three papers that deal with the Better documentation and predictive methods on
forecasting of the movement of the surface slick, one of the fate of the oil could greatly improve the efficiency of
which also includes subsurface transport. Two types of chemical and biological investigations. It wasn't until well
forecasting (real time and risk type) were presented. A after the spill that the magnitude of sediment turnover
seabed drifter study conducted at the time of the spill was fully appreciated by most investigators. Drs. Eva
showed that bottom transport was similar to that Hoffman and James Quinn were the only investigators
observed by previous investigators. The last paper des- that mentioned it could have a significant impact on the
cribes attempts to measure oil dispersed in the water results obtained from sediment samples.
column. Much of our state of knowledge concerning the
A workshop on physical studies was held following physical aspects of an oil spill is contained in the papers
the presentation of the papers to assess our present that follow and the reader is urged to study them. At the
state of knowledge and understanding of the various time of this writing, over a year has passed since theArgo
physical processes occurring during an oil spill and the Merchant went aground. We are not sure where all of the
present capabilities of forecasting spill movement and oi I went except that it was dispersed in the North Atlantic
behavior. along with the oil from the Grand Zenith and a host of
Spill forecasts are used to (1) aide the OSC in clean- other lesser spills. The large number of tar balls found
up operations, (2) to help in damage assessment and along the shoreline of southern New England in the spring
(3) to establish risks in environmental impact statements. of 1977 might be an indication of the fate of some of this
Risk type forecasts based on climatological data, etc., are oil; however, chemical analysis never proved that they
primarily used in environmental impact statements came from eitherof these two ships. Hopefully by nowthe
whereas real time forecasting is relied on by the OSC. oil has been consumed in the ocean by processes of
There was a great deal of diversity of opinion concerning which we know very little about.
present capabilities to forecast the motion of a spill; this Mason P. Wilson, Chairman
might in part be due to the diversityof personnel attending Physical Studies Session
the workshop. It was generally agreed that the gross
motion of a spill can be predicted with some degree of
certainty provided that wind direction and magnitude are
accurately forecasted as well as tidal currents. Some of
the limitations of our present forecasting capabilities are
due to:
(i) the lack of good tidal current data for much of the
U.S. coastal waters
(ii) lack of information on pancake formation and the
inability to forecast slick thickness
(iii) the inability to predict with any degree of cer-
tainty the total area covered by oil
(iv) the inability to accurately predict the motion of
the slick around obstructions and islands
(v) the inability to predict the amount of entrainment
of oil in the water column and sediment.
The recommendation of the workshop was that
much more research is needed to understand oil behavior
on water and its fate following a spill. Much more docu-
mentation must be done on actual spills. It was suggested

The Role of Physical Studies Before,
During, and After Oil Spills

Jerome H. Milgram

Department of Ocean Engineering
Massachusetts Institute of Technology
Cambridge, Massachusetts

Abstract tion of the overall effects of an oil spill by combining them
When oil is spilled on water, an enormous number of with the results of chemical and biological studies. How-
processes take place. Those that fall into the category of ever, there is another role for the scientific community
11 which is often overlooked. The laws of the United States
physical processes" include: transport by wind,. waves, require that the scientific community provide the on-
and currents; spreading on the surface; dispersion into scene coordinator in charge of cleaning up an oil spill with
the sea, principally by breaking waves, evaporation and the scientific information he needs to guide him in
dissolution. The distinction between physical processes choosing optimum cleanup technology and logistics. It is
and chemical processes cannot be made entirely clear important to direct our studies in a way that will achieve
since the effects are often intimately interwoven. Except that requirement.
for mass transport, all of the processes listed above have
connections with physical or surface chemistry, although
none involve chemical reactions involving molecular
changes which are clearly within the domain of chemical Introduction
processes. When the problem of cleaning up spilled oil is
considered, additional physical processes become When oil is spilled on water, a great many processes
important. Because spilled oil often forms a thin layer on take place. The categorization of these into physical,
the surface, effective cleanup requires gathering of the oil chemical, and biological processes is not entirely dis-
with a barrier having a large sweep width. Since the tinct. For example, the spreading of oil is usually thought
device must have forward motion, we are led to a con- of as a physical process,.but surface chemistry has been
sideration of the problem of containment of oi I by a barrier found to play a major role in the spreading process.
in a current. The dynamical instability of the oil-water Similarly, many of the biochemical reactions that take
interface is an important consideration and a limitation on place in living organisms can be considered under both
sweeping speed. The oil droplets resulting from these chemical and biological processes. This paper centers
instabilities can pass beneath the barrier and result in on physical processes, but includes some consideration
containment and collection failure. Thus the mechanics of those aspects of surface chemistrywhich could also be
of these droplets, including recoalescence with the slick, considered as chemical processes.
are of major importance. Substantial study of the various The naturally occurring processes that will be con-
physical processes can be made theoretically, in labora- sidered here which occur when oil is spilled on water are:
tories, and with relatively small controlled oil spills. These (1) Spreading
are the studies which we refer to as "studies before oil (2) Mass transport due to:
spills." However, many of the important physical (a) wind stress
phenomena can only be accurately observed with large (b) water currents due to tides and winds
quantities of oil in deep water so that the only acceptable (c) waves
way of making the studies is to carry them out during (3) Dispersion of oil into the water
accidental spills and their subsequent associated clean- (4) Sedimentation
up efforts. These studies generally point to a lack of suf- (5) Evaporation
ficient knowledge in certain areas which can then be (6) Dissolution
remedied by studies after a spill or during a subsequent When we try to clean up the spilled oil, other
spill. The physical processes which occur after oil is important physical processes come into play. These are:
spilled on water are scientifically interesting in their own (1) Seakeeping characteristics of cleanup
right and have results which are needed for a determina- equipment,

(2) Mass transport due to waves reflected by
collection and containment equipment,
(3) Drainage failure of oil booms,
(4) Interfacial oil-water hydrodynamics associated
with a layer of oil restrained by a boom above a moving
current; a situation which can lead to entrainment of oil
droplets into the water which then pass beneath the
boom.
It is appropriate to consider why we care about 4
these physical and "weakly chemical" processes. Three
reasons for our caring come immediately to mind. First of
all, we do not understand some of the physical and
chemical effects verywell, and as scientists wewould like Figure 1. A Photograph of Oil Spreading on Water. The oil is
to understand them better. Secondly, the physical and opaque so that thick and thin regions can be distinguished by
chemical effects influence the fate and environmental dark and light areas.
impact of the spilled oil. Thirdly, the effects influence the Source: Exxon USA, Third Quarter, 1977, Vol. XV1, No. 3.
optimum cleanup logistics for any particularoil spill. Each
spill with its own oil characteristics and its own external
environment will have different optimum cleanup theory to a spreading pool of oil in the sea, even underthe
logistics. idealization of a pool with monotonically decreasing
The latter two reasons for our caring are recognized thickness from center to* edge. For example, Milgram
by the laws of the United States. The National Oil and (1975) has shown that the gravity-viscous phase of the
Hazardous Substances Pollution Contingency Plan cites Fay-Hoult spreading must be inapplicable when the
the particular relevance of the organization of a standby boundary layer flow in the water becomes turbulent. The
scientific response capability. From a consideration of Reynolds numbers of the flow for real oil slicks at sea
the matters considered by that plan, the scientific indicate that this flow must indeed beturbulent. Secondly,
response capability can be seen to have two major roles: he has shown that for many conditions to which the Fay-
(1) Advising the on-scene coordinator of the likely Hoult gravity-viscous theory has been applied (Offshore
course of future events so that optimum response Oil Task Group, 1973), the surface tension forces which
measures can be chosen. are neglected actually exceed the gravitational spread-
(2) Aiding the Environmental Protection Agency in ing forces. DiPietro and Cox (1975) and Foda and Cox
assessing the damage caused by an oil spill which is a (1977) have shown that even for a single substance, the
,matter required of the EPA. assumption of single values for oil-water and oil-air
Here, it is important to point out that assessing the tensions is inappropriate. They have shown that at the
damage of an oil spill does not diminish that damage in spreading edge of an oil slick a very thin layer of oil
any way; whereas optimizing response plans does spreads ahead of the bulk region of the slick and have
diminish the ultimate damage. As a result, the role of the concluded that in this thin layer the spreading pressure,
scientific community in aiding in the achievement of S, varies with position. The spreading pressure is given
optimum response measures is the more important role, by:
although it is a role which has not been adequately
exercised in the past. S@ Twa-Toa -Tow

where Twa water-air surface tension
Naturally Occurring Physical Processes Toa oil-air surface tension
Tow oil-water interfacial tension
Oil Spreading. The spreading of oil on the sea is a subject The force balance for the region of varying spreading
which has been under essentially continuous study for pressure is achieved by the shear stress of the water on
the past eight years. During the first part of this period, the oil being equal to the gradient of the spreading
analytical and laboratory studies dealt with the spread of pressure.
a liquid having single values of density, viscosity, tension All of the considerations of spreading described
against air, and tension against water. The most well above are for a pool of oil whose thickness decreases
known of these theories is the Fay-Hoult theory (Fay, monotonically from the center to the edge. However, oil
1969, and Hoult, 1972). This theory considered a pool of does not generally spread on the sea in this way. An
oil of constant volume whose thickness decreased mono- example is shown in Figure 1. As exemplified by the
tonically with distance from the center to the edge. Initial- figure, oil often spreads into relatively thick regions (thick-
ly, when the oil was very thick, the theory was based upon ness more than 0.5 mm) surrounded by relatively thin
a balance between the gravitational spreading forces and regions (thickness less than 0.1 mm). The reason for this
the acceleration of the oil and surrounding water. Later, phenomenon has been discussed and disputed for years.
with intermediate oil thicknesses, the theory was based A postulated, but until recently unproven, reason for the
on a balance between the gravitational spreading forces phenomenon was that the stick fractionated into diff erent
and the viscous retarding forces on the oil slick associ- chemical compositions within the thick regions from
ated with the boundary layer flow in the water beneath those in the thin regions with differences in spreading
the oil. At the times when the slick was quite thin, the pressure resulting from the compositional differences.
theory was based on a balance between surface tension Such a difference in spreading pressure could be
spreading forces and viscous retarding forces. A number balanced by a difference in the gravitational spreading
of objections have been cited to the application of this force associated with the observed thickness variation.

An important experiment indicating the possibility of such
a fractionation by spreading was reported by Phillips and
Groseva (1975). In their experiment, octane, decane, and
toluene were mixed to form a "mixed oil". A drop of this
mixed oil was placed at the center of an initially clean
water surface. The drop spread and samples of the sur-
face were taken at different times and at different dis- -12
tances from the center of the spreading layer. The molar
fraction of each type of oil in each samplewas determined
by gas chromatographic analysis. Fractionation did occur
with relatively high molar fractions of toluene and rela-
2
5
tively low fractions of decane and octane found far from
the center, while higher fractions of decane and octane
with lower fractions of toluene were found closer to the
center. For the octane-decane-toluene system, toluene is
more strongly surface active by comparison with the other
constituents. An unrestrained drop of toluene will spread
to a layer having a thickness on the order of the mole-
cular dimensions (approximately 20 Angstroms). Dis-
appointingly little information is available about the
surface active agents in crude and refined oils. water le@el
Recently, the author and Mr. Riyaz Fazal have begun
laboratory experiments oriented towards revealing the
processes that occur when oil spreads into the charac-
5
teristic thick and thin regions. In one type of experiment
that has been done, a drop containing 1 ju t of oil was
allowed to spread in a non-restrained fashion on a sur-
face. The overall extent of spread was determined by first . (c)
sprinkling talc on the surface and then noting the size of Figure 2. A Film Balance. 1, torsion wire control; 2, sweep control;
the region from which the talc was pushed away by the 3, sweep holder; 4, trough; 5, sweep; 6, floal; 7, mirror; 8, calibra-
spreading oil. By dividing this area into the initial volume tion arm; 9, head; 10, torsion wire; 11, gold foil barriers; 12, sight-
of 1 p 1, the average oil thickness was obtained. Experi- ing wire; 13, elevation control; 14, guide; 15, traverse.
ments were done with four different types of crude oil and When an oil film floats on the water to the left of the float and the
the spreading was found to stop after a certain area was water to the rightof the floatis clean, thespreading force of the film
reached. The average thickness of the oil for each case is exerted on the float. This force is measured by the torsion in the
was: with an Arabian oil called Arabian Light, the thick- torsion wire needed to hold the float fixed.
ness was 1110 Angstroms; with an Algerian crude oil Source: Adamson, A. W, Physical Chemistry of Surfaces.
called Arzew, the thickness was 2500 Angstroms; with a
California oil called THUMS, the thickness was 160
Angstroms; and with a Lybian crude oil called Zuetina, the have been able to show conclusively that the fractionation
thickness was 833 Angstroms. Although we do not know effect observed by Phillips and Groseva in their model
the form of the surfactant molecules in these oils, it is system also occurs in crude oil slicks.
certainly unlikely that their molecular dimensions are as Much more remains to be learned about the nature
large as the observed thicknesses, except possibly for of the oil spreading processes. Important questions
the THUMS. This immediately indicates that the nature of about the spreading of the surface active compounds on
the spreading of crude oil is particularly complicated. the water remain unanswered. In particular, we do not
When larger amounts of oil were put on the water know precisely the conditions under which the boundary
surface, the characteristic thick and thin regions were layer in the water is laminar and those at which it becomes
observed. By doing the spreading on water in a film turbulent. Neither do we know the nature of the boundary
balance (see Figure 2), the spreading pressures could be layer flow in the turbulent region which we must under-
measured. By varying the location at which oil was put on stand in order to be able to predict the shear stress and
the surface, it was possible to measure separately the thereby the ultimate spreading rates. Many of these
spreading pressure of the thin region against the matters can be understood by theoretical and laboratory
measurement bar and the thick region against the studies. However, they must be complimented by field
measurement bar. Differences in spreading pressure measurements for at least two possible reasons. First of
were observed. For example, in.the case of the Zuetina all, if the theoretical and laboratory studies lead to an
oil, the spreading pressure of the thick region was found understanding sufficient for generating a method of pre-
to be 1.5 dynes/cm and that of the thin region to be 2.5 dicting how crude oil spreads on the sea, this prediction
dynes/cm. This indicates that the thin regions can have must be compared with actual measurements in order to
higher spreading pressures than the thick regions and assess its accuracy and determine an appropriate con-
thereby inhibit the spreading of the thicker regions of oil. fidence level -for the predictive capability. Secondly, the
The observed difference in spreading pressure of 1 dyne/ analytical and laboratory studies may lead to a semi-
cm can account for a thickness in the thick regions of empirical theory which requires experimental measure-
1.5mm, this being of the same order of magnitude as is ments for determination of certain needed constants. In
frequently observed at sea. The spreading pressure this case, measurements would be required for some
variation must result from fractionation of the slick and we oil spills to determine the constants and for entirely

In the case of the Argo Merchant oil spill, the oil went
away from shore, a condition for which the combination of
the present limits of available cleanup equipment and
present knowledge of oil spill effects would indicate that
no cleanup actions should be undertaken. If, however, the
A oil were coming towards shore, cleanup of certain por-
tions of the slick and protection of certain vulnerable
coastal areas would have been indicated. Under such a
circumstance, it would have been a role of the scientific
community to inform the on-scene coordinator of the
regions most likely to be impacted by the oil. An error in
predicted slick size of a factor of 3 would impose severe
restrictions on the applicability of the usefulness of the
information that the scientific community could provide.
Obviously, the accuracy of trajectory models can be
no better than the basic oil spreading and oil transport
information used by the model. It has already been
indicated that our ability to predict how oil spreads in a
....... quantitative way is severely limited. However, it has also
t been pointed out that this is not the reason for the over-
prediction of the slick area since if spreading were added
Figure 3. Comparison of Observed and Predicted Limits of Argo to the model the area would be overpredicted even more
Merchant Oil Slick, Dec. 21, 1977. strongly. The mass transport phenomena considered by
Source: Grose, A L. and Mattson, J. S. (ads.), The Argo Merchant spill trajectory models are tidal currents, currents due to
Oil Spill, A Preliminary Scientific Report. wind effects on the water, shear stress on the oil imposed
by the wind, and the mass transport of the oil caused by
separate oil spills to check the validity of the resulting the water waves.
predictive capability. Considerable effort has taken place on measuring
In the case of the Argo Merchant spill, the author and then predicting tidal currents in certain areas. For
obtained samples of oil from the thick and thin regions. At each of these areas, there is an appropriate study which
that time, and even now, we do not have a predictive should be undertaken, prior to each tidal measurement
capability with which to compare results, nordowehavea program, which is not done in enough cases. This is the
semi-empirical theory that can be completed by the study of how the number and location of the tidal
addition of field measurements. Nevertheless, it would measurement stations affects the expected accuracy of
have been useful to determine if a difference in chemical oil spill trajectory predictions based on these measure-
composition existed between the oil in the thick and thin ments. Such a study would result in optimization of the
regions. Such a determination was initially planned. tidal data that could be obtained for a fixed amount of
However, the funds to do this work which the EPA had funding. With a limited amount of tidal information avail-
initially indicated would be available did not in fact be- able, predicting the trajectory of an oil spill can be done in
come available. Thus, we had to cancel plans to do the an improved way by utilizing tidal measurements made
work because we were unable to pay for it. If we are to during and immediately after the spill. Clearly, such infor-
gain the information that actual case studies can mation can only be utilized in an effective way if adequate
provide us, it is necessary to have funded plans made Plans for its acquisition and use are made in advance.
in advance. The current induced bythe direct action of wind over
water and the shear stress exerted on oil by wind are
The Mass Transport of Oil. The prediction of the mass subjects about which little experimental information is
transport of spilled oil, coupled with oil spreading theory, available. The reason for this is that experiments with
is frequently considered under the topic of oil spill wind generate waves which have mass transport of their
trajectory modeling. An extensive review and biblio- own that transports the oil. Thus, the motion of the oil in
graphy of this subject is given by Stolzenbach, Mattson, such experiments is affected not only by the water cur-
Adams, Pollack, and Cooper (1977). Hence, it is not rents directly due to the wind and the shear stress of the
appropriate to delve into the most well known details wind on the oil, but also by the waves. On the other hand,
here, but rather some of the problem areas and someof it is possible to generate waves in a laboratory without
the recently collected information should be pointed wind so wave effects on oil transport can, in principle, be
out. Figure 3 shows the observed limits of the Argo measured. However, since laboratory apparatuses have
Merchant oil slick on December 21, 1976 as well as the walls which are not important for waves at sea, and since
limit of the slick predicted by a U.S. Coast Guard R&D the size scale in laboratories is necessarily much smaller
Center trajectory model (this figure came from Groseand than the size scale at sea, laboratory experiments cannot
Mattson, 1977). The figure shows that the direction taken be precisely representative of the situation at sea. As a
by the oil was generally as predicted (downwind) but the result, most work on the mass transport due to water
area occupied by the slick was only one third of the waves has been done theoretically with specially de-
predicted area (as determined by planimeter measure- signed laboratory experiments used for confirming the
ments of the figure) in spite of the fact that the model results. Most of the past work, both theoretical and ex-
apparently did not consider spreading of the oil. If the perimental, was done for depths shallow enough for
model considered spreading as well as mass transport, it effects of the bottom to be important.
would have predicted an even larger area. Until very recently, there was not even a theory suit-

_"1_14_1 I I I I- whether or not it is the third term of Equation (2) that
accounts for the incremental oil speed over that of the
1A I JAL
I sec I increasing time water measured for the thick layers in the Argo Merchant
spill. For this particular instance the relative effects of the
_J -P I 'k- I j", "N A I
"J. 7- mass transport term and the shear stress due to the wind
-74- are not known well enough to make fully definitive
wave breaks conclusions.
All of the existing theories for mass transport con-
Figure 4. Horizontal Velocity vs Time Measured at a Fixed Point sider laminar flows without mixing. Certainly breaking
2.9 cm Beneath the Mean Free Surface During Passage of a waves are a source of m'1xing in the ocean. Figure 4 shows
Breaking Wave. The sinuous component of the velocity is due to a measurement of the horizontal velocity beneath a
the underlying wave motion. The jagged component of the velo- breaking wave in the Marine Hydrodynamics Laboratory
city is due to the turbulence generated by the breaking process. at MIT. This figure shows turbulent velocities of about
Turbulent motion results in mixing. The sinuous motion has no
mixing. These measurements were made in a laboratory. one-third the fluid orbital velocities in the waves. How-
ever, the turbulent velocities die out very quickly indi-
cating that a fairly rapid succession of breakers would be
needed to maintain a significant turbulence level. There is
able for the understanding of the mass transport of water insufficient information available about the near surface
without oil for waves in deep water. The theories which turbulence level in the ocean to be able to estimate the
had been developed for water of finite depth led to results effects of natural mixing on the mass transport of oil. Such
which were obviously incorrect when the water was deep. information is much needed. It can be obtained at sea
Two of these theories which are correct for shallow water without oil, but similar studies should be made in the
are those of Longuet-Higgins (1953) and Unluata and Mei presence of oil since it is known that oil inhibits wave
(1970). These theories are applicable to laminar (non- breaking and therefore its presence can diminish the
mixing) flows in the water. Recently, the author has con- amount of natural mixing. Extrapolating the existing
sidered the problem of mass transport both with and theory without mixing indicates that natural mixing should
without oil in deep water for laminar (no mixing) flows increase the mass transport of oil due to waves. Thus, the
(Milgram, 1978, 1). The result for the mass transport speed existing theory provides a lower bound on an estimate of
at the top of the oil is given by: the wave induced mass transport velocity.
03 2o2\/,)- 2 All commonly used oil spill trajectory models
.t 2o
V=-A2 (1 +_+ + small term) (2) assume a net drift of the oil as some fraction of the wind
9 g r (3/2) 9 speed commonly taken as between 2 and 4%. This is to
where account for the effects of direct wind shear as well as the
V is the surface fluid flow speed, mass transport due to waves. It is important to distinguish
a is the radian frequency of the waves, between the two effects since different wind conditions
A is the wave amplitude (half the height), can lead to different waves. For example, in a developing
g is the acceleration due to gravity, storm with increasing wind speeds the waves will be
v is the kinematic viscosity of the water, shorter and steeper than waves for the same wind speed
t is the time since the waves first reached the fluid in a diminishingwind afterthe peakof astorm has passed.
being observed, and Accurate predictions of oil spill trajectories will have to
r' is the gamma function. take into account the direct wind shear effects and the
The first term in Equation (2) is the predicted surface wave induced mass transport separately. Much informa-
transport speed given by Stokes (1947) for an inviscid tion about determining these effects separately can be
fluid. Naturally, such a term exists in the water as well. The found from laboratory experiments coupled with fluid
second term is a slowly increasing term found by the dynamic theory. Since the laboratory scale Reynolds
author to be caused by the slowdiffusion of vorticity number will necessarily be much smaller than the full
generated at the free surface into the interior of the flow in scale Reynolds number at sea, direct scaling is im-
the water. It exists in the absence of oil as well as in its possible. However, it should be possible to develop a
presence. The third term is the result of the rapid diffusion semi-empirical theory for the effects with the needed
of surface generated vorticity through the relatively constants determined from careful measurements at sea
viscous oil. The small term results from non-linear effects both with and without oil.
in the free surface boundary layer. It involves oil and water
densities and viscosities and the oil layer thickness. Its Dispersion of Oil Into the Sea. Observations of oil spills at
detailed form is given in the reference. The author sea show that when breaking waves interact with floating
(Milgram, 1978, 11) has performed a series of experiments oil, some of the oil is broken up into small droplets which
which confirmed the existence of the third term. are driven down deeply into the water. This phenomenon
Experiments made by the National Oceanic and has importance both with regard to control and cleanup
Atmospheric Administration, and reported by Grose and measures and with regard to environmental effects of the
il. Since control and cleanup equipment operates at or
Mattson (1977), of oil transport velocities in the Argo of
Merchant oil spill showed oil velocities as much as one near the surface, oil which is deeply dispersed cannot
percent of the wind speed faster than the velocities of generally be contained or collected. Thus, optimum
surrounding water. Using typical values for wave heights, cleanup measures must be based on the anticipated dis-
wave lengths, oil thicknesses, and wind speeds shows that tribution of dispersion. If breaking waves exist in one area
the third term in Equation (2) is of the same order of but not in another, containment and collection efforts
magnitude as 1 % of the wind speed. The experiments that would generally be best concentrated in the area not
were done were not detailed enough to determine having breaking waves. On the other hand, if dispersants

Figure 5. Photographs of a Breaking Wave Dispersing Oil in a
Laboratory Wave Channel.

(a) The wave breaking has begun at the upper right. Thewavesgo
from right to left. The dark gray at the left is the relatively undis-
turbed oil. The lighter gray beneath it is the far channel window
with a layer of oil on it.

(b) The flow in the breaking wave has carried some oil quite deep.
Note that on the free surface at the right all the oil is gone and an
ai.r-water surface is seen.

(c) The turbulent oil has broken into a cloud of droplets. This is
later after the beginning of breaking than the cases shown in the
two preceding photographs.

(d) This is after the breaking wa ve has passed. Many fine droplets
remain in the water column. The breaker has opened a "clean"
spot on the water surface at the right. The oil that was there has
been dispersed into the cloud further to the left.

4 are to be used, the breaking wave effects can aid in the
dispersion and the mixing of the dispersants so areas
containing breaking waves might sometimes be the most
appropriate ones for application of dispersants.
When there is dispersed oil in the water, there is
more surface area of oil available to biological organisms.
Whether this is good or bad depends on whether the
resulting biodegradation has benefits which outweigh the
harm of the poisoning of organisms by the oil or the intr6-
duction of the oil into the food chain. In addition, when
there are oil droplets in the water, more oil surface is
available to suspended sediments which can result in
sinking of some of the oil. It is interesting to note that
Forrester (1971) found droplets of oil as deep as 80
meters below the surface in Chedapucto Bay weeks after
the grounding of the tanker Arrow.
Only recently have studies been initiated to deter-
mine the details of the mechanism by which breaking
waves disperse oil. Figure 5 shows photographs of
laboratory experiments of the dispersion by breaking
waves phenomenon. In order to be able to make quantita-
tive predictions of dispersion effects for oil spills at sea,
much more work is needed. We have found that the sea
conditions as well as the oil properties affect the degree
of dispersion. Qualitative observations in the field have
All
Ot- shown this also. For example, differences in the nature of

the dispersion of the oil from the Argo Merchant on Table 1. Extrapolated Fresh Water Solubilities of
Nantucket Shoals and the oil from the barge Ethyl H in the n-Paraffins in Heavy Fractions of Crude Oil
Hudson River were observed. Both spills were of No. 6
fuel oil. Observation of the oil from the Argo Merchant
indicated that when it was struck by a breaking wave, it Extrapolated Solubility in
dispersed into the water. but soon was seen to rise to the Fraction Carbon Nos, Fresh Water
surface again. In the case of the Ethyl H, however, fol- gm-3
lowing the breaking wave much less rising and recoales- Kerosene C10-C17 1 x 104 - 2 x 10-1
cence of the droplets into the slick were observed. No
quantitative studies at sea were made during the disper- Gas Oil C16 C25 6 x 10-4 - 10-8
sion process in either case. Such studies would have
been quite helpful. *Similarly, following the oil spills Lube Oil C23 C37 10-7-10-14
laboratory studies of the differences of the oil aimed ai
determining why the dispersion effects were different Bitumen, etc. >C37 < 10-14
would have been quite helpful. Unfortunately, no funds
were made available for such studies and they were not Source: Parker, Freegarde and Hatchard (1971)
carried out.
Sinking and Sedimentation. Very few oils or oil products indicates that dissolution is not significant for paraffins
are heavy enough to sink in sea water. In fact, the only larger than C23. For the smaller molecules, the important
constituents in oil in substantial quantity that by them- factor controlling dissolution is the dissolution rate. The
selves are heavier than sea water are some of the fused author has found no literature indicating these dissolution
aromatic ring compounds and many of these, in fact, float. rates for oceanic conditions. Laboratory experiments
Thus, unassisted sinking is a rarity. with sea water indicate that in most instances the intro-
If heavier-than-water sediments contact the oil, duction of oil into the water column by dissolution is
some oil can be absorbed or adsorbed by the sediments negligible in comparison with its introduction by droplet
with the oil-sediment combination going to the bottom. dispersion and by adhering to biological material.
This can occur either by oil contacting the bottom by .Evaporation can be a major effect in removing oi I from
escaping from a sunken vessel or by dispersed oil drop- the sea. A spill of a relatively volatile oil (including volatile
lets being driven to the bottom by breaking wave tur- crudes) in warm conditions can easily result in half of the
bulence; or by suspended sediments contacting the oil entering the atmosphere in a day or two. Prediction of
surface slick or dispersed droplets in the water column. evaporation loss can aid in planning optimum response
There is inadequate information available about the logistics. Some years ago, some important contributions
conditions and oil type-sediment type combinations that towards an understanding of the oil evaporation problem
can lead to significant sedimentation. Thus, predictions appeared in the literature. Blokker (1964) hypothesized a
about sedimentation can only be made in the simplest mathematical model for the evaporation of oil in the
situations. presence of wind. Blokker did laboratory tests and found
Immediately after the Argo Merchant began spilling his model in poor agreement with measurements for short
oil, several people pronounced that the damage would be times and in good agreement for long times. Sivadier and
severe because the oil would sink and foul bottom life. Mikolaj (1973) measured oil loss rates at sea for a small
These pronouncements were made without any scientific spill of 15 liters of oil and obtained reasonable agreement
basis. with Blokker's model. However, they noted the afore-
The author and Dr. Edward Kern took samples of the mentioned thick and thin oil regions in their tests and their
oil and the water and found that there was no significant measurement methods for determining evaporation rates
quantity of suspended sediment, that oil dispersed by from slick samples in the non-uniform slick are far from
breaking waves appeared to rise rapidly, and that the convincing as regards accuracy. No pre-planned quanti-
specific gravity was 0.96 with this figure not changing tative measurements of evaporation rates during large oil
significantly after the cutter stock was evaporated off. spills seem to have been made. Such measurements
These facts indicated the oil would not sink. could be quite helpful in establishing the accuracy of
In many spill situations, the evidence will not be so Blokker's model and in pointing the way to more accurate
clear. It would be helpful for the on-scene coordinator to predictive methods if they are needed.
know where oil would sink and when this would occur so
he could plan optimum cleanup logistics. The planning of
surveys to establish damage assessment would benefit Physical Processes That Occur During Oil Spill Cleanup
from an accurate prediction of sinking distribution. We
cannot make accurate predictions of sedimentation in Seakeeping Characteristics of Cleanup Equipment. When
those cases where the correct answer is more obscure oil containment and cleanup equipment is used at sea,
than it was in the case of the Argo Merchant. Our predic- the characteristics of the motion of the equipment affect
tive ability could be improved by studying the problem in the containment and cleanup efficiency. For example, an
laboratories and in the field during and after oil spills. oil barrier (boom) is of little utility if the oil it is to contain
passes over or under the barrier. This will occur if the
Evaporation and Dissolution. The evaporation and dis- barrier does not follow the vertical motions of the waves
solution of spilled oil is usually considered in the field of with adequate accuracy. In addition, if a floating barrier
chemical studies. However@ since these effects are often does not maintain an essentially vertical fence, but rolls
overlooked, a few words will be devoted to them here. back and forth in the waves to an excessive degree, this
Table 1 gives the solubilities of the n-paraffins pre- rolling motion will allow the oil to pass over or under the
dicted by Parker, Freegarde, and Hatchard (1971). This barrier. If the barrier does not follow the sway motions of

the fluid particles in waves, temporarily high currents dence factor to place on the work that is to be done.
between the boom and the water will exist and such cur-
rents can draw oil beneath the boom. Generally, the sea- Modes of Hydrodynamic Failure of Oil Barriers. Oil barriers
keeping constraints on skimmers are even higher than on form a fundamental line of defense against damage from
barriers. The reason for this is thatthe skimming elements oil spills. They can be used in several ways. One is to
of these devices must be kept in the oil most of the time for protect vulnerable areas. Another is in skimming systems.
high efficiency. If the skimming elements spend much of Once oil has spread on the sea, it is so thin that the only
the time in the air or in the water, poor oil collection device which can encounter the oil at a high rate is an oil
efficiency will result. Studies of the seakeeping motion of boom. Therefore, massive spill cleanup at sea can only
pollution control devices and the design of equipment take place with towed oil booms having skimmers built
which minimize adverse motions are matters which can into the booms or used within the U-shaped configuration
be handled before oil spills. However, in spite of ourability formed by the towed boom. In the case of the Argo Mer-
to analyze the seakeeping characteristics of a wide chant oil spill, the oil was No. 6 fuel oil and the tempera-
variety of equipment types, very little work in this regard ture was very low. This combination resulted in the oil
has taken place. The only extensive work on the sea- being extremely viscous (the author measured kinematic
keeping abilities of pollution control barriers has been viscosities on the order of 70,000 centistokes). Hardly
thatof Milgrarn (1971)and Milgramand O'Dea (11974). This any pumping systems could handle so viscous an oil so it
work provided a framework for studying the barrier sea- would not have been possible to actually collect theArgo
keeping problem and contained some preliminary experi- Merchant oil from the surface of the sea unless we had
ments. A few barriers have been designed and built in special skimming devices for oil of this type. Since we do
accordance with some of the findings of this work. How- not have such devices, what could we have done had the
ever, most of the pollution control barriers in existence oil threatened shore? It would have been possible to trap
today have far less adequate seakeeping motions. The much of the oil in towed barriers and then to tow these
reason for this seems to be due to the fact that less barriers containing much of the oil far out to sea in hopes
adequate barriers can be built in a way that is less expen- that the oil would be degraded or the weather direction
sive and that leads to barriers that are easier to handle would change before that oil returned to the shoreline.
than those having good seakeeping motions. However, Thus we see another potentially very important use of
their oil containment and collection efficiency is substan- barriers.
tially lower than that of the better devices. At this time, it Early oi I barriers generally suffered from inadequate
would be worthwhile to do more theoretical'and experi- strength and inadequate seakeeping ability. Now, some
mental research on barrier seakeeping in orderto quantify barriers are availablewhich are strong enough to maintain
the degree to which containment and collection efficiency structural integrity and which have sufficiently good sea-
is lost as the features which lead to good seakeeping keeping ability to follow the wave motion in moderate
motions, but which are expensive, are reduced. It is im- waves (typically in waves up to about 2 meters high, and
portant to note that nearly all of the required experiments higher waves when they are of limited steepness).
can be done without the presence of oil and hence the Remaining problems with the use of barriers include two
work can be carried out and the fruits of this work used to modes of hydrodynamic containment failure wherein oil
produce useful equipment without there having been any goes beneath the barrier; one being called drainage
oil in the water. Thus, this is a task to do before an oil spill, failure and the other being called entrainment failure.
not during or after. Of course, evaluation of performance Drainage failure is easier to understand and easier to deal
during an oil spill is useful for comparing. predicted per- with. Basically, when a barrier is towed faster and faster
formance with actual performance in order to determine the thickness of the oil pool at the base of the U-shaped
what, if any, equipment modifications should be made. configuration gets larger and larger. Eventually, the oil
becomes so deep that it passes beneath the barrier. This
Mass Transport by Reflected Waves. It was mentioned occurs before the depth of the pool equals the barrier
above that the seakeeping requirements on skimmers are draft because the pressure distribution associated with
often stricter than those on barriers. In addition, another towing a barrier results in a "suction effect". It is always
seakeeping effect is important in the case of skimmers. possible, in principle, to avoid drainage failure at any
When an external device interacts with waves, it reflects particular tow speed by using a sufficiently deep barrier.
some of these waves. In a system with oil floating on Entrainment failure, on the other hand, cannot be
water, there are two different types of waves of impor- prevented at arbitrary towing speeds because this mode
tance. The normal waves at sea are of one specific type. of failure has nothing to do with the details of the barrier
The second type is one which moves oil rapidly in the itself. It has to do with the fact that the interface between
direction that the wave propagates. However, the re- the floating oil pool and the water becomes unstable
flected waves contain both types. Asaresult, a devicewhich when the water speed relative to the oil is sufficiently
reflects many waves when floating in an oil spill tends to large. Figure 6 is a sequence of sideview photographs of a
drive the oil away from itself. In the case of a skimmer restrained oil layer in an experimental flume at MIT for
under some sea conditions, this effect can lead to a various current speeds ranging from 15 to 46 cm/sec. The
device not being able to collect any oil at all! Thus, the figure shows that the oil-water interface is very smooth in
interaction of the device with the oil through the pheno- a current of 15 cm/sec. At 23 cm/sec, a few waves on the
menon of reflected waves is especially important. This interface are formed and a slightly bulbous region, called
effect has not been studied in detail for many of the a headwave, is generated just behind the leading edge of
skimming devices presently in existence. It needs to be the slick. These features are considerably more apparent
studied now in order to determine which devices are the at a current speed of 30 cm/sec and they are further
most appropriate ones to use on oil spills and it needs to amplified at a speed of 38 cm/sec. At a speed of 46
be studied during spill cleanup in order to obtain a confi- cm/sec the unsteady motion has become sufficiently

4 irect effect on equipment operations and on overall
d
logistics.

PLOW
Slowly rise. Three possible things can happen to a drop-
When oil droplets are formed by entrainment, they

STRUCTUML SUPPOM' let. First of all, it may initially be driven down so deep that
0) 15 ..V.. TLST SMCTION,or FLUME, it does not rise up to the slick before it has. passed
beneath the boom and is therefore lost by entrainment
failure. Secondly, it may rise to the slick but not recoalesce
before it moves to the boom where it is easily torn away
from the slick and drawn beneath the boom. Thirdly, it may
23,1CW M rise to the slick and recoalesce to the slick so that it is not
4n
Al
lost by entrainment. Clearly, the third possibility occurs
A
7., 7@@g more frequently when there is a long oil slick contained in
"d ACV,
a barrier than when there is a short slick. No detailed
observational or quantitative experiments have ever been
(ci
made with long oil slicks in barriers. Two sets of experi-
ments with quantitative measurements and careful under-
'7
water observations were made with relatively short pools
of contained oil as reported by Tierney (1975). These
eriments were made with between 28,000 and 40,000
(d) 38 7, exp
. . ..... ...
gallons of oil in sections of the U.S. Coast Guard barrier
that Were 1,000 feet long. Such barriers can hold about
500,000 gallons of oil. Thus, with the booms holding less
"A than 1/10th of their ultimate capacity, the slicks were
7 i7
quite short and the space available for droplet recoales-
W 46
cence wa
s sma
P,
Generally, it has been found that oil booms lose oil
Figure 6. Photographs of a Restrained Layer of Dyed Heavy by entrainment failure at towing speeds in excess of one
Mineral Oil at Various Flow Speeds. On each photograph the flow knot (52 cm/sec). An important question is: Can oil
speed and length scale are shown. booms which are deep enough to prevent drainage failure
be towed at substantially higher speeds than one knot
and not suffer excessive entrainment failure if the oil slick
violent for the interfacial waves to just begin to break and is relatively long (say 150 meters or more)? That question
thereby generate droplets of oil which can become en- can only really be answered by quantitative experiments
trained in the flowing water. This happens to oil spills at and careful observation of oil booms containing enor-
sea, and when the droplets of oil in the water pass beneath mous quantities of oil during actual large oil spill contain-
the boom, we have containment failure due to entrainment. ment and cleanup operations. Such measurements and
Although the existence of the entrainment problem observations have never been made and should be care-
has been known for several years, it is only very recently fully planned as part of our oil spill response measures.
that the fundamental fluid mechanics which cause it have
begun to be understood. Milgram and Van Houten (1978) Conclusions
have conclusively confirmed the hypothesis of Leibovich
(1976) that the short interfacial waves are Kelvin- In the above, I have attempted to show that there is
Helmholtz waves and the oil droplets result from the an enormous amount of technical work remaining to be
breaking of these waves. done in studying the physical phenomena related to oil
Although we have learned a great deal about the spills. Some of the phenomena can be studied theore-
conditions under which the Kelvin-Helmholtz waves are tically and in laboratories, some can be studied at sea
generated, there is still much to learn about the condi- without oil, some can be studied at sea with small
tions under which they break and form droplets. For quantities of oil, and some must be studied at sea with
example, in the experiments carried out by Milgram and Very large quantities of oil. All but the last two groups can
Van Houten, it was found that droplet generation began be handled by normal well planned scientific efforts, if
with a diesel oil slick at half the relative current speed as properly funded. Studies at sea with small quantities of oil
it began with a slick of the Algerian crude oil called Arzew are often hard to do because it is sometimes difficult to
in spite of the fact that the two types of oil had almost obtain permission to spill oil. Arrangements for the re-
identical densities, viscosities, interfacial tensions, and quired tests need to be made in a reasonable way-
surface tensions. Since these measurements are basically Studies with very large quantities of oil can only be done
"bulk measurements", it seems likely that the differences during the occurrence of major oil spills. It is not possible
have something to do with the fact that the crude oil is a to plan such studies once a spill has occurred. Extensive
multi-component material. The details of droplet genera- initial planning and preparedness is needed in order to
Ition of multi-component materials need to be studied obtain the maximum amount of information which can be
further. We know very little about the interfacial wave learned from each spill. This requires a substantial finan-
breaking and resulting entrainment from slicks of very cial commitment. However, without it we will never be in
viscous oils like that of the Argo Merchant. This informa- an improved position to answe@ the important questions
tion is of crucial importance to the on-scene coordinator that arise each time oil is spilled nor will we be able to
if he is to arrange for spill containment and cleanup since reduce the cost and environmental damage of oil spills to
the speed at which oil booms should be towed has a the extent that is technologically feasible.

Reterences
Brokker, P,C,., 1964, "Spreading and Evaporation of Petroleum
Products on Water", 4th International Harbor Conference,
Antwerp, pp. 911-914
DiPietro, N.D. and Cox, R G., 1975, McGill University, Dept, of
Civil Engineering Technical Report No. 75-4 (FML),
Fay, J, A., 1969, "The Spread of Oil Slicks on a Calm Sea,,, in Oil
on the Sea, Hoult, D. P.,ed., Plenum Press, New York,
PP. 53-63.
Foda, M. A. and Cox, R. G, 1977, "Similarity Solutions for the
Spreading of Oil Slicks", McGill University, Dept. of Civil
Engineering Technical Report No, 77 (FML.
Forrester, W.D, 1971, "Distribution of Suspended Oil Particles
Following the Grounding of the Tanker Arrow", J. of
Marine Res., 29, 2, pp 151-170,
Grose, P. L. and Mattson, J,S. (eds.), 1977, "TheArgo Merchant
Oil, Spill, A Preliminary Scientific Report", U.S. Dept. of
Commerce, NOAA, Environmental Research Labora-
tories, Boulder, Colorado 80302.
Hoult, D. P., 1972, Ann. Rev. Fl. Mech., Vol. 4, p. 341,
Leibovich, S, 1976, Oil Stick Instability and Entrainment Failure
of Oil Booms, J. Fluids Eng., March 1976, pp, 98-105.
Longuet-Higgins, M. S,, 1963, "Mass Transport in Water Waves",
Phil. Trans. A. Roy. Soc., London, pp. 535-581
Milgram, J. H,, 1971, "Forces and Motions of a Flexible Floating
Barrier" J of Hydronautics, 5, 2, pp. 41-51.
Milgram, J. H, and O'Dea, J. F., 1974, "Evaluation of the Strength
Seakeeping Ability of Pollution Contro) Barriers", U.S.
Coast Guard Office of Research and Development Report
No. CGD-56-75,
Milgram, J. H., 1975, Problems with Application of the Fay-Hoult
Theory, Presentation to New England Section of the
Marine Technology Society,
Milgram, J. H., 1975, "Mass Transport of Water and Floating Oil
by Gravity Waves in Deep Water", submitted for publi-
cation,
Milgram, J, H., 1978, "Measurements of the Transport of Water
and Floating Oil by Gravity Waves in Deep Water", to be
published.
Milgram, J. H. and Van Houten, R, J., 1978, "Mechanics of a
Restrained Layer of Floating Oil Above a Water Current",
to be published in J. of Hydronautics.
Offshore Oil Task Group, 1973, The Georges Bank Petroleum
Study, MIT Sea Grant Report No. MITSG 73-,
Parker, C. A., Freegarde, M., and Hatchard, C. G., 1971. "The
Effect of Some Chemical and Biological Factors on the
Degradation of Crude Oil at Sea", in Water Pollution by Oil,
Hepple, P. (ed.), Institute of Petroleum, London, pp.
237-244.
Phillips, C.R. and Groseva, V, M 1975, "Separation of Multi-
Component Hydrocarbon Mixtures Spreading on a Water
Surface", Separation Science, 10, 2, pp. III- 118,
Sivadier, H. 0. and Mikolai, P. G., 1973, "Measurement of
Evaporation Rates From Oil Slicks on the Open Sea",
Proc. of Joint Conference on Prevention and Control of Oil
Spills, Am. Petroleum Institute, Washington, D.C.,
pp. 475-484.
Stokes, G. G., 1947, "On the Theory of Oscillatory Waves",
Trans. Camb. Phil. Soc,, 8, pp. 441-455.
Stolzenbach, K. D., Mattson, J. S.. Adams, E.A., Pollack, A. M
Cooper, C.K, 1977, "A Review and Evaluation of Basic
Techniques for Predicting the Behavior of Surface Oil
Slicks", MIT R. M. Parsons Laboratory Report No. 222.
Tierney, J. M.. 1975, Research, Design, and Development of the
U.S, Coast Guard High Seas Oil Containment System,
U.S.Coast Guard Office of Research and Development
Report No. CG-D-42-76.
Unluata, U. and Mei, C. C., 1970, "Mass Transport in Water
Waves", Jq. Geophys. Res., 75,36, pp 7611-7618.

Chronology of Events and Oil Slicks From
the Argo Merchant

James S. Mattson

Center for Experiment Design and Data Analysis
Environmental Data Service/NOAA
Washington, D.C.

Abstract Deaver was carrying an infrared radiometer for use in
thermal mapping of the Gulf Stream. The Coast Guard
The operational and scientific events associated agreed with the few scientists present that night at Otis
with the Argo Merchant oil spill are. presented in the AFB to send Deaver out the next morning to map the
context of the condition of the vessel and the location extent of the oil slick emanating from the grounded
of the spilled oil. tanker.
The next morning, Friday, 17 December 1977, Peter
Chronology of Events Grose and Gary Hufford accompanied Deaver in the first
mapping effort. Visiting the grounding scene first [Slide 2;
The purpose of this presentation is to give an over- Photograph 2, p. 111-3, Grose and Mattson, 19771, they
view of the Argo Merchant oil spill during the period of reported a heavy concentration of oil extending 5 miles to
15-31 December 1976. While not going into detail on the the northwest, towards Nantucket, then curving 31/2
field and theoretical studies that accompanied and miles to the west fSlide 3, Figure 1]. They reported
followed this spectacular event, this overview is intended that the oil looked like "an asphalt road," with no
to "set the stage" for the physical, chemical, and biologi- thin sheen surrounding it. This is not unusual,
cal studies to be reported on during the remainder of looking back over the event, as the weather that
this Symposium. day was relatively poor, with 17 knot winds, rain and snow,
When the Argo Merchant went aground on the and the seas were showing 2 to 4 ft. waves and swells.
morning of 15 December 1976, no one was around to All through theArgo Merchant experience, much of the oil
evaluate the extent of the initial spill. The first scientific slick would "disappear" on rough days, only to reappear
observations of the scene were made by Elaine Chan of almost undiminished when the weather would clear. A
NOAA and Gary Hufford of the U.S. Coast Guard R&D power failure aboard Deaver's plane terminated Friday's
Center. They flew to the scene in a chartered single-
engine Cessna 182, on Thursday morning, 16 December
1976. [Slide 1; Photograph 1, p. 111-3, Grose and
Mattson, 19771*. Chan and Hufford reported an oil slick
extending approximately two miles from north to south
and four miles east to west, being fed by a streamer of oil
bearing about 240' from the vessel. That afternoon,
Peter Grose, Craig Hooper, and I flew to the scene of the
grounding in the same plane, only to find that fog had
obscured visibility so much that we could not approach
the vessel. What we did see, to the west of the Argo
Merchant, was a nearly continuous oil slick extending
several miles from north to south.
The biggest break of the Argo Merchant spill came
that night when that same fog forced Joe Deaver, of the
Coast Guard's Oceanographic Unit, to land at Otis AFB
on Cape Cod. Deaver's twin-engined HU-16 was far
superior to the rented Cessna for operations at sea, and DEC. 17, 197r.
*Copies of the color slides may be found in Appendix III of Grose
and Mattson, 1977. Figure 1. 17 December 19 76 Slick Map.

Q1.

--- @Vo

DEC. 18. 1976 DE*C. 20. B76

Figure 2. 18 December 1976 Slick Map. Figure 4. 20 December 1976 Slick Map.

overflights of that day [Slide 7; p. 111-12, Grose and
Mattson, 1977].
By Sunday, the 19th, we had become aware that the
'.asphalt highway" seen from the HU-1 6 overflight was in
reality a mixture of thick "pancakes" surrounded by thin
sheen". The thickness of one such pancake was esti-
mated to be on the order of one centimeter. From the
helicopter, Jerry Galt and I estimated the "pancake"
coverage to be no more than 1 /10 of the total observable
slick, but even this estimate produced a slick "volume"
that exceeded the Argo Merchant's entire cargo by about
a factor of three. Meanwhile, the figure arrived at by
Deaver from his HU-16 observations was something
between 90 and 100% "thick" oi I on the 1 9th. Rather than
try to change the aerial estimation of "pancake" coverage,
Deaver had a "look-see" constructed of Plexiglas with a
DEC. 19.1976 50-square grid etched on it, so that he could give consis-
tent estimates from day-to-day. We hoped that his esti-
mates could eventually be "ground-truthed" with some of
Figure 3. 19 December 19 76 Slick Map. the many aerial photographs that were being taken by
search pattern prematurely, and only the limited rectan- NASA and by BLM's contractor, Aero-Marine Surveys,
gular flight path shown on Figure 1 was carried out. One Inc.
of Deaver's primary objectives during his initial overflights, On Monday, the 20th, the weather was reasonably
one of many that lost out to the broader objective of good, and Deaver, Dave Kennedy, Captain Lynn Hein,
"mapping", was to use his infrared radiometer to measure the on-scene coordinator, and I set out in an H-3 helicopter
oil slick temperatures, particularly to see whether the thin to map the slick, and to conduct some differential oil-
and thick 'regions of oil differed either positively or water velocity measurements. The horseshoe shape of
negatively from the surface water temperature. The the previous day had given way to the banana-shaped
temperature isopleths on the slick maps for slick shown in Figure 4 [Slide 81, and when this information
December 17, 18, 21, 23, and 24 represent that was telephoned to Jerry Galt at WHOI, just before
effort. By the 31 st, and January 2nd and 3rd, the rationale Oceanus departed on her first cruise during the incident,
for thermal mapping had changed to one of locating the it appeared that the *oil had broken out of the tidally-
edge of the Gulf Stream, as well as a suspected "ring" on influenced regime and was headed to the northeast.
the western edge of the Stream. On Oceanus, -this was the interpretation given to the
By Saturday, 18 December, the counter-clockwise observations, and the cruise was planned to the north-
shape of the slick, due to rotary tidal currents, was east of the oil slick, as shown, on 20 December.
distinctly visible [Slide 4, Figure 21, and on Sunday, a On Tuesday, 21 December 1976, all hell broke
beautifully calm day, the slick had grown into a 16-mile loose. The Argo Merchant 'broke in two aft of the king
long "horseshoe", as shown in Figure 3 [Slide 5], and in post at 0835 [Slide 9; Photograph 7, p. 111-4, Grose and
the photograph taken from a Coast Guard helicopter Mattson, 19771, releasing 1.5 million gallons of oil
from the east of the grounded tanker [Slide 6; Photo- all over the place," according to Deaver and
graph 4, p. 111-3, Grose and Mattson, 1977). The shape Dave Kennedy, who were in the HU-16. The slick
of the slick on 19 December is best exemplified by the map prepared by Deaver that day included much
composite photomosaic prepared by NASA from their of the oil released in the first breakup, and is shown

V11 -------

DEC. 23, 876
DEC. 21, 976
Figure 7. 23 December 19 76 Slick Map.

Figure S. 21 December 1976 Slick Map.

D
DEr 2Z mc. 24. FYM
EC 2Z 8%
L
Figure 6. 22 December 19 76 Slick Map. Figure 8. 24 December 19 76 Slick Map.

in Figure 5 [Slide 101. By then, there literally was oil
"all over the place".
On Wednesday the 22nd, while many of our
colleagues, who were to eventually show an interest in
the spill, were attending an EPA-called meeting at the
Federal Building in Boston, the Argo Merchant split again
at 0730, releasing another 1.5 million gallons (the amount
estimated to be contained in three tanks across th
e
vessel) into a heavy sea (winds up to 50 knots, seas to
fifteen feet). While Deaver and Kennedy carried out a six-
hour long mapping flight, with the result shown in Figure 6
[Slide. 111, Elaine Chan and Scott Fortier (USCG R&D
Center) went out to the scene in a helicopter to take
samples of the thick oil that was surrounding the wreck-
age at that time. In a part of the Argo Merchant
story not well-circulated to date, the HU-16 carrying
Deaver and Kennedy lost an engine on this mapping
flight, and the engine caught fire on landing at Otis AFB.
On Thursday, 23 December, a number of valuable DEC. M UM
pieces in the puzzle were collected. First, Dave Kennedy
took a team of four U.S. Navy divers out to the scene in Figure 9. 25 December 19 76 Slick Map.
order to photograph the underside of the floating
"pancakes", looking for streamers, "skegs", or anything
else that might be peculiar to that situation, and to
photograph the sandy bottom in the vicinity of the wreck of underwater slides showing the underside of the slick,
to determine whether any of the oil had sunk in noticeable and their own descriptions of what they had seen. In addi-
quantities in the seven days since the grounding. They tion, Deaver conducted another 6-hour mapping survey,
returned before noon with 15 minutes of movie film, a roll with the results shown in Figure 7 [Slide 13].

I Pa

DEr 9Z am

Figure 11. 27 December 1976 Stick Map.

DEC 26.
764 _(.@-W- 6M
42N 7" _J k 4rN-
Figure 10. 28 December 1976 Slick Map.
On the 24th and 25th, Joe Deaver and Al Kegler (of
the Alaska Dept. of Fish and Game) continued the
mapping, as shown in Figures 8 and 9 [Slides 14, 15],
although by Christmas Day, the mapping took just under -40'N- -WN
seven hours in the HU-1 6, and it had become impossible
to map the easterly extent of the slick in that time,
Christmas Day did produce a bonus, however, as Deaver
and Kegler located a huge (300'x 700', probably over 1/2
million gallons) pancake, shown in Figure 9 [Slide 16; WN
Photograph 38, p. 111-21, Grose and Mattson, 1977).
Because of its size, we guessed that "Pancake I" com-
prised the entire cargo of one of the tanks that split
open on December 21 or 22.
At that point, by Sunday December 26, mapping of 36*N ?rW 70V ",W WN
the extent of the slick, because of its huge size, had to be W 617W
reduced in priority as other objectives became important. Figure 12. Track Lines of Drifting Buoys.
The first of these originated from a weather forecast on
Christmas Day that called for sustained easterly winds
that night. Computer projections being carried out by the December, Deaver made a flight to the edge of the Gulf
Coast Guard based on Deaver's maps predicted a near- Stream in an effort to determine whether the slick had
landfall of the eastern edge of the slick on Nantucket entered the Stream. That evening, in afive-hour helicopter
Island sometime Sunday. Therefore, when Deaver, flight to the eastern portion of the slick, Peter Grose and
Kegler, and Ed Myers (of NOAA) took off in sleet and snow Marilyn Pizzello (NOAA) dropped a satellite-trackable
at 0915 on the 26th, they deployed some 4,000 drift cards buoy into the middle of what they thought was "Pancake
at locations between the last reported easterly edge of
the slick and Nantucket Island. Fortunately, the easterly Relying on the buoy, rather than straining people
winds were short-lived and the oil never came ashore. and aircraft to their limits, marked the beginning of the
Neither, for that matter, did a single one of 9,000 drift New Year. The slick, if it could still be called that, covered
cards deployed on December 26 and 27. After dropping thousands of square miles of sea surface, and was simply
their drift cards on December 26th, Deaver and his crew unmappable. The buoy positions for the next 25 days are
proceeded to relocate "Pancake I", shown on Figure 10 shown in Figure 12 [Slide 201. This buoy stopped trans-
[Slide 17]. mitting in September 1977; at that time, the buoy was
On the 27th of December (Monday), the final reported to be about 200 miles off the Azore islands.
"complete" map of the slick was made, though the Thus was marked an end to this, one of the most specta-
easternmost boundary could not be ascertained as cular, and probably the most closely-watched, oil spills
shown in Figure 11 [Slide 18]. In addition, the Coast Guard of all time.
conducted a "burn test" that day [Slide 19; Photograph 43,
p. 111-22, Grose and Mattson, 19771 on one of the huge
floating pancakes (this'one was estimated to be about
300 ft. in diameter). Needless to say, the pancake did
not burn.
Poor weather precluded any aerial observations in
the next two days, December 28th and 29th. On the 30th,
the Coast Guard HU-1 6 lost an engine and caught fire on
takeoff, terminating that day's operations. On the 31 st of

The Behavior of Floating Oil From the
Argo Merchant

P. L. Grose

Center for Experiment Design and Data Analysis
Environmental Data Service/NOAA
Washington, D.C.

Abstract is most likely chemical fractionation (Phillips and
Groseva, 1975), where the lighter fractions physically
The Argo Merchant oil formed thick pancakes separated from the bulk of the oil in the pancake and
surrounded by a thin sheen most likely because of spread into a thin lens because of lower surface tensions.
fractionation of the bulk oil and spreading of the lighter The pancakes were fluid and changed shapes rapidly.
components controlled by lower surface tensions. These shape changes appear to be related to the dif-
Measured differential velocities of the pancakes and ferential velocity of the oil relative to the surface waters,
surface waters are consistent with a wave/oil interaction rather than to the spreading of the oil into a thinner lens.
model that predicts oil velocities of twice the surface This conclusion is drawn from the observation that the
Stokes drift. The thicknesses of the oil pancakes are not pancakes did not diminish in thickness with time.
consistent with a model that considers only static forces. On December 23, 1976, divers from the U.S. Navy
However, when the dynamics of the oil movement are Audio Visual Command dived beneath the oil slick and
included in the model, the predicted results are more photographed the bottom and edges of an oil pancake
realistic, with thicknesses related to the square root (Slide 30). Three important conclusions can be drawn
of the surface tension at low speeds and to the square of from the photographs and the observations by the divers.
the oil/water differential velocity at higher speeds. First, the bottom of the pancake they saw was similar in
appearance to its top surface; that is, it was relatively
Discussion flat and smooth and, most important, did not have a keel or
other protuberances extending into the water below.
During the Argo Merchant oil spill, unique oppor- Second, the edges of the pancake were well defined and
tunities arose for observing the physical behavior of the relatively square in shape. They did not appear to taper in
floating No. 6 oil. This paper describes some of the thickness as might be expected with a lens shape. Also,
observations made during the oil spill (Grose and the interface between the oil and water was well defined
Mattson, 1977). In particular, the appearance of the and there was no visible boundary layer between them
pancakes that the oil formed, their morphology, and their other than the interface itself. In other words, while the
movement are discussed. These observations are pancake was not well coupled to the water column,
examined in light of models that attempt to explain them. neither was it totally decoupled by streamlining. Third,
As the oil flowed out of the tanker, it was advected when the pancake was broken by the divers'air bubbles,
away by primarily tidal currents. Although in gross the rend would heal within a matter of seconds, indicating
appearance the oil slick appeared to be continuous that there were forces acting on "or within the pancake
(Slide 4*), in reality it was composed of thick oil pancakes to maintain it as a distinct entity.
surrounded by a thin sheen and open water. The pro- It has been known for some time that oil moves
portions of these three components of a typical oiled faster than surface waters. Several experiments were
area shortly after the spill were estimated roughly as 40% conducted to determine this relative velocity during the
clear water, 60% sheen, and less than 1% of the area Argo Merchant spill. The best documented experiment
covered by the thick pancakes (Slide 39). The pancakes was conducted at 1100 on December 19,1976, at the end
could be differentiated from the sheen by their smooth of a 1000-foot streamer of oil heading east near the far
black surfaces, while the sheen was irridescent and end of the slick (Slide 22, right hand corner near 41 03'N;
easily broken by waves. The cause of the sheen formation 69 19'W). A line of dye pills was dropped from a heli-
copter onto the water directly downwind of the oil and the
*Copies of the color slides may be found in Appendix III of Grose oil overtaking the dye was observed (Slide 35). The dye
and Mattson, 1977. pills were dropped approximately 90 ft. apart. After 205

under the static conditions that apply to Langmuir's
25 model. The oil was moving faster than the surface water
and thus one would expect a compressive force to be
20 7 S acting on the oil caused by a drag exerted on a pancake
as it moves through the surface waters and by the force
accelerating it. This compressive force would act to
L5 increase the thickness of the pancakes. For this model
E we have a balance between the integrated hydrostatic
forces and the surface tensions of the two oil surfaces
within the pancake, and the equivalent integrated hydro-
2 1 static pressure, water surface tension, and the drag force
outside the oil. This model, which is identical to Lang-
I first break (21 Dec 76)
05 2 second break (12 Dec 76) muir's model with the addition of the drag force, is
(wind speed) represented numerically by

0 0,5*po* 9*T 2
6. 81 1; ' 1; to _ _foW = 0. 5 * PW * 9 * (T polp,,V) 2 . -yW + F D 0
TIME (days)
Figure 1. Estimated Oil 7hickness As a Function of Estimated where the subscript 0 refers to the oil and W to the water,
Age. i are surface tensions, P are densities, g is gravitational
acceleration. T is the thickness of the oil lens, and FID is
the drag force. From dimensional reasoning, this drag
force should be proportional to the thickness, the square
of the oil velocity (U), and a drag coefficient (CD), which is
sec (Slide 36) the oil had reached the nearest dye pill. in turn a function of the Reynolds number for the water:
The velocity of the oil relative to the surface water (5 cm)
was calculated to be 5.6 cm/s in an easterly (downwind) FID = CID * T * U2 (2)
direction. With these two slides one can also measure the
elongation of the dye tails that resulted from a shear
between the surface and about 50 cm depth as well as
from some windage on the pills themselves. The Substituting (2) into the quadratic (1) and solving for the
measured elongation velocity was 5.6 cm/s, again in the thickness:
downwind direction. Relative to wind speed, both of these U2
speeds are about 1.1 % of those measured at 19 rn eleva- T=CD* A + ( U4 - 2 * A - FS)1/2 / A (3)
tion by a Coast Guard cutter 7 miles away. The differential
velocity measurements agree quite nicely with the model In eq. (3) we have defined the spreading force FS and a
of oil movement proposed by Milgram (1 977A, B). In this density function A to simplify computations:
model the oil surface is predicted to move at about twice
the Stokes drift speed as an upper limit because of FS =YW - Ito -'YoW (4)
interaction with waves generated by local winds. The
elongation velocity of the dye tails should be a good and
approximation of the Stokes drift. A=9*(PW-Po)*PO/Pw
On several, occasions visual estimates of the thick- (5)
ness of the oil pancakes were made. Figure 1 illustrates
four such estimates as a function of the age of the pan- If we assume that the drag coefficient for the oil lens is
cake. The thickness appears to increase with age, but approximated by the coefficient for an infinite cylinder, at
there were differences in wind speeds, and the thinner Reynolds numbers of approximately 2000 for the
estimates were made under lighter wind conditions. The observed thickness and speed we have:
increase in thickness is of considerable interest because
thick oil is easier to recover and because this phenome- CD= 1. (6)
non appears to be fundamental to the dynamics of some
spilled oils. One possible explanation for the variation is An estimate of the spreading force (FS) can be made by
an increase in surface tension as the oil ages. Langmuir using measured values for the surface tensions of the oil
(1933) developed a relationship between the surface and the water of 35 and 72 dynes/cm, and assuming the
tensions and densities of a two-part immiscible fluid water/oil interface tension is 69 dynes/cm. This assumed
system with the thickness of the fluid lens. In his model the value is based on Milgram's measurements of the surface
thickness is a function of the square root of the spreading tension of water with a thin film of oil on top of 59 and of
force (FS) times the buoyancy term (A) derived from the water pipetted from directly beneath oil of 79 dynes/cm
densities. Applying actual measurements of the Argo (Grose and Mattson, 1977, pg. 71). These measurements
Merctiant oil gives a computed thickness for the pan- and assumed values result in a spreading force of -32
cakes of 1 cm. Using an estimated maximum surface dynes/cm. A numerical value of A, eq. (5), is computed to
tension (double) to which the oil might have aged results be 62 from measured specific gravities of the oil and
in a maximum computed thickness of only 2 cm. This water of 0.96 and 1.028 respectively. Furthermore, we
computation does not agree with the observations in can approximate the speed of the pancake as 1. 1 % of the
spite of being based on unreasonably large surface wind speed from the experiments described earlier. Sub-
tensions. However, the observations were not made stituting the above values into eq. (3) thicknesses of 1.5,

25 A 19 Der 76 This paper has attempted to demonstrate that it is.
B 23 Dec 76 not sufficient to consider oil as a static substance on the
C 25 D r76
C, 25 .:r76 12 hours earlier water surface. The local dynamics appear to be critical in
D,
D 11 11:1 116 L2-24 hours earlier
It 0 c I trying to predict its behavior. Some aspects of the
postulated dynamics agree with the observations: Sheen
0 Cd - 2 formation controlled by fractionation and spreading
15 D dependent on surface tensions; differential velocities
related to interaction of wind waves with oil pancakes;
and to a limited extent oil thickness determined by
5 C surface tensions at low pancake speeds but by drag
5 0 o' Cd - I forces at higher speeds. However there are still some
inconsistencies and many unknowns. Future research
5 01 r' od.1 should be directed toward resolving these inconsisten-
L..... !@ So,
Pr P.S . del cies and unknowns. Among these are: Where is the shear
0 (Cd - drag coefficient) located between the oil surface and the water below?
0 ............ What is the drag coefficient for an oil pancake? What are
5 10 J5 20 25 the streamlines for the water flow in the vicinity of a
WINI) SPEED (MIS) pancake? Once these questions have been answered,
Figure 2. Oil Thickness As a Function of Wind Speed. the models will be capable of predicting how the oil will
behave in a more realistic manner.

&0, 5.4, and 8.8 cm are computed for wind speeds of 5, References
10, 15, and 20 m/s (Figure 2). Doubling the drag coeff i- Grose, P. L. and J. S. Mattson, 1977: TheArgo Merchant Oil Spill.
cient (CD) to 2 yields an even better agreement as indi- National Oceanic and Atmospheric Administration, Dept.
cated in this figure. While these thickness values are not of Commerce. 133 p. (Available from the Superintendent
as large as the visual estimates indicated in the figure, of Documents, U.S. Government Printing Office,
they do bring the theory closer to the observations and Washington, D.C. 20402).
indicate that the hydrodynamics of the oil pancakes is Langmuir, 1., 1933: Oil Lenses on Water and the Nature of
important in determining their behavior. What remains is Monomolecular Expanded Films. J. Chemical Physics,
to properly measure the drag coefficient for an oil V. 1. pp. 756-776.'
pancake, but this is beyond the scope of our discussion Milgram, J., 1977A: Mass Transport of Water and Floating Oil by
here. Gravity Waves in Deep Water. MIT. 48 pp. Unpublished
As the oil on the water surface ages, it will weather Manuscript.
by fractionation and evaporation. Because of the Milgram, J., 1977B: Measurements of the Transport of Water
and Floating Oil by Gravity Waves in Deep Water. M.I.T.
weathering, one would expect the viscosity to increase. 46 pp. Unpublished Manuscript.
This viscosity increase does not cause the pancake to Phillips, C. R. and V. M. Groseva, 1975: Separation of Multi-
become thicker other than by slightly increasing its rela- component Hydrocarbon Mixtures Spreading on a Water
tive velocity, as indicated by Milgram's model, but Surface. Separation Science, V. 10. pp. 111-118.
viscosity does effectively increase the amount of time
required for the pancake to reach equilibrium with the
forces acting on it. Therefore, some form of time inte-
gration of the forces acting on the pancake must be used
when relating Eq. (3) to the real world, which is in dynamic
rather than static equilibrium. This may help to explain the
deviant point in Figure 2 for December 31. While the winds
averaged only 7.5 m/s for the 4 hours before this
measurement, from 24 to 12 hours before the measure-
ment they averaged 18 m/s. That the viscosity of the oil
was high is not the question, since a piece of 2 x 4 lumber
on the pancake surface was observed to be sitting with its
long face vertical.
Without postulating increases in surface tensions,
we thus find the data on pancake thickness increase to be
consistent with almost all the facets of a model that moves
oil by wind waves. The only prediction not supported by
observations is the change in pancake speed between
the top and bottom of the lens as observed by Milgram
(1977B). Drift cards placed on pancakes did not tend to
move toward the leading edge as is implied by a shear
profile throughout the thickness of the pancake. While a
shear must exist, it may be that it is concentrated at the
oil/water interface. Certainly this needs to be investi-
gated further, as the shear will directly influence the
amount of oil that gets accommodated into the water
column.

Can Oil Spill Movement Be Predicted?

Ivan Lissauer and Pat Welsh

U.S. Coast Guard Research and Development Center
Groton, Connecticut

Abstract for predicting the movement of oil spills in estuarine and
offshore areas. These techniques involve vector addition
On-1 5 December 1976 theArgo Merchant grounded of the forces that move the oil. For. estuarine areas this
on Nantucket Shoals approximately 28 miles southeast of generally includes addition of wind, tide, and fresh water
Nantucket Island. Immediate forecasts of, the movement flow vectors; and wind, tide, and semi-permanent current
of the oil should a spill occur from the ship were required vectors for offshore areas as shown by Hufford et al.
by Coast Guard Marine Safety Office (MSO), Boston. A (1975) and Lissauer and Bacon (1975). These techniques
description of the forecasts provided to MSO, Boston, is were applied to forecast the movement of oil from the
provided. The method used for predicting the oil move- Argo Merchant Results of the forecast are given, as well
ment was a simple vector addition of the tidal velocity as a retrospective discussion of the validity of the tech-
and 3.5% of the wind speed on an hourly basis. The lateral niques used based on the actual movement of the spill.
spread of oil was determined from the tidal velocities.
Results of the forecasts compared to the actual move-
ment of the oil are given, as well as a retrospective Forecast
discussion of the validity of the techniques that were
used. Forecast of the movement of the oil began on 15
December 1977. Data for the forces which transport the
oil were obtained and examined. The magnitude of the
Introduction wind vector from predicted values of wind speed would
move the oil at a rate of 0.5 to 1.5 knots in a downwind
On 15 December 1976 theArgo Merchant grounded direction; the magnitude of the tidal currents would move
on Nantucket Shoals approximately 28 miles southeast of the oil at 0.5 to 1.3 knots; the magnitude of the permanent
Nantucket Island (Figure 1). The Coast Guard R&D Center currents in the area are approximately 0.1 to 0.2 knots.
has been involved in developing forecasting techniques After comparing the magnitude of the forces that move
the oil, it was determined that for short-term predictions
NANTUCKET (i.e., <12 hours) the tides and winds would control the
MLAND movement of the oil. For long-term predictions the winds
alone would dominate the movement of the oil. The
method used for predicting the movement of the oil was a
simple vector addition of the tidal vector and 3.5 percent
<
of the wind speed on an hourly basis. The lateral move-
J
ment of the oil was determined to be the magnitude of the
tidal movement. This combined with the hourly vectorial
movement of the oil showed the extreme estimate of
the areal extent of the oil.
On 16 December at 0930, the U.S. Coast Guard
NANTUCKET TO AMBROSE Marine Safety Off ice (MSO), Boston, requested a forecast
NANTUCKET WHT of the movement of oil should the tanker rupture. At 1135,
AMBROSE TO NANTMKET -
MSO, Boston, was informed that the oil would move south-
east for the next,24 to 48 hours. In addition, the oil would
continue to move southeast to east through the weekend
SCALE(MLES) of 18 and 19 December. This forecast was based on the
I A I I I - 1
0 5 10 15 20 25 JU J5 40 45 50 tides and the predicted winds supplied by the National
40*t + --j-40- Weather Service (NWS). MSO, Boston, was also informed
70* 69* 68. that the oil would continue to move offshore as long as
Figure 1. Location of the Argo Merchant the winds were offshore.

%NANTUCKET 1600 IS DEC-0600 16 DEC
ISLAND NANTUCKET
70' 0700-1300 DEC 16 Be* ISLAND PREDICTED LIMIT OF SLICK
41. + + 41' 70* AS OF 0700 DEC 23 Be*
1300 DEC 16-1900 DEC 17 41*+ +41-
0700-1300 16 DEC ':@@
<.SPC
PREDICTED LIMIT OF <.?o
SLICK AS OF 1900 DEC 17
1300 DEC 16-1900 DEC 17

NANTUCKET
NANTUICKETLIGHT LIGHT

PREDICTED WINDS HEAVIEST CONCENTRATIONS OF OIL
SHOULD BE E-SE OF ARGO-MERCHANT
1600 DEC 15-0600 DEC 16 W@NW 15-20KTS. (8 MILES)
0700-1300 16 DEC NE 15-20KTS. 16 MILES) PREDICTEDWINDS
1300 DEC IS- 1900 DEC 17 NW 15-25KTS. (21 MILES) W.NW35KTS. 467Md..)
40*4- + +40- Figure 5. Predicted Limit of Slick as of 0700 23 Dec.
70* 69* 68,

Figure 2. Predicted Limit of Slick as of 1900 17 Dec.

NANTUCKET NANTUCKET
ISLAND ISOODEC17-2100DEC19 ISLAND

70. 681 70* 68'
41'+ +41' 41'+ 11 OBSERVED LIMIT OF +417
REDICTED LIMIT OF SLICK SLICK AS OF 21 DEC
PREDICTED LIMIT OF SLICK AS OF 2 100 DEC 19
AS OF 1900 DEC 17 i@O
1900 DEC 17-2 100 OP
DEC 19 0'e
?400
NANTUCKET occ I
LIGHT ?0
PREDICTED WINDS NANTLICKET
ISOODEC17-2100DECIS W-NW2SKTS. (44MQS)

Figure 3. Predicted Limit of Slick as of 2100 19 Dec.

40'+ + +40'
70' 69, 68*
Figure 6. Comparison of Observed and Predicted Limit of Slick
as of 21 Dec.
I? PREDICTED LIMIT OF
% C:A Q SLICK AS OF 2400 DEC 20
NANTUCKET
ISLAND
70' the forecast as well as the total movement of the oil
4 J.+ elf +41
a caused by the predicted winds are shown. From 1600,
N 15 December to 0600,16 December, the west-northwest
PREEN TED LIMIT OF
SLICKCAS OF 2100 DEC 19 winds and tides moved the oil eight miles to the east-
northeast. From 0700 to 1300, 16 December, the north-
east winds moved the oil towards the southwest a
distance of six miles. The spill was treated as a con-
NANTUCKET tinuous leak. Therefore, the oil movement was con-
LIGHT PRE @CTED WINDS tinuously predicted from the site of the Argo Merchant.
2 100 DEC 19-2400 DEC120 S-SW30XTS. (28MLEs) This is the reason there are two vectors labeled 0700-
Figure 4. Predicted Limit of Slick as of 2400 20 Dec. 1300 16 December. One of these vectors shows the
transport of the oil that moved northeast prior to these
wind conditions. The othervector shows the movement of
oil from the wreck site during the period 0700-1300,
Figures 2 through 5 depict the predicted movement 16 December. This process was used for the entire
of oil. These figures were the basis for the information period. Thus the boundaries of the oil spill shown on
:@@07
00 @ @RF 'I-
P MTE"MIT. 3 :.
1 30.. A. OF 117- DEC _1@.

N
T

OF
.NT

PFIEGICTE...

REE"C'
AS.
N:ET

which was transmitted to MSO, Boston. The long-term Figures 2 through 5 are, in essence, an estimate of the
movement is based on the predicted winds as supplied extreme outer limits of the oil slick.
by the NWS. Figure 2 shows the predicted limit of the slick Figure 6 is a comparison of the observed limit of the
as of 1900,17 December. The predicted winds used for slick on 21 December (obtained from overflights) and the

NANTUCKET
ISLAND NAITrUCKET ISLAND
70* a
41 41'
OBSERVED LIMIT OF SUCK AS OF 23 DEC

NANTUCKET
UGW 84/C'4,1

Figure 7. Comparison of Observed and Predicted Limit of Slick
as of 23 Dec.

SCALE(WLES)

5 10 is .70
0

NANTUCKET Figure 9. Comparison of Observed Slick and Predicted Move-
GLAND ment Caused by Tides and Wind
To- Be.
41 +41'

CAPE COD

NANTUCKET
AND
NANTUCKET ISL
WAIT
Figure 8. Progressive Hourly Wind Vector Diagram Using 3.5%
of the Wind Speed 1Z, 'ri"* 14"
41' +
4 ib.100%" 68*
_@o

estimate of the slick limit predicted from the winds as of
2400, 20 December. Figure 7 is a comparison of the
observed limit of the slick on 23 December and an
estimate of the slick limit as of 0700,23 December. Again, ATLANTIC 100
the observed limit of the slick was obtained from over-
flights of the area. In Figures 6 and 7 the predicted
direction of movement of the oil is in excellent agreement
with the observed movement. The predicted areal 1000 F.
coverage for both figures is about twice the observed + 100 + +
coverage. This entire prediction was accomplished in 70* 69- 68.
approximately three hours from the time of request. It Figure 10. Observed Area of Oil Spill on 17 Dec.
indicates that vectorial addition of the forces that move
the oil is an excellent method to obtain quick response
answers to where the oil will go and when it will get
there. Had the winds been onshore instead of offshore USCGC Vigilant. This wind data and tidal data (from the
this method would have enabled cleanup equipment to be charts) can be used to compare actual short-term move-
placed at strategic areas before the oil came ashore. In ment of the oil with the predictive technique of adding
addition, it seems likely that had actual on-scene winds winds and tides vectorially. In addition, the validity of
been used in the forecast rather than long-term predic- using 3.5 pe Ircent of the wind speed in a downwind
tions from NWS, the predicted movement and dispersion direction can be examined.
would have been even more precise. For short-term drift, Figure 9 gives a comparison of
observed slick and predicted movement caused by tides
Observations and winds. The vector shows the predicted movement of
the oil for the period 2400, 15 December 1976 to 0900,
Figure 8 is a progressive hourly wind vector diagram 17 December 1976. The outline of the slick was taken
.11. 1.11 OF 1111:11 AS OF 23 MC

741

IO..Fm

using 3.5 percent of the wind speed for the period 1600, from Figure 10 which shows the observed slick deter-
15 December to 0700, 23 December. The diagram was mined near noon on 17 December. There is excellent
prepared from on-scene wind data collected by the agreement between the actual directional movement of

ce
A t"

j@
f. V
NANTUCKET NANTUCKET
SLA %
$LAND

LORAN
%
PANCAKES TO Z
9.
+
4 41'+ 41 416 41
So-g! 68. 14
90-96%

ATLANTIC
100
F@

I DOO r-
40'+ 100 F. 40* 40+ 1 000 Fm 100F.
70' 69. 68.
Figure 11. Observed Area of Oil Spill on 18 Dec. Figure 12. Observed Area of Oil Spill on 19 Dec.

the oil and the predicted movement as determined from The observed movement of the oil spill is docu-
tides and wind. In fact, it appears that 3.5 percent of the mented in Figures 11 through 16. Table 1 indicates that
wind speed adequately describes this transport vector. the oil was moving westward bearing 240*T on 16
The greatest error occurs in predicting the tidal December. The wind vectorduringthe morning of the 16th
component for each hour of movement. was 2457. On 18 December, pancakes were found 27
During the period 15 December to 19 December, the miles east of the ship (090'T). The maximum eastward
total extent of the spill was not clearly defined in over- movement of the oil caused by the wind (Figure 8) is 31
flights. However, several observations of the movement miles bearing 130'T. Assuming this is the maximum
of the oil during and after this period verify the techniques eastward extent of the oil the computed wind factor for
used. These observations are shown in Table 1. moving it would be 3.05 percent. Commencing on the

Table 1. Observations of Oil Movement

Predicted Movement of Oil
Date Actual Observations Using 3.50% of Wind Speed
of Oil Movement in a Downwind Direction

12/16 Oil 2 mi. N to S and 4 mi. E to Winds during the morning of 16 Dec.
W; oil moving west streak of would move oil on a bearing of
oil (bearing about 240'T). 245T.

12/18 Pancakes 27 miles east of ship (090'T). Maximum movement east, 31 miles,
Computed wind drift 3.05% bearing 130'T.

12/20 Main plume 16 miles long bearing Winds 0900 19 Dec. to 0900 20 Dec.
040*T. Computed wind factor would move oii 14 miles, direction
4.00%. 048*T.

12/21 Maximum eastward movement of Maximum eastward movement 60 miles,
oil 53 miles, direction 090*T. direction 095*T.
Computed wind factor 3.09%.

12/22 Maximum eastward movement of oil Maximum eastward movement 82 miles,
95 miles, direction 108'T. direction 106*T.
Computed wind factor 4.05%.
'00
F-
@1.. F.
41 @411

12/23 Maximum eastward movement of Maximum eastward movement 97 miles,
oil 86 miles, direction 1 00*T. direction 11 O'T.
Computed wind factor 3.10%.

CAPE COD + CAPE COD
68'
J
PANCAKES

14 NANTUCKET
NANTUCKET _1SLAND PANCAKES
LIGHT
1:@L @A
RAINBOW PANCAKES
M,
10-20%
q RAINBOW
+ 41'
<11% &
I IGHT PANCAK S 67
RAINBOW
4 1@0; RAINBOW
LIGHT
68 ANCAKES &
t"@ 90% 5% PANCAKES
HEAVY 10%
PANCAKES
20-30%
ATLANTIC 100 ATLANTIC

IOOIFM 69- 1 OOFM +
68'

Figure 16. Observed Area of Oil Spill on 23 Dec.
100 F. 1000 Fm
Figure 13. Observed Area of Oil Spill on 20 Dec. 20th, the wind factor for the observed movement and the
direction of movement of the oil are given in Table 1 on a
daily basis. A summary of these values compared to
predicted values is shown in Table 2.
CAPE COD One way analysis of variance without replication
was
performed to determine if statistically significant
differences were detectable between observed and pre-
dicted wind factor and direction. The analyses are
NANTUCKET PANCAKES
Is. summarized in Tables 3, 4, 5, and 6 respectively.

LORAN
Conclusions
TOWER
+41'
PANCAKES 1. For short-term predictions (i.e., <12 hours) a
i C. 68*
90-95% vectorial addition of wind and tides provided good agree-
MIXED HEAVY@ S ment between the actual direction movement of the oil
To PANCAKES and the predicted movement as determined from tides
and winds. The greatest error occurs in predicting the
100 tidal component for each hour of movement.
ATLANTIC L!_1

1OOF. 1000 F. 1000 F. Table 2. Comparison of Observed Versus Predicted Movement
Figure 14. Observed Area of Oil Spill on 21 Dec. Date Observed Predicted
Wind Factor Direction Wind Factor Direction

12/20 4.00% 0400T 3.50% 0480T
12/21 3.09% 090*T 3.50% 095'T
HEAVY OIL 12/22 4.05% 108*T 3.50% 106'T
CONCENTR
- ON A
(90-95%)
12/23 3.10% 1 00*T 3.50% 11 O'T
NANTUCKET
ISL:4
ME"
MODERATE CONCENTRY
ATIONS
PANCAKES PANCAKES & STREAKS

Table 3. Data Array for Wind Factor
+ 41'
170'1@ 67'

Date Observed Predicted
LIGHT
PANCAKES 12120 4.00 3.5
12/21 3.09 3.5
F.

..F
1 .00

12/22 4.05 3.5
1OOFM
IN 1000 @FM@ 12/23 3.10 3.5
3@,@.56 i 3.5 _x 3.53
Figure 15. Observed Area of Oil Spill on 22 Dec.

2. For quick response answers to where oil will go
and when it will get there, vectorial addition of the forces
that move the oil is a good method. Use of the 3.5% value
for the wind drift factor is of great benefit if no model is
availableto depict the wind drift current for the area where
a spill occurs.

Table 4. ANOVA for Wind Factor

Degrees
Source of Amount of of Estimated Observed
Variation Variation Freedom Variance F Ratio

Among 0.0072 7 0.0072
Within 0.8662 6 0.1444 0.0499

Total 0.8734 7

Conclude that for the 5 percent level of alpha there is no significant
difference between the observed and the predicted wind factor.

Table 5. Data Array for Predicted Movement

Date Observed Predicted

12/20 040*T 048*T
12/21 090 095
12/22 108 106
12/23 100 110

T8450 T 89.7 5 x 87.13

Table 6. ANOVA for Wind Direction

Degrees
Source of Amount of of Estimated Observed
Variation Variation Freedom Variance F Ratio

Among 55.13 1 55.13 0.06
Within 5247.75 6 875.63

Total 5302.89 7

Conclude that for the 5 percent level of alpha there is no significant
difference between the observed and the predicted wind directions.

References

Hufford, G. L., Lissauer, 1. M. and Welsh, J. P. (1975). "Movement
of Spilled Oil Over the Beaufort Sea Shelf - A Forecast,"
National Technical Information Service.
Lissauer, 1. M. and Bacon, J. C. (1975). "Predicted Oil Slick
Movement from Various Locations Off the New Jersey-
Delaware Coastline", National Technical Information
Service.
Mardia, K. V. (1973). Statistics of Direction Data, AcademicPress,
New York.

Risk Forecasting for the
Argo Merchant Spill

Timothy Wyant and Richard A. Smith

U.S. Geological Survey
National Center
Reston, Virginia

Abstract vicinity of a spill long after the actual onset of spillage.
An oilspill trajectory model, originally developed to The model was designed to provide estimates of such
assess environmental risks of Outer Continental Shelf long-term risks. It thus serves a different purpose from
oil production, was used during the Argo Merchant spill models developed to provide quick accurate forecasts of
to forecast the risk to various shoreline and marine imminent movement of a slick based on instantaneous
resources. The model indicated a low risk to these receipt of short-term weather forecasts, current and
resources given the location and season of the spill and weather conditions, and the precise state of the slick.
the particular wind conditions under which the spill Risk arising from the Argo Merchant spill was fore6ast
occurred. Oil from the Argo Merchant in fact contacted for periods up to 60 days under different assumptions
few of these resources. However, had a spill at this concerning initial wind conditions and spill duration.
location occurred either under other typical wind condi- The model can be used to answer numerous
tions for the season or at a different time of year, the special case" questions of interest which inevitably
model also indicated that risk would have been much arise in assessing risk from any particular spill or offshore
higher. Quantitative estimates of risks were constructed production scenario. As an example, it was unclear at
assuming different initial conditions, seasons, and the time of the Argo Merchant spill whether considerable
durations of spillage. amountsof oil might remain in the hull. The question arose
as to the time of year to destroy the remaining hull frag-
ments to minimize the risk from the resultant release of oil.
Introduction The model can provide estimates of risk as a function of
spill date as an initial answer to such questions, and was
An oilspill trajectory model, originally developed to employed for this purpose during theArgo Merchant spill.
assess olispill risks associated with Outer Continental
Shelf oil production, was used during the Argo Merchant
spill to give quantitative estimates of resulting risk of Description of the Model
beaching and impacting various resources. The model
has been used to assess the risks associated with Outer The model simulates the movement of hypothetical
Continental Shelf oil production in several frontier areas spills using deterministic ocean currents and stochasti-
(Slack and Smith, 1976; Slack, Smith, and Wyant, 1977; cally generated winds. In the North Atlantic area, these
Smith, Slack, and Davis, 1976a, b). Because the Argo currents are derived from drift-bottle studies (Bumpus,
Merchant broke up in one of these areas, the data to 1973), and the statistical descriptions of winds are
drive the model had been fully prepared in the region derived from 5 years' record at Georges Shoals and
surrounding the grounding site before the incident Nantucket Shoals weather towers.
occurred. Thus, risk forecasts could be made beginning A first-order Markov model for winds is the heart of
with the first grounding reports. the risk-forecasting abilities of the model. Historical
Risk is defined to be the probability that within a records of winds for 3-hourly observations are grouped
given time spilled oil passes through an area occupied by into 41 classes depending on magnitude and direction
a shoreline or marine resource at a time of year when the and analyzed for frequencies of transitions between
resource is considered potentially vulnerable to oil. In classes. Given the date and location of a spill, an array
general, for large oilspills, spillage can continue for many of possible future wind patterns on the basis of past wind
days, and oil, once spilled, can persist in observable transition frequencies can be generated. Superimposing
quantities in the ocean for a considerable period. In the effects of the forces exerted by the ocean currents,
either case, spilled oil may contact resources in the winds, and coriolis effect, a hypothetical oilspil! trajectory

Table 1. Assumed Seasonal Vulnerability of Resources.
T, assumed vulnerable;
F, assumed not vulnerable

Resource Month

Dec. Ja n. Feb. Mar. Apr. May June July A ug. Sept. Oct. Nov.
1 Beaches and recreation areas T T T T T T T T T T T T
2 Wildlife sanctuaries and wintering areas T T T T T F F F F F F T
3 Coastal bird breeding areas T T T T T T T T T T T T
4 Pelagic bird nesting areas T T T T T T T T T T T T
5 Pelagic bird wintering areas T T T T T F F F F F F T
6 Eagle and osprey nesting sites T T T T T T T T T T T T
7 Cod and haddock spawning areas F T T T T T T F F F F F
8 Silver and red hake spawning areas F F F F F T T T T T F F
9 Sea herring spawning areas F F F F F F F F F T T T
10 Atlantic salmon migration routes F F F F F T T T T T T T
11 Shortnose sturgeon areas T T T T T T T T T T T T
12 Shellfish areas T T T T T T T T T T T T
13 Harbor sea[ whelping areas F F F F F T T T T T F F
14 Grey seal whelping areas F F T T T T T F F F F F
15 Salt marshes T T T T T T T T T T T T
16 Eel grass beds T T T T T T T T T T T T
17 Kelp beds T T T T T T T T T T T T
18 Rocky coastline T T T T T T T T T T T
19 Sandy beaches T T T T T T T T T T T T
20 Nova Scotia T T T T T T T T T T T T
21 Sewage clumpsites T T T T T T T T T T T T
22 Sand and gravel deposits T T T T T T T T T T T T

for each of these samples of wind patterns can be com- Validityof thehfodeL Theabilityof any model toac'curately
puted. Tabulating the contacts of these trajectories forecast. is directly tied to the quality of the driving data.
with land or other resources then provides quantitative Clearly, wind records from -particular sites over finite
estimates of the risk to each resource. Because these periods of time do not precisely represent wind patterns
estimates depend upon the initial conditions under which to be found, at other sites and other time periods; eq u*afl'y'
a spill occurs, the model can be used to assess sensitivity clearly, drift-bottle studies imperfectly represent the
of risk to the location and duration of a spill, the date at current patterns of' any large area. Unfortunately, little
which it occurs, and the specific wind conditions under quantitative assessment of the sensitivity of model
which it occurs. results to these imperfections can be made; judgments
Risk arising from the Argo Merchant spill was fore- as to the validity of oilspill trajectory model predictions
cast for periods up to 60 days. Implicitly, the model in light of uncertainty in the basic data remain largely a
provides a pessimistic estimate of risk in that no allow- matter of "engineering judgment".
ance is made forweathering of oil in that period. However, More extensive and precise work has been done
in model runs elapsed time between oil spillage and investigating the behavior of simple vector addition
resource contact is tabulated, allowing refinement of risk advection models such as used here for risk forecasts -
estimates in light of whatever weathering assumptions advection = current velocity + (wind drift factor) x (deflec-
are desired. tion adjusted wind velocity).
The hypothetical spills generated by the model A general review of oilspill trajectory modeling
move in straight lines over each 3-hour time step in the techniques (Stolzenbach et al., 1977) concludes
simulation. The direction and distance of the 3-hour unsurprisingly that such a simple advection model
movement is calculated as the vector sum of the current "represents a considerable oversimplification of a very
and 3.5 percent of the prevailing wind. At each step, the complex process" and discusses a number of potential
location of the hypothetical spill is checked against a improvements. However, the review goes on to note that
table of resources locations and the simulated date most modelers have found such simple advection
checked against atable of seasonal resource vulnerabili- equations to be an adequate approximation to the actual
ties. Any simultaneous occupation of a one-mile square advection process. Indeed, the review also points out
area by the simulated spill and a resource is tabulated (section 5, p. 50) that at present most potential improve-
as a contact. Table 1 shows the resources considered ments to this simple advection model cannot' b ,e im-
and their assumed seasonal vulnerabilities. The locations plemented for models meant to have general applica-
and seasonal vulnerabilities were estimated by Bureau of bility; many of the relevant physical processes "have no
Land Management personnel during the oilspill risk analytical description available", where analytical des-
analysis for the proposed North Atlantic Outer Continen- criptions exist, they are often unproven; where proven,
tal Shelf production areas (Smith et al., 1976b). Maps there is seldom enough widely available data to make
of the assumed resource locations appear in that inclusion in a multipurpose model possible.
report. The review also details laboratory experiments and

Table 2. Risk to Land As a Function of Initial Wind.

-42* Estimated
Estimated mean days
Initial wind pattern probability to come for spills to
ashore come ashore

Initial wind selected 0.10 8
randomly from his-
Model prediction toric winter record

Argo Me rchant
Initial wind northeast 0.24 5
10 knots (reported by
USCG, December 16@
Observed slick

Initial wind northwest 0.07 8
-40* N at 20 knots (reported
from Nantucket L igh t
701 W 689 W Ship, December 17)
I
Figure 1. Model Performance.

a discussion of the effects of different deflection angle
actual ocean observations used to determine appropriate assumptions on performance of the model in tracking
values for wind drift factor and deflection angle for the the actual spill from the Argo Merchant, In the course of
simple advection models. In varying circumstances, the previously mentioned North Atlantic oilspill risk
empirically determined values have been reported in a analysis (Smith et al., 1976b), risk was estimated
range of 0.008 to 0.058 for drift factors and a range of assuming deflection angles of both 00 and 20' to the right.
0.3' left to 13.2' right for deflection angles (section 3, Estimates of risk to shore varied from 0.21 to 0.08 over
p. 81). The model used to forecast risk for the Argo this range of drift angles. This roughly indicates the
Merchant spill assumes a drift factor of 0.035 and a sensitivity of the model's risk forecasts to deflection
deflection angle of 0'. The National Oceanic and angle assumptions.
Atmospheric Administration preliminary report on the The model represents a spill as a point and neglects
Argo Merchant spill (Grose and Mattson, 1977) contains spreading effects, although resource locations are

Table 3. Risk to Resources As a Function of Initial Wind.
n, less than 0.005; g, greater than 0.995

Initial wind selected Initial wind northeast Initial wind northwest
Resource randomly from historic 10 knots (reported by at 20 knots (reported
winter record USCG, December 16) from Nantucket Light
Ship, December 17)

1 Beaches and recreation areas 0.11 0.24 0.07
2 Wildlife sanctuaries and wintering areas .11 .24 .07
3Coastal bird breeding areas .03 .03 .01
4Pelagic bird nesting areas n n n
5Pelagic bird wintering areas 9 9 9
6Eagle and osprey nesting sites .01 .02 n
7 Cod and haddock spawning areas .23 .23 .2
8 Silver and red hake spawning areas n n n
9Sea herring spawning areas n n n
10 Atlantic salmon migration routes n n n
11 Shortnose sturgeon areas n n n
1.2 Shellfish areas .02 .06 n
13 Harbor seal whelping areas n n n
14 Grey seal whelping areas n n n
15 Salt marshes .01 .02 n
16 Eel grass beds n .01 n
17 Kelp beds n n n
18 Rocky coastline .01 .02 n
19 Sandy beaches .11 .24 .07
N@l

20 Nova Scotia n n n
21 Sewage clumpsites n n n
22 Sand and gravel deposits .97 .94 .98

3-hourly winds reported by the Nantucket Light Ship from
the Argo Merchant site, is shown in Figure 1 super-
1.0- Initial wind imposed on an outline of the actual spill's oil pattern.
z
--- NE 10 knots Such a picture seems to be the best way of communi-
WNW 20 knots cating the validity of the risk forecasts using the model. It
0.8-
quickly and concisely imparts the level of approximation
0 represented by this model to users of the forecasts who,
W 0.6-
M ultimately, must subjectively assess the worth of the
0
forecasts in consideration of their own needs.

0.4-
Risk Forecasts
0.2- As the Argo Merchant began to break up, the model
was used to forecast the risk to shore and marine re-
0
Q: sources from the spill conditional on initial wind condi-
(L 0.0 tions. This was done by giving hypothetical spills an
0 5 10 15 20
SPILL DURATION IN DAYS observed wind for the first 3-hour time step and thereafter
driving them with stochastically generated winds based
on the past records. These conditional risk forecasts
1.0- Initial wind reflect not only the movement imparted by the wind in the
rn
LU first 3-hour period, but also the past observed persis-
LU
cc - - - NE 10 knots tence of this wind pattern.
0 < WNW 20 knots
Cr 0.8- Table 2 shows the estimated risk to shore as a
00
_J z
5 z function of initial wind; Table 3 shows the estimated risk
UJ 3. to resources as a function of initial wind. Each estimate is
2 0.6-
0 (n based on 300 simulated spills assumed to start on
0 le December 15. Table 2 shows that there is an estimated
!XTE 80 0.4- probability of 0. 10 of impacts to shore from spills occur-
-ce ring in winter at the site of the Argo Merchant grounding.
zi X
a 02- For the northwest 20-knot wind reported at about the
M z
< < onset of spillage, estimated risk was even lower. How-
M a
00 ever, had the spill occurred under the northeast 1 0-knot
0.0- wind reported one day before the spill actually began,
0 5 1 15 20
SPILL DURATION IN DAYS estimated risk to shore would have been 0.24, reflecting
the fact that this onshore wind pattern, while unusual in
Figure 2. Risk As a Function of Spill Duration. winter, often persists long enough when it does occur to
be able to drive oil from the Argo Merchant site to shore.
The preceding risk estimates assumed that a spill
widened in model runs so that any hypothetical spill occurs as an instantaneous release of oil. Figure 2 shows
passing near the actual resource locations will be estimated risk as a function of spill duration. These
counted as contact. However, by releasing a number of estimates were made by using the model to generate
point spills in sequence and moving them with an identical sequences of hypothetical point spills and driving each
overall wind at each time step, a 2-dimensional approxi- complete sequence with a common wind at each time
mate picture of a spill can be generated. Such a picture, step. Thus, each spill of more than 3 hours' duration is
using a 0.035 drift factor, a 00 deflection angle, and the simulated as a stream of points such as that in Figure 1.

Table 4. Risk to Land As a Function of Spill Date.

Estimated Estimated Total
probability to come probability to come estimated
Date ashore in U.S. ashore in Canada probability to come
(Nova Scotia) ashore

December 0.10 0.00 0.10
January .08 .00 .08
February .14 .02 .16
March .33 .09 .41
April .40 .12 .52
May .33 .34 .67
June .34 .58 .92
July .34 .57 .91
August .29 .38 .67
_'N
W
E
N
W
0
knots
20 knots

September .16 .05 .21
October .19 .03 .22
November .21 .01 .22

Argo
Merchant

Argo Merchant

Figure 3. Simulated Trajectories Winter. Figure S. Simulated Trajectories Summer

Argo Merchant

Figure 4. Simulated Trajectories Spring. Figure 6. Simulated Trajectories Autumn

Estimated risk was defined for this case to be the At one time, it was thought that demolition of
probability that one or more points of each generated lingering hull fragments might be necessary and that such
sequence crossed an area occupied by a seasonally demolition might release some remaining oil. Con-
vulnerable resource. The estimated probability that some sequently, risk forecasts for different seasons of spillage
oil from a long spill would reach shore rose to 0.40 for a were made. Figures 3-6 show 10 simulated trajectories
20-day spill from the Argo Merchant site, and for 15-day for hypothetical spills from theArgo Merchant site in each
spills and up the differences in estimated risk due to season. The actual spill followed a path similar to that of
different initial wind assumptions disappears. In the the hypothetical slicks in Figure 3. Table 4 shows risk
model it was assumed that the Argo Merchant grounding estimates for spills starting from the Argo Merchant site
site was in a cod and haddock spawning area, but that for each month. Estimated risk to any shore reaches
this area was not potentially vulnerable to oil until January peaks in June and July and estimated risk to the American
1. The sensitivity of risk forecasts to such assumptions shore reaches a peak in April. These estimates clearly
of seasonal vulnerability is also shown in Figure 2; fore- supported the natural preference for an early demolition
casts of risk to cod and haddock spawning areas rose to of remaining ship segments should it have been
near certainty for20-dayspills beginning on December 17 necessary.
from the Argo Merchant site.

Summary
The spill from the Argo Merchant was driven south-
east and out to sea, passing through fishing areas but not
contacting shore resources to any appreciable extent. A
risk forecasting mode) predicted this to be the most likely
path for this spill, given the location and season and the
particular wind conditions under which the spill occurred.
However, the model's risk forecasts indicated that, had a
spill at this location occurred either under other typical
wind conditions for the season or at a different time of
year, risk would have been much higher.

References

Bumpus, Dean F., 1973, A description of the circulation on the
continental shelf of the east coast of the United States:
Progress in Oceanography, v. 6, p. 111-157.
Grose, P. L., and Mattson, J. S., eds., 1977, The Argo Merchant
oil spill - a preliminary scientific report: National Oceanic
and Atmospheric Administration, Environmental Research
Lab., Boulder, Colo. 80302.
Slack, James R., and Smith, Richard A., 1976, An oilspill risk
analysis for the South Atlantic Outer Continental Shelf
lease area: U.S. Geol. Survey Open-File Report 76-653,
54 p.
Slack, James R., Smith, Richard A., and Wyant, Timothy, 1977,
An oilspill risk analysis for the Western Gulf of Alaska
(Kodiak Island) Outer Continental Shelf lease area: U.S.
Geol. Survey Open-File Report 77-212, 57 p.
Smith, Richard A., Slack, James R., and Davis, Robert K., 1976a,
An oilspill risk analysis for the mid-Atlantic Outer Conti-
nental Shelf lease area: U.S. Geol. Survey Open-File
Report 76-451, 24 p.
Smith, Richard A., Slack, James R., and Davis, Robert K., 1976b,
An oilspill risk analysis for the North Atlantic Outer
Continental Shelf lease area: U.S. Geol. Survey Open-File
Report 76-620, 50 p.
Stolzenbach, K. D., Mattson, J. S., Adams, E. A., Pollack, A. M.,
and Cooper, C. K., 1977, A review and evaluation of basic
techniques for predicting the behavior of surface oil slicks:
Ralph M. Parsons Lab. Report 222.

Near-Bottom Transport in the Vicinity of
the Argo Merchant A Seabed Drifter
Study

Barclay P. Collins, Clement A. Griscom, and Eva J. Hoffman

Graduate School of Oceanography
University of Rhode Island
Kingston, Rhode Island

Abstract evidence that the oil would not sink. Several processes
Seabed drifters were released from early January which may cause oil to sink include: 1) loss of soluble low
through March, 1977, by helicopter and ship in an area molecular weight components by weathering, 2) adsorp-
bounded by latitude 400N. and 420N. and longitude 670W. tion of oil onto sediment increasing the density, and
and 72.5*W. to determine the possible transport of Argo 3) direct physical mixing of the oil into the sediments by
Merchant oi I by near-bottom currents. A total of 176 of the turbulence. It is now estimated that less than one percent
1800 drifters released were recovered as of December 1 of theArgo Merchant oil reached the bottom (Hoffman and
1977. Of these, only 38 are known to have been recovere@ Quinn, 1978). If Argo Merchant oil had sunk to the bottom,
intact. Ship-launched drifters released at University of it could have drifted to the northwest based on
Rhode Island (URI) stations 8 and 9, located 8 nautical miles seabed drifter studies of near-bottom circulation on the
south southwest of the wrecksite and at the Argo continental shelf showing that from the Great South
Merchant site, respectively, had the largest number of Channel to south of Rhode Island, the predominant near-
intact drifter returns. Based on a 12 percent return from bottom drift direction is to the northwest at a rate of
these stations there appears to be a near-bottom drift 0.7 � 0.2 nautical miles/day with a transport tendency
component to the northwest. These results corroborate toward the mouth of Narragansett Bay (Bumpus, 1973).
the findings of Bumpus (1973) in this area. Drift rates of Drifter returns of earlier releases in the vicinity of
0.3 - 0.6 nautical miles/day are similar to those reported theArgo Merchant are about 10 percent (Bumpus, 1973),
by Bumpus (1973). Only three drifters released east of suggesting a substantial offshore drift component
longitude 690W. were recovered suggesting a diver- (Bumpus, 1976). To the west of Nantucket Shoals the
gence of near-bottom drift in this area with an offshore returns are substantially greater. In general, Bumpus
component to the east of 690W. (1973) notes a divergence of near-bottom flow, with an
inshore drift component landward of the 50 meter contour
and an offshore tendency seaward of this line.
Background and Previous Studies
:TATIONS:
An immediate concern after the grounding of the 110- 1-77-1/W7
42. _*MFS12/2 77-315/77
.URI
Argo Merchant was that the No. 6 fuel oil carried by the 216,77 IMM
tanker would sink to the bottom, contaminate the
sediments, and be transported towards Massachusetts A
and Rhode Island shores. The specific gravity of heavy
residual fuels, such as that carried by the Argo Merchant,
41.
may increase as a result of evaporation of volatile com-
ponents, allowing the oil to sink to the bottom (Grose and
Mattson, 1977; p. 66). Determinations of the physical SEABED DRIFTERS
oroperties of the cargo oil from the Argo Merchant by RELEASE LOCATION
J. H. Milgram (Grose and Mattson, 1977; p. 71) revealed
the specific gravity of the No. 6 oil to be 0.96, not the value _L
72. TO- 66.
of 0.996 reported by the U.S. Coast Guard. This early
finding, indicating the oil would not sink, was based on Figure 1. Seabed drifter release locations in the vicinity of the
distillation results, and was not considered conclusive Argo Merchant.

II I I I I .I I I _F_
JAN-FEB RETURNS SEPT-NOV RETURNS 0
STATIONS: STATIONS@
42- ONMFS 1/4/77-1/9/77 42' ONMFS 1/4/77-1/9/77 0
&URI 1/6/77, 1/19/77 &URI 1/6/77, 1/19/77
AURI 2/26/77 &URI 2/26/77
010 00 0 10 0
NAUT. MILES coo L___j NAUT. MILES
CA "i,

MARTHA's MARTHA%
VINEYARD VINEYARD

0
@0
",-A! 0 0 0

41* 41*

72' 71* 70- 69. 721 71* 70. B9.
Figure 2. January through February, 19 77, seabed drifter returns Figure 5. September through November, 1977, seabed drifter
in the vicinity of the Argo Merchant Solid lines show inferred returns.
travel paths of all drifters that were recovered broken or whose
condition is unknown. Dashed lines show inferred travel paths
of all drifters known to have been recovered intact. The University of Rhode Island and the National
Marine Fisheries Service (NMFS) of Woods Hole
deployed approximately 1800 Woodhead type seabed
MARCH-IMAY RETUIRNSI I I drifters (Lee et al., 1965) over Nantucket Shoals, Georges
STATIONS: Bank, and Rhode Island Sound to determine the possible
4:2- - ONMFS 1/4/77- 1/9/77 transport direction of any oi I from the Argo Merchant that
AURI 1/6/77, 1/19/77 sank to the bottom (Figure 1). A total of 1150 drifters were
AURI 2/26/77
0 to 0 released by helicopter and ship at 9 URI stations and 535
L@ NAUT. MILES released during two NMFS cruises. Bundles of 75, 100, or
0 150 drifters were deployed at the URI stations and groups
of 5 at each NMFS station. Clusters of drifters were
weighted for rapid descent to the seafloor where a water
NANTUCKETI, soluble link dissolved in about ten minutes, releasing the
drifters to respond to near-bottom currents.
-------------- 0 0
41* Results and Discussion
During the period from January 6, 1977, to
December 1, 1977, approximately 10 percent of the
drifters were recovered. Of the total of 176 returned, 84
72* 71- 70- 69* are known to have lost their stems and weights. When this
Figure 3. March through May, 1977, seabed drifter returns. occurs the drifter loses its negative buoyancy and
becomes a surface drifter. The plastic drifters, purchased
from Insul-Tab in Woburn, Massachusetts, may have
become brittle in the cold waters and broken apart as they
JUNE-AUGUST RETURNS moved through the surf zone. Follow-up letters confirm
STATIONS: that only.38 of the drifters were found intact; the condition
42'- ONMFS 1/4/77- 1/9/77 of the remaining 54 is not known.
&UR 1/6/77. 1/19/ 77
&UR: 2/26/77 A seasonal analysis of the recovery data (Figures
0 10 0 2 - 5) shows a 5 percent return in January and February
NAUT, MILES
C' increasing to 25 percent in March through May. The return
maximum of 44 percent in June through August (Figure 4)
falls off rapidly to only 14 percent in September, October,
and November. The three-month recovery plots show that
NANTUCKETI. although the majority of the drifters were recovered
broken, the predominant direction of drifter movement
was the same as the intact seabed drifters. During
41' December, January, and February the wind direction is
predominantly from the northwest (Grose and Mattson,
1977; Morgan and Anthony, 1977). Therefore, if the
M

0
A
A
A

N
U
U
R

M
R
R
C

0
SS_
2
T
A

6
M

2
to
A
T

6
y
'o
7
7
7
NA RE TU' RNS
S* I
SEP T-NOVRETURN
STAT'ONS'
0NM FS1/4/77_I/g/
U R I/6,77,'/ 9/
. R 2 26 77

0 to
UTMI'ES
S
I

77
0 '0
N,

.ART-A%
V VIAR.

@N
71/9 /77
/'9/ 77
77

UT ILES

NAAT

A drifters had broken apart at some time before stranding,
L I an offshore direction of travel would be expected.
7i- 71. To- 69. Although the quality of the data is poor, there appears to
Figure 4. June through August, 1977, seabed drifter returns. be a change from a predominantly northeastward drifter

movement during March through May to a strong north- Bumpus, D. F., 1976, Review of the physical oceanography of
westward drift in June, July, and August. Georges Bank: International Commission for the North-
Of the total released, 150 drifters were deployed on west Atlantic Fisheries Research Bulletin No. 12, 119 -
February 26, 1977, at URI stations 8 and 9, located 8 134.
nautical miles south southwest of the wreck and at the Grose, P. L. and J. S. Mattson, eds., 1977, The Argo Merchant oil
spill, a preliminary scientific report: National Oceanic and
Argo Merchant site, respectively (Figure 1). These were Atmospheric Administration Special Report.
ship-launched and have the largest number of intact Hoffman, E. J. and J. G. Quinn, 1978, A comparison of Argo
drifter returns, 13 complete of the 18 returned. Based on a Merchant oil and sediment hydrocarbons from Nantucket
12 percent return from these stations there appears to be Shoals: Presented at the Argo Merchant Symposium,
a component of near-bottom drift to the northwest. These Center for Ocean Management Studies, University of
results corroborate the findings of Burnpus (1973) in this Rhode Island. -
area. The first drifter recoveries from the two stations did Lee, A. J., D. F. Bumpus, and L. M. Lauzier, 1965, The seabed
not occur until mid-June with the majority recovered in drifter: International Commission for the Northwest
July. Drift rates of 0.3 - 0.6 nautical miles/day calculated Atlantic Fisheries Research Bulletin No. 2, 42 - 47.
for stations 8 and 9 are similar to the rate of 0.7 � 0.2 Morgan, C. W. and W. H. Anthony, 1977, Average monthly wind
nautical miles/day found by Bumpus (1973) from Great stress along coastal regions of the United States and
western Canada: U.S. Coast Guard Oceanographic Unit
South Channel to south of Rhode Island. Technical Report 77-1, 29 p.
Only three drifters released east of longitude 69W
were recovered suggesting a divergence of residual drift
in this area. Bumpus (1976) notes an offshore drift on
Georges Bank based on low drifter returns from this
region. Great South Channel, approximately delineated
by the 690W. longitude line, may convey many of the
drifters from both Nantucket Shoals and the western
portion of Georges Bank offshore.

Conclusions
Residual bottom drift in the area of the Argo
Merchant appears to be primarily offshore with a small
onshore component of drift to the northwest. Results
from this study corroborate Bumpus' (1965, 1973, 1976)
findings of a minor northwest drift component trans-
porting drifters at a rate of less than one nautical mile/day.
Higher recovery rates and drift speeds are observed for
drifters released closer to shore. There appears to be a
divergence of residual bottom drift associated with the
Great South Channel.
Although little Argo Merchant oil appears to have
sunk to the bottom, heavy residual fuels such as the fuel
oil carried by the Argo Merchant have been known to sink
as the lighter volatile components evaporate. Seabed
drifter studies appear to be a valuable technique to
determine possible.transport routes of oil spills that reach
bottom. However, the usefulness of the present results
are severely limited because at least half of the recovered
drifters broke apart. The design and quality of materials
used in the construction of seabed drifters should be
evaluated further to insure drifters will remain intact
throughout the study.

Acknowledgments
This study was partially funded by NOAA Contract
03-7-022-35123.

References

Bumpus, D. F., 1965, Residual drift along the bottom on the
continental shelf in the middle Atlantic bight area: Lim-
nology and Oceanography, v. 10 supplement, R50 - R53.
Bumpus, D. F., 1973, A description of the circulation on the
continental shelf of the east coast of the United States:
Progress in Oceanography, v. 6, 111 - 157.

Surface and Subsurface Spill Trajectory
Forecasting: Application to the Argo
Merchant Spill

Malcolm L. Spaulding

Ocean Engineering Department
University of Rhode Island
Kingston, Rhode Island

Abstract was further assumed that the wind induced surface drift
In an effort to meet the needs of the coastal zone was in the same direction as the wind.
management community and the University of Rhode The model was run using two-hour time steps, i.e.,
Island oil spill response team in planning for oil spill the spill was moved at the determined rate and in the
response, trajectories were forecast for. the Argo
Merchant spill. Using a simple surface drift model, the spill
trajectory was estimated for periods up to 30 days. Actual 5-DAY@ MONTE CARLO
spill motion closely followed the predicted mean trajec-
tories. Based on the seabed drifter data of Bumpus, a
simple advective model was constructed to estimate
possible subsurface oil transport trajectories. The results
of the model agree qualitatively with seabed drifter
studies performed during the spill.

Introduction
Shortly after the Argo Merchant ran aground in the
Nantucket Shoals area, requests were made to the Ocean
Engineering Department by the Rhode Island coastal (D
W
management community, the University of Rhode Islana e
Oil Spill Response team, and the governor of Rhode W
Island's energy advisor to provide probable oil spill 0
D
trajectories. This information was needed in order for the
Rhode Island community to develop an appropriate oil
CC
spill contingency plan. _J
To meet this need, two simple computer m+odels
were constructed
one to predict the surface drift of
spilled oil, while the second predicted the subsurface
drift patterns. The results of this effort are presented here.

Surface Drift Model

In the simple surface drift model that was
developed, it was assumed that the wind induced a sur- 70.50 69 00 67 so 66 00 64. so
face drift of 3.5% of the wind velocity. This drift is LONGITUDE (DEGREES)
accounted for by 1.5% for the wind induced water motion Figure 1. Monte Carlo surface drift predictions (wind & tidal
and 2.0% for the relative oil/Water motion (Smith, 1974). It currents) for 5, 10, 15, 20, 25, 30 days after the spill.

10-DAY@ MONTE CARLO 20-DAYi MONTE CARLO

W 8.
Cr
CD CD
UJ W
g

F- d;
a:
_J _J

70.50 69 Do 6?. 50 66.00 64.50 to. so 69.00 STAS $6.00 14.50
LONGITUOE (DEGREES) LONGITUDE (DEGREES)

15@DAYq MONTE CARLO 25-DAY9 MONTE CARLO

C U)
W ca
W W &
W
cc 0
W W
(30 00
C! C!

W W C ff
0

F-
Cr cr
_J _J

0
c! c!

C,
o

TO.50 69 00 C.T.50 66.00 64.50 T0.50 63 go or so 66:00 $4.50
LONGITUDE (DEGREES) LONGITUDE (DEGREES)

determined direction fora period of time corresponding to from the National Oceanic and Atmospheric Administra-
two hours before the wind was changed. The necessary tion Navigation Chart 13006. All cases were run starting
wind data was obtained from the Grant Point Coast Guard on December 18 and calculated 30-day trajectories.
Station on Nantucket Island. Tidal currents were taken Monte Carlo runs were made both for wind-driven

w 00-0 a) z-r -Z 0- 0 0 0 (P=3 0-=C13 (D 0 C/)< 0 CL - - - 0 LATITUDE (DEGREE
- U - w 0 03 0 CL D c c 0 3 C) (D CL M W W 0' M M CT 7,' ::r D- :7 C
3 (a ch M o CD - - - " CD CD CD 36.00 37.00 38.00 39.00 40.
z w a 0 - - - - CD < Q W =(O 0 W En CD
0 Q M 0 CD -0- CD CD M 0 - CD W @T 0 M ::3' :Z:- CD 0-0 CD
0 2 CD --o 0-0 -- ::r
:3 cn = 0- =rac: - 0 o o :3
=r 0 W 3 6 (n C: 0 M --j CD 0 0 0 Z)
w :3 (D M 0 C) CD - - C -
- Cn W 2. Z3 0 0 co 3 cn 0 0.0 w M - ZT 0. CT
0- Cl 0- CD 0 U(C - 0 0 :T :3 - CD CD aD a
0 CD M 0 co CD D) M W
a- CL Z3
Z) PL =3 3 2. w a N) 0 zy - 0 -1 = 0 0-
CD CD 3 W
U) CD 0) CD 0 '0 --0 CT M 0 ::r CD
C: 0 Cn CL Cl 0 -- Z) - CD
0 CL CL Z = W cr Z 3 0 0
M D
W CD wo 0 c
CD - 0 CD - CD C (n CD =) Cn CD CD W - Cn
=r Cn :r :3 - 0 CL * " a a CD w
.0 0 :3 CD 0 (p C-D U) 0 =3 0 CL CL ::r 0 C M
om 0 W 0 m ==ra o x, 0 0 ::r < w -0 CD Cr CL 0-.
-:E =r :3 0) o cl CD -o
0 Ch 0 Cn =0-0 W -- CD (n
w CL - CL Cn ' -- =3 - Cn :3 CD =1 0 0. W Cr!R, cn a- co -0 w CO CL Z
Z) iw 0. CL G)
C CD cD o 0 Ch CD CL C CD -4 0 -
9L @@ , CL 0 0 3 w - CL--w -5 =;' Q. 0 CD V 0
w r - - CD m - w m m C- CD <
G) Cr cn - CD C =3 c W OL O'D a-W 9-oz-=i@ , @ ::r M c
--Q- -- CL CD 0 0 0--. cn CC CD CD :3
" CL c
0 3 m - CL CD - " - -- :3 cD (D w @ :E 0 rn
w c - (D :3 5. CL CD W -- C 0 Cl (D w 3 :L cl 0
- :3 cn 0 0- CD =!* Z3 Q W -- 0 W W 0) Z) CD CL
0-0 CL -a o (o m - 0 - cD w 0 5 (n w 3 Z (a :3 5'-
W WD a m -- 0 Pcl-=CD 0@ CD =ro-o -;z* a
= CD - - - 0 CL - 00
w --3 CD -- CD 0 W = M - - M 0 :r CD W :y CD 0 CD rn
:
5- 3 =CL -
3, W
CD 0) m (CD =03 --n - P a-- C<D @-E- 0 0 -0 5; 0 :3 3 -o 7-3
c 0 a 3 -(0 0 0 :3 =3 =r
cn =r 0 =r -4
@r o 0 --J--CLO 00 < CD Cn m
W< c- 5- 3 0 = CD N (n CO - a 12 CD W
CD 0 D) M Cl 0) 3 0 :3 zy cn 0 CD x C: C rp CL CL m
=) U) W C-) CD CD cn @ 0 CL W
=3 * -a CTD a
CL :5- C) :3 > :3 M
a (D -0 Z(c o 3-5
0) Cl W CD a) 0 w CD 0 0 CD 0) =r . 0 Cy' 5'= 3 W NIn
(n CD -W - =r " -
:� =3 * 0 =
'<M 0 Cr Z3 CD CD CD CD 0 CD 0 m C-L CMD Z w - o w ZF CL C<L
CL o F o 0- 0 - a 0 'a :3 w - @ CL =r :3 co 0- CD
w 0 5' =-C o OL :3 M OL M
cy 3 = =3 W OL _ - M 0 CL a 0 CD CD 0
CD < -0 =r < rr M 0 CD (A M =
=CD CD CD 0 ZT CL CD C, M 5' :3 3 .0 C:
c (1 0 CL 0 0) F). c o 0 Sm
8 =r CD (n w CD - OL M CL Z3 (D 0 M (n D a 0 = :2.
CD C CD a W CD E 0- E CD :3 W CD =r 0 W
=r CL - < 0- CD ::r c a 0 cn (A 5 - :: cn
CD Cn cn 00 zT w 0,
w .-o CD =r CD 67 @L Elr CL - 0 w
0 - - - =CD CO CD w :3 ::r F)* CD M =r M M 0 ID CD - :3 0) 0 -
- ao N) CD - 9 CD P @ CO CD a - w CD = :3 CL CD 0- a :3 CD a - - 0

CA)

LATITUDE (DEGREES) LATITUDE (DEGREE
36.00 V. 00 38.00 33.00 40.00 41.00 42.00 39.00 V. 00 39.00 33.00 40.

CD
N)

CA) Zt.
Cl CO)
a.
Z.
G-) Ch G)
0) fb" -
zr C) --4
(b C
M- z E; S M.
0 C5

m
G) m

M-
M.
M M!-

m Do).
z
-4

cn :E C) LATITUDE (DEGREES) LATITUDE.WEGRM
0CD C 36.00 37.00 38.00 33.00 40. OD 41.00 4e. DO 3G. 00 37.00 38.00 39.00 40.0
F@ 0
C::rM :3
D En
W CD
En CD

C
C:
D
(o rj) F -
CL --
0
:3Z
(D -0
3 - Ej- r
CD
a CLW 0
0 0- - z . - Z
0 3.w * 0 -- G)
m CD CD
D) co w
:7 0
(D fD CD M z M -
74
0 U)
CD m
0 =r m
=,< M2 G) z
5. ;o
Q FD':E m M
CD CD m
cn :@ V);., in
C-D
CD Co CL
CD 0- m
2. CD z
CD CL W
o = 6
0CD 0 -
M 0) -1-
CD E-r @'
W CD

co- A) 0 LATITUDE (DEGREES) LATITUDE.(DEGREEl
o =r o w
IW -- 36.00 3?. 00 38.00 33.00 40.00 41.00 42.00 36.00 3?. 00 39.00 33.00 40.0
C0
w
W Cn
= CD
0 C)-
cr C=)
CD <
=r CD 0
0 C)
=)- r
w 01. 0-.
CD -CD, CY z
Z o
0 G)
CD -4
- CD CD C:
M 0 z M
Cl) M
0 0
x 0 -0
(0)0 57 m
nm Zy m
C2 = CD G)
c=3 0 m
3r- K) =r M, M .
M. mr
MM C.n CD U) n cn .
CA) - -
C:
!zt
CD
ZF m
LD 0
0 CD =3
7
CD 5-0
w P@ --
0) > =r
CD CD

Table 1. Summary of residual bottom drift literature for southern New England shelf

Observation Residual
Type of Bottom
Reference Location L ength Observation Drift (cm/sec) Drift Direction Comments

Beardsley & 40o45'N March 8- Richardson 6.2 Westward Two Northeast Storms
Butman 7 1'03'W April 10, Type (Along Shore) (Lasting a Total of 4
(1974) 18 m 1974 Current Days) Produced 2/3 of Net
from Meter Westward Displacement
bottom Observed During the
(60m Experiment
depth)

Bumpus Cape 1961- Sea Bed 1,4� A Offshore of a Line Drift Toward Mouth of
(1973); Cod 1970 Drifters Great South Drawn about 1/2 to 3/4 Estuaries, Particularly
a I so see to Channel to of Distance Between the Evident South of Rhode
Saila Cape South of Shore and the 50 Fathom Island
(1973) Henlopen R.I. Contour at Depths of 30-
.2-1.4 35 Fathoms, the Ten-
but more dency is Toward an Off-
frequently shore Drift. Inside
.4-1.0 this Line, the Tendency
is for the Flow to
be Westerly or Southerly
with a Component Toward
the Coast

east heading. This marked change in direction by the slick possibility for oil sinking, either by itself or attached to
from the wind induced drift could be considerably im- sediment particles, and being advected along with the
proved by addition of Gulf Stream generated currents. bottom currents was considered.
A simple advective transport model was developed
Subsurface Drift Model to predict probable trajectories if the oil were to be
transported by bottom currents. In an effort to provide
Since all the preliminary information we could obtain estimates of bottom drift direction and speeds, a quick
at the time indicated that Argo Merchant oil had a high review of the literature was undertaken and revealed
specific gravity, .96 (Grose and Mattson, 1977), and it several investigations on seabed drifter studies and
appearea tnat tne suspended sediment concentrations current meter measurements on the southern New
would be extremely high due to the high wind and wave England shelf. These studies are summarized in Table 1.
environment and shallow depths of the spill site, the Using the inferred bottom drift directions and

PARTICLE STARTING
DECEMBER LOCATION

ARGO MERCHANT

72 7 t 70 69

Figure 3. Inferred bottom drift for December, 1961-1970 Figure 4. Bottom drift trajectories basedon Bumpus (1973) sea-
(Bumpus, 19 73). bed drifter data.

PARTICLE S' '@A,@TING Computer facilities were furnished by the URI Academic
LOCATION Computer Center.

References

Beardsly, R. and B. Butman, 1974, "Circulation in the New
England Continental Shelf: Response to Strong Winter
Storms", Geophysical Research Letters, Vol. 1, No. 4.
kZZ6 ARGO MERCHANT Bumpus, D. F., 1973, "A Description of the Circulation on the
0 Continental Shelf of the East Coast of the United States".
In Progress in Oceanography, Vol. 6, Pergamon Press,
New York: pg. 111-157.
0 Collins, B., C. Griscom, and E. Hoffman, 1978, "Near Bottom
Transport in the Vicinity of the Argo Merchant A Seabed
Drifter Study", In the Wake of the Argo Merchant Sym-
posium, Center for Ocean Management Studies, Univer-
sity of Rhode Island.
Draft Environmental Statement, 1976, U.S. Department of
Interior, Outer Continental Shelf Oil and Gas Lease Sale,
72 71 70 69 OCS Sale No. 42.
Grose, P. and J. Mattson (eds.), 1977, "The Argo Merchant Oil
Figure S. Bottom drift trajectories based on Bumpus (1973) sea- Spill", National Oceanic and Atmospheric Administration,
bed drifter data. Environmental Research Laboratories.
National Ocean Survey Chart #13006, 19th Edition, April 1976,
National Oceanic and Atmospheric Administration.
speeds for December based on Bumpus'(1 973) review of Noll, C., P. Comillon, and M. Spaulding. 1977, "URI-ERDA Oil
the 1961-1970 seabed drifter data, as shown in Figure 3, Spill Drift Prediction Model - Preliminary Report", Depart-
several bottom drift trajectories were predicted. Simu- ment of Ocean Engineering, University of Rhode Island,
lated oil particles were started from several locations in Kinqston, Rhode Island.
the area of the spill; also, the path of the surface Saila, S. led.), 1973, "Coastal and Offshore Environmental Inven-
slick and trajectories were simulated using a one-daytime tory - Cape Hatteras to Nantucket Shoals", Marine Publica-
step. Advective velocities for the oil particle were deter- tion Series No. 2, Marine Experiment Station, Graduate
mined by taking the closest velocity to the particle at School of Oceanography, University of Rhode Island,
Kingston, Rhode Island.
each time step. Figures 4 and 5 display typical bottom Smith, C. L., 1974, "Determination of the Leeway of Oil Slicks",
drift trajectories. Department of Transportation, U.S. Coast Guard, Report
These figures clearly show a bottom drift toward No. CG-D-60-75.
Rhode Island's coastal waters. Employing the several U.S. Naval Weather Service Command, 1970, "Summary of
estimates for bottom drift speed, noted in Table 1, Synoptic Meteorological Observations (SSMO) for North
estimates for time of arrival of oil particles at shore would American Coastal Marine Areas", Vol. 11, National
be 30-150 days after entrainment in the water column. Technical Information Service, AD707699.
Comparison of these trajectory estimates with the sea-
bed drifter study performed by Collins, Griscom, and
Hoffman (1978) in the area of the spill show good
agreement.

Summary
The rapid development and application of these
simple computer models to predict both surface and sub-
surface drift of the Argo Merchant oil spill proved to be
very beneficial to the Rhode Island management com-
munity in planning spill response efforts as well as to the
URI Oil Spill Response team in planning the initial
oceanographic cruises. While the surface drift predic-
tions were adequate for planning purposes, much work
remains to be done in order to predict the fate of oil
discharged into the marine environment.

Acknowledgments
Financial support for this effort was provided by
ERDA Contract DY-76-S-02-4047. Valuable contributions
in developing and coding the computer programs and
data were provided under severe time constraints by
@410. @_
@ARGO MERCHA

Dr. P. Cornillon, C. Noll, and R. Gordon of the Ocean
Engineering Department at the University of Rhode Island.

Oil Droplet Measurements Made in the
Wake of the Argo Merchant

Peter Comillon

Ocean Engineering Department
University of Rhode Island
Kingston, Rhode Island

Abstract relating to the structure of larger oil droplets, 100M to
In an attempt to determine the size of entrained oil 500 p, were obtained, and these along with the reasons
particles following the Argo Merchant spill, one literwater for attempting, the droplet size measurements, the
samples were filtered and the resulting filters examined apparatus and experimental technique used and the
problems encountered are discussed below.
for signs of oil. The technique used was similar to that
used by Forrester following the grounding of the Tanker Why Make the Measurement? The first question which
Arrow In the case of this spill only several fairly large comes to mind is why should one want to make such a
(> 100 1a) oil particles were detected in all of the samples measurement (oil droplet size distribution) rather than
examined. The scarcity of observed oil is most likely a re- simply relying on the values of concentration obtained
sult of the high rate of dispersion of entrained droplets, if through chemical analysis which are more easily made
any, the observational difficulty resulting from large sedi- and more accurate? The main reason is that the shape and
ment and biological concentrations in the water and the mean value of the oil droplet size distribution define a set
fact that the bulk of the samples we 're obtained approxi- of variables which are extremely important in determining
mately seven weeks aftertheArgo Merchant broke up and the fate of entrained oil. Some of the more important
long after the surface spill had left the area. physical and biological processes affected by this
It was possib le, however, to obtain interesting distribution are:
results with regard to the structure of 100 p and greater 1. Residence time in the water column. This refers to the
droplets specifically whether or not they contained length of time for which a droplet will remain submerged.
sediments. Three general droplet structures were This time determines the distance that the droplet is
identified: advected as well as the distance that it diffuses which,
1. Pure oil with no sediments in or adhering to the for a cloud of entrained droplets, defines the areal extent
droplet; of the cloud and its center of mass as a function of time.
2. A large oil droplet, between 100p and 500p, with The residence time also affects the probability that a
small particles of sediment inside; and, droplet finds its way to the bottom.
3. A large piece of sediment, between 10011 and 500 p 2. Rate of biological, consumption. Current theory indi-
coated with oil. cates that the rate of microbiological consumption
Those samples with droplets falling in class 1 were taken depends critically on the oil-water interfacial area rather
from the surface (bucket samples) or one meter below the than simply on the concentration of oil (Traxler and
surface (Nisken bottle). Those samples'with droplets Bhattacharya, 1977).
falling in classes 2 and 3 were taken at 6 meters and a 3. Suspended sedimenVoil droplet interaction. The size
meter above the bottom. of an oil droplet will most likely aff ect the probability of its
adhering to or absorbing suspended sediments. The size
of the droplet will also determine whether or not it remains
introduction buoyant when and if it sticks to a given piece of sediment.
The objective of the experimental work described A decrease in buoyancy may lead to the sinking of the
herein was to measure the droplet size distribution of droplet or to a substantially longer residence time in the
entrained oil in the 1 p to 1 00,u range under actual spill water column..
conditions. Unfortunately this effort was unsuccessful Those of us at the University of Rhode Island
due to the large number of microorganisms and the high working on the use of chemical dispersants with oil soills
sediment load in the water at the time. Interesting results are especially interested in the droplet size distribution

because the primary functions of a dispersant are to FUNNEL
promote entrainment and to increase the stability of the
oi I-in-water dispersion, therefore enhancing the impact of SEPARATORY FUNNEL
the above processes on the fate of spilled oil.
How May the Measurement Be Made? Several ap-
proaches have been taken in the past to measure the oil
droplet size distribution in the water column. These are
briefly outlined below. STOPCOCK
a. Use of a Coulter Counter. The Coulter Counter is an
electronic instrument which measures the oil droplet size FILTER MILLIPORE VACUUM
IN-LINE TUBING
distribution directly. This is done by forcing the oil-in- FILTER
water dispersion through a narrow orifice and inferring the
size of each oil droplet from the change in capacitance of
the water as that particular droplet moves through the
orifice. The technique suffers from two major disadvan- RESERVOIR TO
tages for actual oil spill work. FOR VACUUM
FILTERED PUMP
i. It cannot distinguish between oil droplets and micro- UATER
organisms. TRAP
ii. The flow rate is low because of the narrow orifice, thus
making the sampling of large volumes of sea water an
extremely lengthy process. Figure 1. Experimental apparatus.
In open ocean spill work the number of oil droplets per
liter is so low (<100/liter) that the two problems men-
tioned above render the technique impractical. Coulter below was finalized. This set-up was used on the last two
Counters have, however, been used quite effectively in cruises of the R/V Endeavor to the site of the Argo
the laboratory. Merchant, cruises EN-004 and EN-005. On the first two
b. Counting the Oil Droplets in a Drop of Water Using a cruises EN-002 and EN-003 the concept was the same
Microscope. To determine the droplet size distribution but the experimental set-up was somewhat different.
using this method (Jasper et al., 1977) a drop (@_-.025 ml) of A schematic of the filtering apparatus is shown in
the oil-in-water dispersion is examined under a microscope. Figure 1. Although this figure shows only one in-line
The oil droplets in the dispersion are measured and count- filter with the associated glassware, there were actually
ed. This method is quite straight-forward but the analysis three such assemblies each leading to the same vacuum
time for a small volume (a few milliliters) of the dispersion trap. In this way it was possible to filter three different
is extremely long. Furthermore sediment particles often samples simultaneously. This approach was selected
resemble oil droplets. For small numbers of droplets per because the data collection procedure used generally
liter (<105) these two problems make this method provided three water samples, the first taken approxi-
impractical. mately one meter below the surface, the second taken
c. Filter the Oil-in- Water Dispersion and Count Droplets on approximately six meters below the surface and the third
the Filter. This technique was first used by Forrester taken approximately one meter above the bottom. From
(1971) for the Arrow spill. The procedure involves 500 to 1000 milliliters of sea water were filtered through
filtering the oil-in-water dispersion through a millipore the millipore filters with the pore size ranging from .8 ju to
filter and then examining the filter under a microscope. A 8
fairly large volume of the dispersion, one liter say, may be
filtered in several minutes and then because the area of
the filter is small (<10 CM2) the entire filter may be Results
scanned rapidly for oil droplets. In this way even a very To begin with, the objective of the experiment, to
small number of oil droplets per liter is readily detectable. measure the oil droplet size distribution in the 1 p to
One disadvantage with this approach is that the area of a 100p range, was not met. This was primarily due to the
droplet on the filter must be correlated with the diameter fact that at most locations the large sediment load and the
of a droplet in suspension. Another problem is that small large number of microorganisms made it impossible to
oil droplets (< 100 M) are very difficult to distinguish in determine whether or not there were any oil droplets less
water containing nigh concentrations of suspended than 100 p on the filter. Oil droplets larger than 100 p
sediments and/or microorganisms. were detectable however and the distribution of samples
with such droplets is discussed below.
Methods RIV Endeavor Cruise EN-002. On this cruise, due to the bad
weather only several samples were taken. Only one oil
Experimental Apparatus. For this project the filtering droplet was detected, this in a sample taken approxi-
technique described above was selected. The main mateiy 30 kilometers east-southeast of the Argo
reason being that the results of Forrester indicate a very Merchant wreck. This droplet was irregular in shape and
low concentration of suspended oil droplets for a spill measured approximately 100 p by 300 M. It was dis-
quite similar in nature to theArgo Merchant, thus ruling out solved with carbon tetrachloride and showed no sedi-
the other two methods. ment inside of it. A photograph of the droplet is shown in
Several iterations were performed in the actual Figure 2 as seen through a microscope under the
experimental set-up before the apparatus described magnification X100. The results of the analysis of the

filters for this cruise are summarized in Table 1. The
locations of the various stations as a function of depth are
indicated in Figure 3 where + indicates a sample with no
oil and 0 indicates a sample with at least one oil droplet.
The location of the wrecked Argo Merchant is indicated
by a A.
RIV Endeavor Cruise EN-003. The samples from this
cruise showed no oil in a preliminary scan. The filters
were considered to be too damaged to carry out a
detailed analysis.
RIV Endeavor Cruise EN-004. A significant number of
samples from this cruise were analyzed in detail and no
indication of oil was found on any of the filters. Table 2
summarizes those analyzed and Figure 4 shows the
.......... ......... sampling locations.
Figure 2. Oil droplet on filter paper as seen under microscope RIV Endeavor Cruise EN-005. Oil droplets exceeding
magnification X100. 100 pt were detected at several locations from this cruise.
These locations are shown in Figure 5. The location of the
bow and stern sections of the Argo Merchant are marked
in the figure with a A and a "b" and an "s", respectively.
The sizes of the O's designating locations where oil
droplets were found have been made proportional to the
square of.the number of droplets observed, th is to give the
reader a feeling for the density of droplets. There are
several points worth stressing with regard to the analysis
of the filters from this cruise (see Table 3).
4j a 1. Oil droplets were found at all three levels for which
4 samples were taken: surface, mid-depth (generally below
the surface 6m) and 1 meter above the bottom.
7B 19 59 138 68 11 W W I sa I aas 139 2. The oil droplets observed may be divided into three
ALL SAMPLES SURFACE SAMPLES groups determined by their structure.
a. Oil with no sediment in or adhering to the droplet.
b. Oil droplet with sediment particles inside.
c. Sediment coated with oil.
3. The droplets in the three classes listed above had a
characteristic length in the 1 bo u to 500 p range.
4. The concentration of droplets (number of droplets per
liter) is order-of -magnitude similar to that observed by
41 al 4t a Forrester; mean separation of 10cm to 20cm or about 1
droplet per liter.
5. The trail of oil droplets observed for this cruise, i.e. to
J. I- as L a the southwest from the wreck, is similar to that observed
MID-DEPTH SAMPLES BOTTOM SAMPLES for oil in the bottom sediments (Hoff man and Quinn, 1978).
LOCATION OF FILTERS WITH (0) AND WITHOUT (@) OIL DROPLETS Small oil droplets were observed to leave the sediment
R/V ENDEAVOR CRUISE EN-002 12/28/76 - 12/30/76 samples when stirred. One might conclude from this and
Figure 3. Location of filters with (0) and without (+) oil droplets. the fact that some of the oil droplets contained sediment
that the droplets observed in the filtered samples are
being released from the sediments.

Table 1. Millipore Filter Results from R/V Endeavor Cruise EN-002

Station Depth Sampling Volume #of Concen tra tion Oil in
Meters Technique LatitudelLongitude Sampled Drops PP8 Sediments
ML

1 1 Nisken 40'41.5'6959.4' 850 0 0 ---
1 6 Nisken 4041.5' 6959.4' 1100 0 0
1 39 Nisken 4041.5'69*59.4' 800 0 0
2 0 Bucket 40'50.1'69'19.4' 1000 0 0
2 6 Nisken 40'50.1'69'19.4' 1100 0 0
3 0 Bucket 4051.0' 68048.3' 1500 1 3.1 No

*The concentration is that resulting from the oil droplets only.

Table 2. Millipore Filter Results from R/V Endeavor Cruise EN-004

Station Depth Sampling Volume of Concentration Oil in
Meters Technique LatitudelLongitude Sampled Drops PP8 Sediments
ML

A34 1 Nisken 41' 2.5' 69'17.5' 750 0
A34 6 N isken 41' 2.5' 69*17.5' 750 0
A34 BOT Nisken 41o 2.5' 69*17.5' 750 0
G41 1 Nisken 41* 6.5' 6922.3' 750 0
G41 6 Nisken 41' 6.5' 6922.3' 750 0
G41 BOT Nisken 41* 6.5' 69o22.3' 750 0
G42 1 Nisken 40-54.8' 69*36.1' 750 0
G42 6 Nisken 40*54.8'6936.1' 750 0
G42 BOT Nisken 40'54.8'6936.1' 750 0
D36 I Nisken 41' 0.9' 6931.9' 750 0
036 BOT Nisken 41' 0.9' 6931.9' 750 0
C39 1 Nisken 41* 3.3'6926.2' 750 0
C39 6 Nisken 41' 3.3' 6926.2' 750 0
C39 BOT Nisken 41* 3.3' 69*26.2' 750 0
815 1 Nisken 40'39.5'6822.0' 750 0
B15 6 Nisken 40*39.5' 6822.0' 750 0
B11 1 Nisken 40'46.8'6822.0' 750 0
B10 6 Nisken 4051.6' 69'20.7' 750 0
A40 6 Nisken 40o56.8' 6931.5' 750 0
A33 BOT Nisken 41* 2.4' 69*13.9' 750 0
E30 BOT Nisken 40'50.2' 69 *15.9' 750 0

*The concentration is that resulting from the oil droplets only.

4 dl
t 4 41 S

4
4
4

AL_3L_ -

ae 4@ ae .36 W 20 aQ 09 .39 ap .28
ALL SAMPLES SURFACE SAMPLES ALL SAMPLES SURFACE SAMPLES

M W MIS wim min MIS N IM so In 09 in 0 1 44 09 In 69 2$
MID-DEPTH SAMPLES BOTTOM SAMPLES MID-DEPTH SAMPLES BOTTOM SAMPLES
LOCATION OF FILTERS WITH (0) AND WITHOUT (+) OIL DROPLETS LOCATION OF FILTERS WITH (0) AND WITHOUT (+) OIL DROPLETS
R/V ENDEAVOR CRUISE EN-004 2/9/77 - 2/13/77 R/V ENDEAVOR CRUISE EN-005 2/22/77 - 2/27/77

Figure 4. Location of filters with (0) and without (+) oil droplets. Figure 5. Location of filters with (0) and without (+) oil droplets.
P.4

Table 3. Millipore Filter Results from R/V Endeavor Cruise EN-005

Station Depth Sampling Volume * of Concen tra tion Oil in
Meters Technique LatitudelLongitude Sampled Drops PPB Sediments
ML

48 0 Bucket 40'56.6'69*41.9' 500 0 0
57 0 Bucket 41* 2.0'69*33.5' 500 0 0
72 0 Bucket 40'58.8'6929.2' 500 0 0
74 0 Bucket 40'59.6' 6931.2' 1000 0 0 - - -
59 1 Nisken 41* 2.5' 6928.0' 500 1 1.6 N o
59 6 Nisken 41' 2.5'6928.0' 500 1 3.8 Yes
59 BOT Nisken 41* 2.5' 6928.0' 500 0 0
75 0 Bucket 41* 2.0' 6929.0' 1000 0 0
70 0 Bucket 41* 1.6' 6926.3' 1000 4 100. No
70 1 Nisken 41o 1.6' 6926.3' 500 0 0
70 6 Nisken 41* 1.6' 6926.3' 500 0 0
70 38 Nisken 41* 1.6' 6926.3' 500 1 ? Yes
58 1 Nisken 41' 0.6' 69o33.8' 500 0 0
58 6 Nisken 41o 0.6' 6933.8' 500 1 4.5 Yes
58 BOT Nisken 41* 0.6' 69*33.8' 500 0 0

*The concentration is that resulting from the oil droplets only.

Conclusions Thanks are also due to the Division of Environmental
Control Technology of the Energy Research aind Develop-
None of the techniques now available are suffi' ment Administration and to the National Oceanic and
ciently sophisticated to detect small, < 100 p, oil drop- Atmospheric Administration for their financial support of
lets under conditions with high concentrations of sus- the project.
pended sediments in the water column and/or a large
number of microorganisms. Furthermore, the volume of
water sampled is too small to obtain useful results on the References
distribution of larger droplets, > 100 u.
It was possible, however, to obtain interesting Traxler, R. W. and 13hattacharya, L. S., "Effect of a Chemical
results with regard to the structure of 100 p and greater Dispersant on Microbial Utilization of Petroleum Hydro-
droplets specifically whether or not they contained sedi- carbons", presented at the ASTM Committee Symposium
on the Use of Dispersants for the Control of Oil Spills.
ments. Three general droplet structures were identified: 1977.
1. Pure oil with no sediments in or adhering to the droplet; Jasper, William L., Kim, Thomas J. and Wilson, Mason P., "Drop
2. A large oil droplet, between 100 g and 500,u, with Size Distributions in a Treated Oil-Water System", pre-
small particles of sediment inside; and, sented at the ASTM Committee Symposium on the Use of
3. A large piece of sediment, between 100 p and 500 Dispersants for the Control of Oil Spills. 1977.
coated with oil. Forrester, W. D., "Distribution of Suspended Oil Particles
Those samples with droplets falling in class 1 were taken Following the Grounding of the Tanker Arrow", Journal of
from the surface (bucket samples) or one meter below the Marine Research, Volume 28-29. 1971.
surface (Nisken bottle). Those samples with droplets Hoffman, Eva J. and Quinn, James G., "A Comparison of Argo
Merchant Oil and Sediment Hydrocarbons from Nantucket
falling in classes 2 and 3 were taken at 6 meters and a Shoals", presented at the Argo Merchant Symposium,
meter above the bottom. Center for Ocean Management Studies, University of
Unfortunately, very little is understood about the oil Rhode Island. 1978.
droplet-sediment interaction. More data of the type
described above are required to determine if the trends
observed here are real or not. If indeed they are real it
may indicate that the droplets from different classes
come from different sources. Specifically class 2 and 3
droplets may come from the sediments having been
trapped there earlier while class 1 droplets may still
have been leaking from the wreck.

Acknowledgments
Thanks go to Dave Konigsberg for constructing the
apparatus and collecting the samples on EN-004, to
Tatsusaburo Isaji for collecting the samples on EN-005,
and to Sergio Antunes for analyzing part of the samples.

Chemical Studies

0
James G. Ouinn, Chairman

Summary of Chemical Studies

Chemical analysis of water samples indicated that needed to determine if these values are due to the Argo
physical or chemical fractionation of the Argo Merchant Merchant spill or normal winter ship traffic in the area.
oil occurred with only the lighter hydrocarbons entering (Data from the Mid-Atlantic region should be reviewed
the water column. These components were probably with this point in mind). The information obtained from
associated with the cutter stock used in blending this additional analyses would be important in evaluating the
particular oil. effects of the Argo Merchant spill as well as providing
Fluorescence spectroscopic analysis of water background data on the Georges Bank region.
samples collected in late December, 1976 from the area In order to provide an adequate response for future
surrounding the grounded vessel indicated concentra- oil spills, the chemical studies workshop endorses the
tions of petroleum as high as 340 ppb. Subsequent pending proposal to form an expanded spilled o .il
analysis of samples from the same area collected in research team. This interagency team would consist of
January and February, 1977, showed reduced levels of personnel from the EPA, FWS, NOAA and USCG, and be
less than 20 ppb. organized under the National Response Team. In our
Water samples were also collected from the view, a coherent team effort appears to be the best
Georges Bank/Nantucket Shoals region and analyzed by approach to the study of fate and effects of oil spills
gas chromatography and combined gas chromato- including ecological damage assessment.
graphy/mass spectrometry. Samples collected in
February, 1977 showed concentrations varying from 10- James G. Quinn, Chairman
100 ppb total hydrocarbons with a mean of 44 ppb. Chemical Studies Session
Concentrations fell off rapidly in May, 1-50 ppb (R = 11
ppb), and levelled off thereafter to the 1-20 ppb range in
August, 1977.
Tar balls found on the shores of Massachusetts and
Rhode Island between two and three months after the
spill were analyzed by infrared spectrometry. Based on
these analyses, it was concluded that theArgo Merchant
oil was not the source of these tar balls. There was some
indication that they may have come from the lost tanker,
Grand Zenith.
Thirty-seven samples of biota, collected from the
vicinity of the spill area in December, 1976 and January,
1977, were analyzed by gas chromatography and com-
bined gas chromatography/mass spectrometry. Three
samples of fish stomach contents gave hydrocarbon
distribution patterns which resembled the Argo Merchant
oil.
Gas chromatographic analysis of sediment samples
collected in February, 1977 indicated that three stations
at the wreck site contained Argo Merchant oil in the form
of small tar particles. In July, 1977, only one of these
stations contained petroleum hydrocarbons and the level
was greatly reduced over that found in February. The
high degree of turbulent mixing on the shoals was
probably responsible for the observed patchiness and
removal of contaminated sediments.
In order to provide additional information on the
Argo Merchant spill, the chemical studies workshop
recommends the following:
1) Obtain samples of cargo oils from the refinery in
Venezuela. These samples would be of use in chemical
matching of samples collected from the spill area
(e.g. biota samples analyzed by NOAA).
2) Collect and analyze water samples from the
Nantucket Shoals-Georges Bank region during the winter
of 1978. The February, 1977 values reported by Dr. Paul
Boehm were high and additional winter samples are

Water Soluble Fraction of Argo Merchant
Cargo

J. Richard Jadamec

U.S. Coast Guard Research and Development Center
Groton, Connecticut

The Argo Merchant, carrying 7.7 million gallons of cracked in half and discharged approximately one and a
blended Venezuelan No. 6 residual fuel oil (having a 2'C half million gallons of its cargo into the ocean. During the
pour point and a 0.96 specific gravity) ran aground on the next few days practically all of the remaining cargo was
shoals of Nantucket Island, subsequently discharging the discharged into the environment. Predictive models to
bulk of its cargo into the marine environment. A series of describe the surface drift of oil as a function of the
water column samples was collected in the area sur- prevailing wind and surface currents were available and
rounding the wreck site shortly after the grounding to were used effectively to predict the surface movement of
determine the extent and level of petroleum oil hydro- oil. However, models to predict the subsurface dispersion
carbons entering the environment as a result of this of oil do not exist, largely because the process of sub-
pollution incident. Fluorescence spectroscopic analysis surface oil dispersion is relatively unknown at this time.
of these samples indicated that a physical/chemical In an effort to gain information on the subsurface disper-
fractionation of the oil occurred with only the lighter sion of oil, a series of water column samples were
aromatic components (onew, two-, and three-membered collected by the U.S. Coast Guard, National Oceanic and
ring aromatic compounds) entering the water column. Atmospheric Administration, Woods Hole Oceanographic
These aromatic hydrocarbons are undoubtedly Institute, and the University of Rhode Island during the
associated with the cutter stock material used in blending period 20 December 1976 through 28 February 1977.
the industrial No. 6 fuel oil cargo carried by the Argo Collected water column samples were analyzed by
Merchant. Analysis of water column samples collected synchronous excitation/emission fluorescence spectro-
during the period of 20 through 28 December 1976 in the scopic techniques. This method as previously described
area surrounding the grounded Argo Merchant and along by Gordon et al. (1976) was selected as a rapid screening
the direction of surface oil movement (to the southeast) technique for handling large numbers of samples for three
indicated petroleum oil concentrations as high as 340 basic reasons: first, fluorescence spectroscopic
ppb. Subsequent analyses of water column samples techniques are quantitatively accurate when the type of
collected throughout this area during January and spilled oil is known and is available for use as a calibration
February 1977 revealed that the relatively high petroleum standard; second, the spectral fluorescence excitation/
levels found in December were reduced to levels less emission response of the polyaromatic compounds
than 20 ppb. present in a sample is both readily distinguishable and
indicative of the type of petroleum oil present; and third,
fluorescence spectroscopic instrumentation is rugged
Introduction and readily adaptable for shipboard use.
During the early morning hours of 15 December Experimental
1976, the Argo Merchant, carrying a cargo of 7.7. million
gallons of an industrial No. 6 fuel oil, ran aground on the Water column samples were collected using
shoals of Nantucket Island. Attempts undertaken to off- Niskin bottle, rosette or sterile bag samplers, and were
load the cargo between 15 and 20 December were made either: (a) extracted and analyzed aboard ship using
impossible by gusting 40 knot winds and seas running up fluorescence excitation/emission spectroscopic tech-
to twenty feet. Surveillance overflights during this period niques; or (b) immediately frozen and returned to the
revealed a daily increase in the leakage of oil into the laboratory for analysis. Portions of the collected sample
marine environment from the abandoned vessel. On the (100 ml or 1,000 ml) were extracted with spectroquality
morning of 21 December 1976, the Argo Merchant hexane. One liter volumes were extracted twice with two

2-1 PT 5 3&4 RINGS 5 AND YORE RT,4rS
--------------%r------------- I----------------N

1. 2 pg/ml

B1 k

280 320 36o 400 4,40 280 320 340 380 420 440
DIISSIDI WAVELDMTH W.

Figure 1. Synchronous excitationlemission spectra of three oil
types., No. 2 fuel oil (-); No. 6 fuel oil ( . . .); Bunker C (---- ).

10 ml aliquots of hexane, whereas 100 ml volumes were
extracted once with 10 ml of hexane. All extractions were 3.5 ug/ml
performed in a separatory funnel by hand shaking the
hexane/sea water mixture vigorously for two minutes.
All fluorescence excitation/emission analyses were
performed using a fully corrected Farrand Mark I Spectro-
fluorometer. All spectra were recorded using an excita-
tion bandpass of 10 nm; an emission bandpass of 2.5 nm;
B
and a scan speed of either 30 or 50 nm/minute, with a
o
constant offset of 25 nm between the excitation and 280 3120 340 3@O 460 4 4'0
emission monochromators.

Discussion

. Fluorescence spectroscopy, and in particular
fluorescence synchronous excitation/emission tech- 8. 2 pg/ml
niqbes, have been used to qualitatively characterize and
identify the presence of petroleum oils based on the
distribution of polyaromatic hydrocarbons (Gordon et al.,
1976; Lloyd, 1971a; Gordon and Keiser, 1975). The
synchronous excitation/emission spectra of several oil
types are shown in Figure 1. As can be seen from Figure 1,
the various oil types can be characterized based on the
distribution of the one-, two-, and multiple-ring com-
pounds. Benzene type compounds emit strongly in the
280 to 290 nrn region, napthalenes between 310 to 3W nm
region, three and four membered ring compounds between 280 320 3'60 40b 440 480
340 and 380 nm, and larger ring compounds above 400 C-IISSIal JAVELMOTH
nm. Additionally, it has been shown that between the con-
centration range from 1 W g/ml to 100 p g1ml the strong Figure 2. Synchronous excitationlernission spectra of Argo
fluorescence emission from both low and high molecular Merchant cargo at three concentrations.
weight aromatic components insures a highly structured
spectral profile (John and Soutar, 1976). Figure 2 shows sterile bag samplers. Gordon and Keizer (11974) have
the synchronous excitation/emission spectra of Argo shown that the sampling device presents a problem in the
Merchant oil at several concentration levels. As can be fluorescence analysis of sea water samples, mainly
seen from Figure 1, the Argo Merchant oil has a spectrum through desorption and adsorption processes taking
similar to that of a No. 6 fuel oil, containing a wide distribu- place between the collected sample and the walls of the
tion of polyaromatic compounds. sampling device. Figure 4 shows three spectral profiles
Fig ure 3 shows a typical fluorescence excitation/ of sea water which had been in contact with sterile bag
emission profile obtained in the analyses of sea water samplers for various periods of time, 1 hour, 2 weeks and
samples collected in December 1976 from the area one month at room temperature. Except for an increase
@1_2
pg/ml
@Bl k

3.
5
ml
131L
L. 2 ,@@ml

surrounding the wreck site of the Argo Merchant which is in the response in the 300 to 304 nm region of the spectral
distinctly different from the spectral profiles of Argo profile the spectral response shows the absence of two
Merchant oil shown in Figure 2. The majority of samples or more ring aromatic compounds. Additionally, Figure 5
collected in this investigation were obtained using Niskin shows the hexane extract of the sterile bag used in

306 3

323

323-

2.AO 5 33'0 -3155 3-8LO 4LO 5 43JO 5 210 3bo 3'20 340 3 k 3 @O 00 42b _4r0__
EMISSION WAVELENGTH EMISSION WAVELZrXTH
Figure 3. Synchronous excitationlemission spectra of sea water Figure 5. Synchronous excitationlemission spectrum of the
sample collected in December 1976 and sea water blank hexane extracted sterile bag containing the sea water sample
shown in Figure 3.

304

3 n'

3o6nm

D
304 IWNGORAYSTUDY AT 5% OFOUTHOORAYSTUDY AT 22'C

304 nm

I HOUR 2 WEEKS 1 MONTH
Sea Water Bleak Sea Water Blaak 11

280 300
TO 3 Or, 320 280 300 320 180 360 3 2b 320 3t5 3tO 2W 300 3 @O
EMISSION WAVELENGTH
DIISSION WAVELENGTH

Figure 4. Synchronous excitation/emission spectra of sea water Figure 6. Synchronous excitation/emission spectra of sea water
after storage in a sterile bag at room temperature foran extended after exposure to Argo Merchant oil for a four day period.
period of time.

collecting the sample shown in Figure 3. Although the 1977 FIN Endeavor cruise, where samples collected
walls of the bag have apparently absorbed some of the using Niskin bottle or sterile bag samplers had identical
aromatic hydrocarbons present, there is no evidence of spectral responses.
higher molecular weight hydrocarbons having been In an effortto explain the repeated occurrence ofthe
absorbed. Figures 4 and 5 illustrate that although the spectral profile shown in Figure 3, two separate labora-
sterile bag samplers do offer a source of contamination tory studies were undertaken. These two laboratory
in fluorescence analyses, and absorb aromatic hydro- studies involved the gentle spreading of Argo Merchant
carbons, neither the fluorescent contaminant induced oil on sea water and allowing it to stand for a four day
from the bag, nor use of the sterile bag itself, can explain period. The sea water used in this study was a composite
the spectral response obtained in the analyses of the of samples collected on board the RN Oceanus 19
typical sea water sample shown in Figure 3. Furthermore, cruise, prior to the break-up of theArgo Merchant. Figure 6
the effect of this contaminant can be minimized if the shows the synchronous excitation/emission profiles of
samples are processed immediately after collection as extracted subsurface water samples from both studies.
shown in Figure 4 (one hour). This was evident in the There does exist a slight difference in the spectral
0
@r- '_ "Y'T"Y AT
IN
R
DOG
0
AY
STUDY IT 5'1
,___ _ k@3
@H.U\R
WEEKS Sea Water Laa@ I' Sea Wat k.@-kl'

shipboard analysis of samples collected on the January response of the 5oC indoor, compared with the average

220C outdoor, test, but in each study there is a complete
absence of high molecular weight aromatics. These
differences can be explained by the relative solubility of
aromatic hydrocarbons in water (Bohon and Claussen,
1951; Boylan and Tripp, 1971) and the effects of
weathering on the change in the synchronous excitation/
emission profile of subsurface water samples taken from
beneath an oil slick (Gordon et al., 1976). Additionally it was
shown in the weathering of a Guanipa crude oil in an out- It
door test tank that the proportion of two tinged aromatics
in samples taken 0.5 meters beneath the surface in-
creased relative to the higher molecular weight aromatics R
during the course of this study. However, the spectral pro-
files shown in Figure 6, for the study conducted at 5*, re-
mained unchanged over a four day period. The only differ-
ence observed in samples taken at 24 hour intervals was
an increase in the relative intensity of the spectral
response, with no evidence of higher weight aromatics ao 300 L
2
being present. ENISSIGN WAVELMM
The spectral responses shown in Figure 6 are
similar to that shown in Figure 3, apparently indicating Figure 7. Fluorescence emission spectrum of the extracted sea
that some type of fractionation of the Argo Merchant oil water sample shown in Figure 3, excitation wavelength 254 nm.
occurred. Studies by Anderson et al. (1974) showed that
when oil is spilled or discharged into the environment two Studies by Milgram (1977) revealed that the Argo
distinct oil-water fractions are formed: a water-soluble oil Merchant cargo contained 20 percent of a light petroleum
fraction (WSF) and an oil-water dispersion (OWD). The fraction. This fraction is believed to be a "cutting stock"
magnitude of each fraction occurring after a discharge is distillate used to blend with the residual fuel oil to improve
dependent on the type and amount of oil discharged; the the oil's handling characteristics in obtaining the
rate and type of discharge (slow leak, pumping, rupturing observed 2'C pour point of the cargo. Furthermore,
of tanks, etc.); and the existing environmental conditions. Milgram found that after removing the light petroleum
In high sea states the OWD fraction will be significant, fraction by distillation the specific gravity of the residual
with its extent being dependent on the physical charac- fuel oil remained unchanged: 0.96. The cargo carried by
teristics of the oil in question, i.e. its pour point, API the Argo Merchant was similar in composition to that
gravity, etc. These same physical parameters will also carried by the tanker Arrow which went aground in
affect the spreading characteristics of oil on water, the Chedabucto Bay, Nova Scotia, during 1970. Both cargos
magnitude of oil-in-water emulsions being formed and the were blended Venezuelan residual fuel oils. Studies by
speed at which the oil will undergo weathering changes Mackay et al. (1973) and Betancourt and McLean (1973)
(solution, evaporation, bio-degradation, etc.). showed that weathering processes (evaporation, biode-
At the time of the grounding, the Argo Merchant was gradation, solution) affects the composition of the spilled
carrying a blended Venezuelan No. 6 residual fuel oil oils by removing the light aromatics and increasing the
cargo having a specific gravity of 0.96 and a pour point of asphaltene concentration which is associated with a
TC. When this cargo was discharged into the environ- large increase in the viscosity of the oil phase preventing
ment, large "pancakes" were formed 11/2 to 2 inches thick both the coalescence and sedimentation of the water-in-
and were reported to behave "like mercury" (Milgram, oil emulsion droplets. This was observed to be particular-
1977). Overflights revealed oil sheens generally upwind of ly true for residual fuel oils spilled in cold water.
the pancakes. The viscous nature of this spilled oil resulting
from both the oil's pour point and existing water and air
temperatures, 50 to 60C, prevented any rapid spreading Results
of oil on water. This also reduced the amount of oil which
could be mechanically driven into the water column to Recent studies (Cretney et al., 1977) have sug-
form the OWD fraction. gested the use of any aromatic standard, chrysene, as a
Anderson et al. (1974) studied the behavior of 4 API standard which could be usecl to report sea water petro-
reference oils (Kuwait and South Louisiana crude oils, leum oil concentrations. Chrysene was selected because
and No. 2 and Bunker C fuel oils) in sea water. They the fluorescence excitation/emission response of chry-
observed that the WSF fractions of the four test oils were sene was found to be similar to that of a heavy fuel oil.
greatly enriched in the aromatic hydrocarbons (one-, two- Others (Hiltebrand, in press; Levy, 1973) have selected
and three-membered aromatic ring compounds) in com- the maximum fluorescence emission band of the parent
parison to the parent oil, whereas the OWD fractions con- oil, when excited at a fixed wavelength, to establish
tained a distribution of hydrocarbons resembling that of a calibration curve to determine sea water petroleum
the parent oil. Similar results have also been reported in oil concentrations. In this study the selection of either
the analysis of water which had been in contact with other approach would have yielded misleading values.
oi I types (Boylan and Tripp, 1971). The water soluble com- Figure 7 shows the fluorescence emission curve of the
ponents in these studies were again benzene and naphtha- extracted sea water sample shown in Figure 3 when
lene type compounds. Additionally these studies indicated excited at a fixed wavelength of 254 nm. There exists no
that oil-in-water emulsion contained a distribution of hydro- significant response in the 365 nm region, the major
carbons similar to that of the parent oil used in these fluorescencing region, when a fixed wavelength of 254 nm
studies. is used, or the, 366 nm, 383 nm and 400 nm region if

Table 1. Estimated Petroleum Concentration Levels (ppb) ly associated with light distillate fuel oils. The high
During December 1976 percentage of "cutting stock" material used in blending
the residual fuel oil carried by the Argo Merchant is
undoubtedly responsible for the petroleum level con-
Latitude (N) 4 1 *07. 0' 40'52.0' 40'45.0' 40'55.0' cehtrations reported in Table 1. Since a sample of this
Longitude (W) 69'56.0' 69'35. 0' 68*25.0' 6z-56. 0' cutting stock material was not available, the concentra-
tions listed in Table I are estimates based on the
Surface 230 250 0 310 response of the Argo Merchant oil in the 306 nm and
5 Meters 310 150 180 140 323 nm region.
10 Meters - 200 270 - Analysis of samples collected in January and
20 Meters - 210 170 February 1977 from the area surrounding the wreck site
indicated that the high levels found in December 1976
had dissipated. Synchronous excitation/emission
Latitude (N) 4 1'0 1. 3' 41*01.4' 40'58.5' 4 1*03.0' spectra of these samples were similar to those obtained
Longitude (W) 6-c?22.0' 69'26.7' 69*24.0' 69'34.0' from samples collected on the R/V Oceanus 19 cruise
prior to the break-up of the Argo Merchant.
Surface 120 90 170 40 This study has demonstrated that the use of
3 Meters 0 0 0 120 fluorescence spectroscopic techniques to quantify
petroleum oil concentration levels is strongly dependent
on the selection of a standard. Prior to selecting a
Latitude (N) 4 1'00. 0' VoOl.6' 40'48.8' standard, whether it be the parent oi I or a single aromatic
Longitude (W) 69'25.5' 69*28.6' 69*04.0' compound or mixtures of aromatics, it is advisable to
analyze the collected samples by synchronous
Surface - 270 79 Meters 170 excitation/emission techniques to determine the distribu-
3 Meters 340 - tion of aromatic compounds present. If only a mixture of
one-, two-, and three-membered ring compounds is
chrysene were used as a standard. If either of these two present, and an oil-in-water dispersion is not obvious,
methods were employed to report sea water petroleum then a mixture of one-, two-, and three-membered ring
oil concentrations in the area surrounding the Argo compounds should be selected as a standard. Based on
Merchant wreck site immediately after its break-up, the work of previous investigators, this would be a logical
negligible concentrations would have been reported. choice for a standard, since thewater soluble compounds
Table I lists "estimated" petroleum oil concentra- which are shown to be readily soluble are light aromatics,
tions for representative samples collected during the regardless of the parent oil. Furthermore, the use of
period 22 through 24 December 1976. All reported con- standards which have responses similar to heavy oils or
centrations are relative to the cargo carried by the Argo relying on the major fluorescencing band of heavy oils to
Merchant. Figure 2 indicates two fluorescencing bands, determine petroleum oil concentrations can lead to
one at 306 nm, and the second at 323 nm. The ratio of erroneous results, particularly if a physical/chemical
these two band intensities were plotted against concen- fractionation of the oil has occurred, as is suspected in
tration. These same two bands are seen in Figure 3, a the analyses of water samples collected after the Argo
typical synchronous excitation/emission spectrum of an Merchant oil spill.
The opinions or assertions contained herein are the
extracted sea water sample. Those samples showing private ones of the writer and are not to be construed as
zero petroleum oil concentrations in Table I were used as official or reflecting the views of the Commandment or the
blanks in calculating the other sample concentrations Coast Guard at large.
listed in Table 1. The synchronous excitation/emission
spectra of these samples selected as blanks had no References
significant spectral features in the 306 nm or 323 nm
region. Additionally, these samples had the same resi- Anderson, J. M., Neff, B. A., Cox, H. E. and Hightower, G. M.,
dence time in the sterile bags, minimizing any possible Marine Biology, vol. 27, 75 (1974).
effects of absorption processes which may have Betancourt, 0. J., McLean, A. Y., J. of the Institute of Petroleum,
vol. 59, 223 (1973).
occurred in storage of these samples between collection Bohon, R. L. and Claussen, W. F., J. of Amer. Chem. Soc., vol. 53,
and analysis times. 1571 (1951).
Boylan, D. B. and Tripp, B. W., Nature, vol. 230, 44 (1971).
Cretney, W. J., Johnson, W. K. and Wong, C. S., Pacific Marine
Conclusions Science Report 77-5 (1977).
The fluorescence emission spectrum of the sample Gordon, D. C., and Keizer, P. D., Technical Report No. 481,
Environment Canada, Fisheries and Marine Service
shown in Figure 7 is typical of that which would be (1974).
obtained if the sample in question was a No. 2, or light Gordon, D. C., Keizer, P. D., Hardstaff, W. R., and Aldous, D. G.,
A distillate fuel oil. The samples listed in Table I all have Environ. Sci. Technol., vol. 10, 580 (1976).
the same spectral response indicating the presence of a Hiltabrand, R. R., Marine Pollution Bulletin, In Press..
No. 2, or light distillate fuel oil. Previous investigators John, P. and Soutar, I., Anal. Chem., vol. 48, 520 (1976).
have shown the rapid solubility of light aromatic com- Levy, E. M., J. Fish. Research Board Canada, No. 3, 261 (1973).
pounds in sea water, and regardless of the parent oil Lloyd, J. B. F., J. Fores Sci. Soc., vol. 11, 83 (1971 a).
used in these studies, the same water soluble com- Mackay, G. D. M., McLean, A. Y., Betancourt 0. J. and Johnson,
B. D., J. of the Institute of Petroleum, vol, 59, 164 (1973).
ponents were found to be present. These one-, two-, and Milgram, J. H., U.S. Department of Commerce, NOAA Special
three-membered ring aromatic compounds are common- Report, "The Argo Merchant Oil Spill" (11977).

Hydrocarbon Chemistry of the Water
Column of Georges Bank and Nantucket
Shoals, February-November 1977

Paul D. Boehm, George Perry, David Fiest

Energy Resources Co. Inc. (ERCO)
185 Alewife Brook Parkway
Cambridge, Massachusefts

Abstract. hydrography, biology, and chemistry of the Georges
Large volume water samples (50-90 liters) were Bank/Nantucket Shoals/Lower Gulf of Maine region. As
collected in mid-February, May, August and November of part of the overall hydrocarbon chemistry program, large-
1977 at twelve stations in the Georges Bank/Nantucket volume water and surface microlayer samples were
Shoals region. Near surface and bottom samples were obtained to describe the dissolved and particulate hydro
filtered on-board and the filtrate extracted in the labora- carbon compositions of the region as they varied
tory, using a countercurrent liquid-liquid extractor. seasonally.
Virtually nothing was known of the detailed water
Aliphatic and aromatic hydrocarbon compositions of the column hydrocarbon chemistry of the region prior to this
particulate and dissolved fractions were determined by set of cruises. In fact, there is only limited information on
glass-capillary gas chromatography and combined gas the distribution of hydrocarbons in the water column of
chromatography/mass spectrometry. the world's continental shelves. A summary of some of the
Samples collected in the winter show markedly available hydrocarbon data is presented in Table 1. A
higher dissolved hydrocarbon concentrations than those wide range of hydrocarbon concentrations has been
analyzed from later seasons. Concentrations throughout reported. In some cases the lack of a consistent reporting
the study area ranged from 10-100 ppb total hydro- format s6mewhat impairs our ability to intercompare
carbons in the winter with a mean of 44 ppb. Concentra- values, but, in general, water column hydrocarbon con-
tions fell off rapidly at the time of the spring cruise, centrations have been reported in the 0.2 to 100 ppb
1-50 ppb (X_ = 11 ppb) and levelled off thereafter to the concentration range. The highervalues arefound in areas
1-20 ppb range. The entire study area showed elevated of heavy shipping traffic.
hydrocarbon. levels at the time of the winter sampling. Our study commenced six Weeks after the breakup
Qualitatively the dissolved hydrocarbon assem-
blages in the winter appear to be a composite of: 1) lower of the Argo Merchant oil tanker and in effect represents
boiling aliphatic and aromatic compounds (boiling range the only follow-up study undertaken which covers the
n-C1 4-n-C22); and 2) a higher boiling unresolved complex entire region of Georges Bank and Nantucket Shoals. We
mixture (UCM). The lower boiling compounds are similar present here a general overview of the data on the hydro-
to those determined in a laboratory sea water accommo- carbon distributions in the water column and the changes
that took place in these distributions in the eleven months
dated fraction of Argo Merchant cargo oil. As. the concen- after the spill.
trations decreased throughout the year so did the
relative contribution of these lower boiling compounds.
The hydrocarbon distributions in the particulate Methods
fraction varied from a direct coincidence with the Sampling. Large-volume water samples (45-90 liters)
dissolved fraction's distribution to ones more suggestive were obtained at twelve stations shown in Figure 1. An
of biogenic inputs combined with pelagic tar. anodized aluminum 90-liter Bodman bottle (Benthos
Corp.) was deployed at the near surface (3 meters) and
Introduction near bottom (3 meters off bottom) at each station. The
design and operation of this sampler permits the bottle to
Four seasonal sampling cruises were undertaken be opened after it has passed through the potentially
during 1977 to characterize the benthic and watercolumn contaminating sea surface film. It is then lowered to

Table 1. Hydrocarbons in Sea Water

Concentration
Location (Ag1l) Comments Reference

Georges Bank Region 0.2-98 Gas Chromatography (GC) This study
South Texas OCS 0.1-2.0 Paraffins only Berryhill (1977)
Alaska OCS GC Shaw (1977)
Gulf of Mexico Loop Current 0-75 GC Iliffe & Calder (1974)
West African Coast 10-95 GC Barbier at al. (1973)
French Coast 46-137 GC Barbier at al. (1973)
Open Ocean (Atlantic) 1-50 IR Brown at al. (1973)
<6 Fluorescence Gordon at al. (1974)
20 1-3 mm Gordon at al. (1974)
Fluorescence
Mediterranean Sea 2-200 Surface OR) Brown at al. (1975)
2- 8 Subsurface OR) Brown at al. (1975)
Atlantic 0.5-6 % Brown at al. (1975)
Baltic Sea 50-60 Non-aromatics Zsolnay (1972)
Gulf of Mexico (coastal) .1-.6 n-alkanes; only Parker at al. 0 972)
Galveston Bay area 8 Brown at al. (1973)
New York Bight 1-21 Brown at al. (1973)
Gulf of Venezuela 50 Brown et al. (1973)
Bedford Basin, Nova Scotia 1-60 Keizer & Gordon (1973)
Gulf of St. Lawrence 1-15 Levy& Walton (1973)
Narragansett Bay 8.5 GC D uce at al. 0 972)
5-15 GC Boehm (1977)
Woods Hole Harbor 11 GC Stegemann &Teal (1973)

W Gillis in the summer (August), and aboard the R/V Knorr in
the fall (November).
Analytical Methods. Dissolved hydrocarbon samples
extracted in an all stainless steel and glass counter-
4r were
current liquid-liquid extractor designed by A. Himmelblau
I:iif
of Energy Resources Co. Two and one-half liters of
chloroform (Baker Resi - Analyzed) were used and 95
p
ercent extraction efficiency achieved in four hours. This
J,I was determined by three successive extractions of a
'Ilk single sample of two hours' duration each.
t 11.12 P!@
The filters were thawed, placed in a round-bottom
C51-
flask, and extracted with chloroform and hexane under
reflux for four hours.
The surface microlayer samples were extracted in
three-batch chloroform extractions in separatory funnels.
All of the extracts were reduced in volume on a
rotary flash evaporator and concentrated to near dryness
Figure 1. Station locations and coordinates. under a stream of water-pumped purified nitrogen. The
total extractable lipid weight was determined by weighing
an aliquot on a Cahn electrobalance. The total extract
depth, closed by messenger and brought aboard. Ninety was then charged to a glass column (1 cm diameter)
liters are then pressure-filtered (110 psig) within one hour packed with 2.5 g 5% deactivated alumina over 7.5 g
after sampling through a 142-mm pre-combusted Gelman fully activated (200'C, 4 hours) 100-200 mesh silica
A-E glass fiber filter set in a stainless steel filter holder. (W.R.Grace Co.). The column was eluted with 18 ml
Water samples were collected in solvent-rinsed stainless hexane (fraction 1) followed by 20 ml benzene (fraction 2)
steel drums (30-gallon) or 12-gallon glass carboys. to collect the aliphatic (fl) and aromatic/olefinic 02)
Surface microlayer samples were collected (7 hydrocarbons. Concentration of these fractions by rotary
liters = 60 M2 of surface) at three stations in the winter, evaporation was followed by their analysis by gas
eight in the spring and seven in the summer, using the chromatography (GQ. A Fjewleft-Packard model 5840A
screen technique of Garrett (11965). Both the microlayer gas chromatograph equipped with the HIP 18835A glass
and dissolved samples were poisoned with chloroform on capillary inlet system was used. The capillary column was
k)4 f

board to retard microbial degradation. a 15 meter (-50,000 theoretical plates) SE-30 column
The samples were obtained on R/V Gyre cruises in (J & W Scientific) and the samples run in the splitless
the winter (February) and spring (May), aboard the R/V mode.

Table 2. Dissolved Hydrocarbon Data Summary

Total Hydrocarbon (11g1l)
Station No. Depth Winter Spring Summer

1 3 23.7 5.0 0.9
4 32 27.8 17.9 1.0

4 3 20.8 4.8
64 16.3 5.8 0.2

7 3 48.9 2.8 -
85 98.5 4.7 - Figure 2A. Dissolved hydrocarbons: winter, fractioril.

13 3 59.8 7.8 -
so 10.2 1.8 3.1

18 3 27.4 9.0 0.3
65 38.9 5.7 0.2

26 3 16A 12.5 2.0/1.0
150 19.7 1.2 1.5

32 3 22.6 17.4 0.8
65 36.5 8.6 0.4

35 3 - 5.7 -
230 4.5 -
Figure 2B. Dissolved hydrocarbons: winter, fraction 2.
36 3 - 49.0 0.2
100 30.9 36.8 0.2
enriched in hydrocarbons relative to the surface waters
37 3 - 5.0. 0.2 while the surface waters at Station 13 are more concen-
30 - 10.7 0.8 trated than the near-bottorn sample. At all other stations
the water column appears well mixed in the winter, an
39 3 32.5 20.4 0.3 observation consistent with the hydrographic data
85 26.9 14.9 8.1 (Strimaitis, personal communication).
Information on sampling replication is not yet
42 3 14.4 7.7 2.9 available. Duplicate subsamples of 45 liters each were
190 47.5 7.3 1.3 analyzed on a split summer sample and results indicate
a precision of � 59 percent for an absolute sample
concentration in the 1-2 p g/t, range.
Concentrations of hydrocarbons were determined At most of the stations, the total hydrocarbon con-
by combining information on individual component con- centrations decreased over the course of three seasons.
centrations, obtained from a PDP-1 0 computer interfaced At Stations 36 and 39 (surface and bottom) and at
to the GC through the HIP 18846A Digital Communications Stations 26 and 32 (surface), the winter and spring values
interface. The unresolved complex mixture (UCM) was remained constant, but then decreased with all of the
measured from the chromatograms directly by planimetry other station concentrations in the summer (August
and quantified relative to the area of the internal standard. 1977). Of special interest is the Station 26 data. The
Combined gas chromatography/mass spectro- water column chemistry at this station is greatly in-
metry was performed on a Hewlett-Packard model 5981 A fluenced by an oceanographic frontal system alternately
quadrupole mass spectrometer coupled to a model 5710 transgressing and regressing over the southern edge of
gas chromatograph. Georges Bank (L. Sick, D. Strimaitis, personal communi-
cations). This frontal system became well developed at
the time of sampling in the spring. The order of magnitude
Results and Discussion differences in the surface and near-bottom "dissolved"
hydrocarbon values attest to our sampling of two different
Dissolved Hydrocarbons. Total hydrocarbon concen- water masses at this station in the Georges Bank surface
trations in the "dissolved" fraction ranged from 10-100 water and the deeper slope water.
ppb in the winter samples (Table 2). The average surface "Dissolved" hydrocarbon concentrations in the spring
water (3 meter) concentration, 29.9 p g/ t , is not signif i- (May) averaged 13.2,ug/t (range 2.8-49.0) in the surface
cantly different from the bottom water samples, 35.0 waters and 9.0 pg/t (range 1.2-36.8) in the near-bottom
g/ t . However, differences do appear when consi- samples. A further sharp decrease was noted in the sum-
dering individual station concentrations. Table 2 indi- mer samples, at which time surface waters averaged
cates that the bottom waters at Stations 7 and 42 appear 0.8 1! g/ I (range 0.2-3. 1) and bottom waters 1.7 Wg/ t

Figure 3A. Dissolved hydrocarbons: spring, fraction 1.
Figure 4A. Dissolved hydrocarbons: summer, fraction 1 (2ppb).

Figure 3B. Dissolved hydrocarbons: spring, fraction 2.
Figure 4B. Dissolved hydrocarbons: summer, fraction 1 (2ppb).
(range 0.2-8.1). Preliminary information on the fall
(November) samples indicate that their concentration
range is similar to the summer samples.
Analytical blanks were run periodically and all
samples were at least twice the blank level (0.1-0.2 ppb).
Values in Table 2 have been corrected for the blanks. In
addition, shipboard samples of all fuel and lubricating oils
were collected and chromatographed. Comparisons of
the gas chromatograms of the oil samples and the water
samples indicated that no shipboard contamination of
the samples occurred.
More revealing than the total concentration data are
the gas chromatographic characteristics of the samples.
The fj (hexane eluate) chromatograms from the winter
samples appear to be a composite of n-alkanes and
branched alkanes in the n-C1 2-n-C22 range, the typical Figure 4C. Dissolved hydrocarbons., summer, fraction 2.
boiling range of a No. 2 fuel oil, and an unresolved
complex mixture (UCM) of coeluting hydrocarbons with a
maximum detector response at n-C28 (Figure 2A). A spring. The relative input of the n-C12-n-C22 alkanes
chromatographic distribution similar to that found in this began to decrease in the spring samples (Figure 3A) and
study has been observed by other researchers in the Gulf its predominance waned as the absolute concentrations
of Mexico Loop Current (Iliffe and Calder, 1974), and in decreased to the 0.2-2.0 ppb range found in the summer
the oil tanker route off the coast of Africa (Barbier et al., samples (Figure 4A and B). The fj hydrocarbon assem-
1973). blage in the low-level (-0.5-2 ppb) summer dissolved
The UCM generally accounts for 60 to 80 percent of samples revealed a relatively smooth n-alkane distribu-
the total determined hydrocarbon concentrations. This f 1 tion from n-C 16 through n-C31 with biogenic inputs Ofn-Cl 5
pattern was quite characteristic of the winter and spring and the isoprenoid pristane rising out of the chromato-
samples, although present in lesser concentrations in the graphic pattern (Figure 413). This summer pattern is similar

Figure 5A. Surface microlayer., winter, fraction 1. Figure 6A. Argo Merchant cargo oil, fraction 1.

Figure 613. Argo Merchant cargo oil, fraction 2.

Figure 513. Surface microlayer winter, fraction 2. Surface Microlayer. Along with the "dissolved" hydro-
carbon samples, the surface microlayer samples
exhibited lower hydrocarbon concentrations in the spring
and summer than they did in the winter, averaging 54
to the surface microlayer samples obtained using the g g/t (range 32-76) in the winter, 20 gg/,t in the spring
stainless-steel screening technique of Garrett (11965). (range 5-41) and 10-20 g g/1 in the summer. This repre-
However, in the case of the microlayer samples, the n- sents about a threefold average winter enrichment in the
alkanes smaller than n-C25 appear depleted, relative to surface microlayer relative to the subsurface waters
the dissolved samples (Figure 5A vs. Figure 413). (range 1.4-3.7), and a fivefold spring enrichment. Pre-
The fraction 2 (benzene eluate) hydrocarbons of the liminary data on the summer microlayer samples indi-
winter (Figure 2B) and spring dissolved samples contain cates that the microlayer was enriched to the same
prominent methylated naphthalene components and extent as it was in the spring.
phenanthrene and alkyl phenanthrene compounds (0.02- The chromatograms of the surface film samples
0.05 ppb). This distribution suggests a petroleum source. from all seasons show various patterns resembling pelagic
Likewise the surface film winter samples (Figure 5B) tar inputs (Butler et al., 1973), probably in the form of small
exhibit significant (0.05-0.1 y g/t ) amounts of the particles (0.1-1000 jim in diameter; Wade et al., 1976).
naphthalene and phenanthrene series. Many samples It is also possible that an Argo Merchant input is con-
also contain small amounts of fluoranthene and pyrene tributing to the observed pattern in the microlayer and the
(m/e 202) compositions as determined by combined gas lower boiling alkanes (n-Cl 2-n-C22) are depleted (solubi-
chromatography/mass spectrometry. Figure 213 is typical lized?) and appear in the dissolved fraction (see Figure 2).
of a winter f2 distribution in those samples of intermediate The summer microlayer samples have similar GC patterns
hydrocarbon concentrations (10-20 ppb). In this sample to the winter and spring samples, although much enriched
type, many aromatic hydrocarbons ride atop a broad in planktonic components, pristane and n-C1 5. However,
LICK an aromatic distribution closely resembling the the summer "dissolved" hydrocarbon samples do not
Argo Merchant f2 distribution (F'gures 613 and 7). In the show the striking n-C14-n-C21 alkane predominance
spring samples, the f2 UCM dominates and resolved as did the winter and spring samples.
aromatic components are less frequently observed
(Figure 3B). The summer samples, when containing total Particulate Hydrocarbons. Total hydrocarbon concentra-
hydrocarbon concentrations of >1.0 ug/ t exhibit an tions in the particulate fraction remained fairly constant in
UCM along with a series of high molecular weight bio- the 0.3-5 p g/ t range in the winter and spring samples
genic olefins (Figure 4Q. As pristane achieved a chroma- and decreased (0. 1 -2 M g/1 ) in the summer. The particu-
tographic prominence in the summer fj sampi 'es, so did lates exhibit several hydrocarbon fj patterns (Figures 8A,
these olefins. A planktonic input is indicated. 9A, 9B) ranging from (1) mid to high molecular weight

JA)

2di-myl
4ph..d=
4 -1hV1 [email protected]_
.'A@ lhw.@- @ II -
dhd ,ph-
T 8--d-
7 P".-
T 10 .1-ft]
4 '0
T

a
TI . . . . . . . . . .. 1-121", Figure 9. Particulate hydrocarbons. spring cruise (May 1977).
- S 1. 55 A. fl (hexane), Station 42, near bottom. B. fl (hexane), Station
Figure 7. Reconstructed gas chromatograms. A. Argo Merchant - 32, near surface.
F2. B. Winter dissolved - F2.

sampling bottle which permits about 200 ml below the
(At bottle outlet to escape filtration (Gardner, 1977).

Discussion
Hydrocarbon values in the water column of the study
region were markedly elevated in the winter of 1977. Is
this a normal occurrence due to increased oil-tanker
AN traffic during winter months in this region, or are we
lot seeing a widespread pulse of Argo Merchant spilled oil?
To answer this, a further winter sampling of the water
column seems appropriate. For now, with the existing
data, we must compare the cargo oil with the hydro-
carbons found in the samples.
Analysis of an Argo Merchant cargo oil sample
obtained from J. Milgram (M.I.T.) reveals (Figure 6A) a
smooth distribution of n-alkanes ranging from n-Cl 1 to
n-C36 overriding a bimodal UCM in the f1 fraction. The
aromatic 02) fraction contains prominent methylated
naphthalenes, phenanthrenes and smaller quantities of
Figure 8. Particulate hydrocarbons: winter cruise (February four- and five-ringed aromatics. Several thiophenes are
1977). A. tj (hexane), Station 1 near bottom. B. fl (hexane), also observed (Figure 7).
Station 32, near bottom. Apparently, the cargo contained approximately 20
percent of a cutter stock, having a component distribution
similar to No. 2 fuel (Grose and Mattson, 1977). Due to low
n-alkanes, (2) a type 1 input with marked biogenic inputs, sea-water temperatures the spilled oil soon formed large
and (3) near total zooplankton input. The biogenic input iceberg-like pancakes. However, the rough seas
became apparent in the spring and summer samples. A probably allowed for substantial dispersion of oil and
fourth type of distribution (Figure 8B) closely resembles hence dissolution of the more soluble fractions in the
the dissolved hydrocarbon distribution. water column and dispersion as small droplets. Two
The apparent lack of correlation of the dissolved and attempts have been made in our laboratory to simulate
particulate hydrocarbon concentrations, along with the the chemical fractionation that may have taken place. In
fact that there is no major winter particulate hydrocarbon one case the oil was placed on sea water and slow stirring
increase is puzzling. The explanation may be due to the of the aqueous layer achieved. After 24 hours, gas
presence of the so-called dissolved hydrocarbons in (1) a chromatographic analysis of the aqueous phase resulted
truly dissolved state, (2) a collodial state (solubilized), as in an n-alkane distribution closely resembling the n-C1 2-
well as (3) a particulate (<0.5p) state (Boehm and Quinn, n-C22 n-alkane distribution observed in the winter and
1974) which is not trapped under pressure filtration by a spring water samples. In a modification of this experi-
Gelman A-E glass fiber filter. Furthermore, the larger ment, the aqueous layerwas stirred turbulently and the oil
particulate material (-4 11 m and up) may have settled out dispersed as small particles. Upon sampling this aqueous
rapidly (< 15 minutes) prior to filtration and hence may not layer, after allowing the mixture to stand, an unaltered
ak I _J@

have been filtered out due to an imperfect design of the Argo Merchant fl distribution resulted.

These preliminary observations, which will be Duce, R. A., J. G. Quinn, C. E. Olney, S. R. Piotrowicz, B. J. Ray,
repeated and refined, suggest that a chemical fractiona- and T. L. Wade. 1972. Enrichment of heavy metals and
tion of the cargo may have occurred, owing to processes organic compounds in the surface microlayer of Narragan-
of dispersion, true solution and micellar solubilization sett Bay, Rhode Island. Science 176: 161-163.
(Boehm and Quinn, 1973, 1974). The components most Gardner, W. D. 1977. Incomplete extraction of rapidly settling
affected were those in the boiling range of a typical No. 2 particles from water samplers. Limnology and Oceano-
graphy 22: 764-767.
fuel oil. This fractionation did not affect the aromatic (f2) Garrett,W. D. 1965. Collection of slick forming materials from the
distribution as evidenced in Figure 7, where a distribution sea surface. Limnology and Oceanography 10: 602-605.
similar to the f2 of the oil itself was seen. In studying the Gordon, D. C., Jr., P. D. Keizer, and J. Dale. 1974. Estimates
solubility behavior of a light distillate, Boehm and Quinn using fluorescence spectroscopy of the present state of
(1974) observed that micellar solubilization affected petroleum hydrocarbon contamination in the water
mainly aliphatic components and that the solubility column of the northwest Atlantic Ocean. Marine
behavior of the aromatics was independent of the Chemistry 2.
behavior of the aliphatics. A similar process may be Grose, P. L. and J. S. Mattson, eds. 1977. The Argo Merchant oil
involved here. However, more laboratory studies dealing spill - a preliminary scientific report. NOAASpecial Report,
with such subtleties are needed, and, further, detailed U.S. Dept. of Commerce, Washington, D.C.
Iliffe, T. M., and J. A. Calder. 1974. Dissolved hydrocarbons in the
chemical studies of spills of opportunity are required to eastern Gulf of Mexico Loop Current and the Caribbean
help to confirm what may be very complex solubility and Sea. Deep Sea Research 21: 481-488.
solubilization phenomena associated with spilled oil in Keizer, P. D., and D. C. Gordon, Jr. 1973. Detection of trace
sea water. amounts of oil in seawater byfluorescence spectroscopy.
Without further studies both in the field and in the Journal of the Fisheries Research Board of Canada 30:
laboratory, two possibilities remain: 1039-1046.
(1) The hydrocarbon concentrations were en- Levy, E. M., and A. Walton. 1973. Dispersed and particulate
hanced in the winter. of 1977 because of normal petroleum residues in the Gulf of St. Lawrence. Journal of
shipping operations associhted with bringing oil to the the Fisheries Research Board of Canada 30: 261-267.
Parker, P. L., J. K. Winters, and J. Morgan. 1972. A base-line
northeast U.S. study of petroleum in the Gulf of Mexico. Pp. 555-581 in
(2) The Argo Merchant spill affected most of the Base-line studies of pollutants in the marine environment.
Georges Bank/Nantucket Shoals region and caused a National Science Foundation, IDOE, Washington, D.C.
large increase of the hydrocarbon burden of the water Shaw, D. G. 1977. Hydrocarbons: Natural distribution and
column. dynamics on the Alaskan Outer Continental Shelf. Annual
Report, NOAA, Boulder, Colorado.
Stegemann, J. J., and J. M. Teal. 1973. Accumulation release and
Acknowledgments retention of petroleum hydrocarbons by the oyster
Crassostrea virginica. Marine Biology 22: 37-44.
This investigation was supported With funds from the Wade, T. L., J. G. Quinn, W. T. Lee, and C. W. Brown. 1976.
Bureau of Land Management (Contract No. AA 550-CT6- Sourceand distributionof hydrocarbons in surface waters
51). We thank Dr. Warren Steele of ERCO for the of the Sargasso Sea. In Proceedings of Symposium on
GC/IVIS analyses. Sources, Effects, and Sinks of Hydrocarbons in the
Aquatic Environment. AIBS, Arlington, Virginia.
Zsolnay, A. 1972. Preliminary study of the dissolved hydro-
References carbons and hydrocarbons on particulate material in the
Gotland Deep of the Baltic, Kieler Meeresforsch 27:
Barbier, M@, D. Joly, A. Saliot and D. Tourres. 1973. Hydrocarbons 129-134.
from sea water. Deep Sea Research 20: 305-314.
Berryhill, H. L. 1977. Environmental Studies, South Texas Outer
Continental Shelf, 1975: An Atlas and Integrated Sum-
mary. U.S. Bureau of Land Management, New Orleans,
Louisiana.
Boehm, P. D. 1977. The transport and fate of hydrocarbons in
benthic environments. Ph.D. Dissertation, University of
Rhode Island, Kingston, Rhode Island.
Boehm, P. D., and J..G. Quinn. 1973. Solubilization of hydro-
carbons by dissolved organic matter in sea water. Geo-
chimica et Cosmochimica Acta 37: 2459-2477.
Boehm, P. D., and J. G. Quinn. 1974. The solubility behavior of
Number.2 fuel oil in seawater. Marine Pollution Bulletin 5:
101-105.
Brown, R. A., T. D. Searl, J. J. Eliot, B. G. Phillips, D. E. Brandon,
and P. H. Monaghan. 1973. Distribution of heavy hydro-
carbons in some Atlantic Ocean waters. Pp.,505-519 in
Proceedings, Joint Conference on Prevention and Control
of Oil Spills. API, Washington, D.C.
Brown,.R. A, J. J. Eliot, J. M. Kelliher, and T. D. Searl. 191@3.
Sampling and analysis of non-volatile hydrocarbons in
ocean water. In Analytical Methods in Oceanography,
Advances in Chemistry Series No. 147, American Chem.
Soc ', Washington, D.C.
Butler, J. N.' B. F. Morris, and J. Sass. 1973. Pelagic tar from
Bermuda and the Sargasso Sea. Bermuda Biological
Station Special Publ. No. 10.

Where the Argo Merchant Oil Didn't Go

Chris W. Brown, Patricia F. Lynch, and Mark Ahmadjian

Department of Chemistry
University of Rhode Island
Kingston, Rhode Island

Abstract Table 1.
Between 60 and 90 days after the grounding of the Information on Tar Balls
Argo Merchant, tar balls were found on shore along
Jamestown (RI), Martha's Vineyard, Nantucket, and on Tar Ball No. Site -Date Comment
Cape Cod from Nauset Beach to Provincetown. Twenty- 1 Jamestown, R.I. 2/9/77 a,-1 lb., composite
two of the tar balls were analyzed in our laboratory. 2 Jamestown, R.I. 2/9/77 a,-1 lb., outside
Infrared spectra of the tar balls were compared with each 3 Jamestown, R.I. 2/9/77 a,-1 lb., inside
other, with the cargo from the Argo Merchant, and with the 4 Jamestown, R.I. 2/9/77 b,-5 lb., inside
oil loaded onto the lost tanker, the Grand Zenith. The 5 Jamestown, R.I. 2/9/77 c,-15 lb., inside
fingerprints of many of the tar balls were very similar to 6 Martha's Vineyard 2/12/77
each other; however, they were completely different from 7 Nantucket 2/15/77
the fingerprint of the Argo Merchant cargo. There was r to
some similarity between the fingerprints of the tar balls 8 Nantucket 3/1/77
and the neat oil loaded onto the Grand Zenith. 9 Nantucket 3/1/77
If these tar balls came from either theArgo Merchant 10 Nantucket 3/1/77
or the Grand Zenith, they were in the ocean for over one 11 Nauset Beach 3/1/77
month and the fingerprints could have changed due to
"weathering". Thus, we artificially weathered the two neat 12 Nauset Beach No. 11 weath. at GSO
oils under similar temperature conditions in flowing sea 13 Marconi Beach 3/1/77
water at the URI aquarium. Weathered samples were 14 Marconi/Lecount 3/14/77
collected periodically for one month and their fingerprints 15 Race Pt. Beach 2/10/77 Sewage
compared to those of the tar balls. 16 Race Pt. Light 3/1/77
The probabilities- that the tar balls came from the 17 1000 Vds. from 3/1/77
same source, that they came from the Argo Merchant, Race Pt. Light
that they came from the Grand Zenith, and that they came 18 Long Pt. 3/12/77
from some other source have been determined and are 19 Provincetown 3/15/77
discussed in the report. Wharf (a)
20 Provincetown 3/15177
Wharf (b)
Introduction 21 Pilgrim Beach (a) 3/12/77
22 Pilgrim Beach (b) 3112/77
During February and March of 1977, tar balls were
found along the New England coast from Jamestown,
Rhode Island, to Provincetown, Massachusetts, including
the islands of Martha's Vineyard and Nantucket. Con-
sidering the location of theArgo Merchant grounding and 30 tar balls collected along the Massachusetts shores.
the time frame, i.e., approximately two months after the Seventeen of these were from Nantucket. We randomly
tanker broke apart on Nantucket Shoals, there was selected and analyzed 4of theseplus 12 from othersites.
considerable speculation that the stricken tankerwas the Furthermore, we artificially weathered one of the tar balls
source of the tar balls. to study the effects of additional weathering.
Our first involvement with the tar balls was on Informational data on the tar balls are given in Table
February 9 when we were notified that oil had come 1 and the sites are located in Figure 1. The Nantucket
ashore in Jamestown (Rhode Island). We collected 12 collection dates are approximate, since the samples
tar balls ranging in weight from a few ounces to 15 lbs. were collected by the Coast Guard and stored at Woods
During the next two months, we received samples of Hole until April.

TO
MASS. TO 18- 2 %
CAK COD NEW ENGLAND
SAY OIL SPILLS
%
%
TBI

CAPE COD
M
TO It

NAWUCKET
jj
SOUND ",10 Ar

TOI-5 ,%I
TO 6
1444
TO 7-10 4V
V
QU
Ilk- %

Figure 1. Map of southern New England with locations of tar balls.

Experimental give a single value for estimating the diff erences between
Most of the tar balls contained water, sand and spectra (Brown et al., 1976a). This value is obtained
from the sum of the squares of the differences between
marine debris; thus, all samples were pretreated prior to the log-ratios and the average log-ratio, i.e.,
analysis. Approximately 5 ml of each sample were
centrifuged for 2 minutes and the top oil layer transferred 18 18
to another test tube. Five MI Of CC14 were added, the S2 (109 Aj, 1,9 A-1 )2
sample shaken vigorously and centrifuged for 5 minutes. Ai2 18 i=1 Ai2
The top three-quarters of the sample was transferred to
another test tube, anhydrous MgS04 added, and it was where Ail and Ai2 are the absorbances for the ith band
centrifuged for another 5 minutes at 35'C. The latter step in spectra 1 and 2, respectively. For a perfect match, the
was repeated several times until all of the water was value of S2 would be zero; thus, the magnitude of S2
removed, and then the CC14was removed by evaporation. reflects the dissimilarity between oils.
The treated samples were placed in demountable We (Brown et al., 1976b) measured spectra of 198
AgCI infrared cells with a 0.05 mm spacer and spectra neat and 647 weathered oils (including several
were measured on a Perkin-Elmer Model 521 infrared weathered samples for each of 80 neat oils). S2 values
spectrometer. The digitized spectral data were stored in for all possible pairs of oils were calculated and placed in
a computer data file and all data analyses were performed one of two categories: same oils (neat and weathered oils
on an IBM 370/60 computer. from the same origin) and different oils (neat and
weathered oils from different origins). There were 5,534
Results and Discussion pairs with the two oils originating from the same source,
and 345,030 pairs with the two oils from different sources.
Probability of Matching Infrared Spectra of Petroleum. The pairs in each category were then ordered according
Recently, Kileen and Chen (1976) proposed a method for to increasing S2 values, and two histograms forfrequency
obtaining the probability of matching spilled oil to one or of'occurrence vs S2 in increments of 0.01 were plotted.
more suspects from infrared 'Spectra of the samples, The histograms provided distributions for pairs of
Their method is an extension of the ratio method oils from the same sourses and for pairs from different
developed in our laboratory (Ahmadiian et al., 1976). sources. Killeen and Chien (1976) used these distributions
In the ratio method, absorbances at 18 frequencies in to determine the probabilities of guilt for each suspect
the spectrum of one sample are ratioed with absorbances in a spill case and the probability that a sample from the
at the same frequencies in the spectrum of another "true spiller" was not included. Their method is based
sample. The log of each ratio, the average log-ratio, an 'd on Bayes Theorem and is described completely in their
the differences between each log-ratio and the average report.
log-ratio are determined. Initially, we ratioed the absor- .In the present report it is important to note that the
bances in the spectrum of each suspect to those in the distributions used to obtain the probabilities are based on
spectrum of the spill sample; the best match was all types of oils, i.e., they are comprised of light through
assigned to the spectrum having the most ratios within heavy crudes, fuels and lubricating oils. This tends to
10% of the average. Later, the method was extended to give slightly higher probabilities especially at the low end

ARGO MERCHANT
CARGO a

Ui
0
Z

(n b

Ile
0

W
0
Z

ARGO MERCHANT
U) SPILLED OIL
Z
< 1400 900
FREOUENCY, CM-'
Figure 3. Infrared spectra of tar balls., a, Jamestown, and b,
Martha's Vineyard.

1 12131415 7 IS 19 110111 12 13 1411511611T 18 11 91?OJ21 @2
IN/
2
3
1200 1000 860 600 H+ I
-L WII
WAVENUMBER, (CM-1)
7
Figure 2. Infrared spectra of the Argo Merchant cargo and of the6 4+ I-HI I
spilled oil collected 2 days after the spill. 9 1 N/ I-H
10, 1 M+
of the 0 to 1 probability scale. For example, infrared I I - - - - - -
fingerprints of two No. 6 fuels may be completely differentI -WV
and we would expect that the probability that they match 13
would be close to 0.0. However, since they are both No. 61
fuels, their fingerprints will be more similar than the finger- 15
prints of a No. 6 and a No. 2 fuel. Thus, in many cases, the 16
probability that two different No. 6 oils match will be higher 17
than expected because they are the same type of oil. ... I I .
Argo MerchantSpill. Infrared spectra of theArgo Merchant
cargo and of the spilled oil collected two days after the 21
tanker broke apart are shown in Figure 2. (These samples 22 - - - LLLLL-
were obtained from Dr. Jerome Milgram, MIT.) When the
digitized spectrum of the spilled oil was compared with >0.99
that of the cargo, the following probabilities were 0-29
obtained: 05-95
Probability .70-.05
30--.70
2

9
20

Argo Merchant 0.986
Another Source 0.014 Figure 4. Probabilities of tar balls matching.

Table 2.
Probabilities Of Tar Balls Matching
(In % Units)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

1 - 66 96 96 97 99 95 97 97 99 46 55 43 20 0 86 66 16 89 16 22 18
2 - 76 70 67 79 77 77 71 71 45 56 41 33 0 56 51 16 58 18 19 20
3 - 96 100 97 96 99 97 98 38 45 36 18 0 90 77 20 91 17 19 18
4 - 100 98 93 100 97 99 33 41 31 18 0 87 83 17 90 16 18 17
5 - 97 93 99 98 100 35 42 34 18 0 90 80 20 91 14 18 18
6 - 98 98 97 99 47 60 44 22 0 84 76 17 89 18 22 18
7 - 95 98 95 55 72 53 26 0 74 73 16 82 18 28 19
8 - 96 99 38 48 35 19 0 84 84 19 84 16 17 18
9 - 97 45 50 42 21 0 91 63 16 73 17 23 18
10 - 40 48 38 19 0 91 76 18 86 16 20 18
11 - 94 99 81 0 33 23 21 33 34 55 59
12 - 94 57 0 31 37 18 44 32 39 36
13 - 80 0 31 23 19 33 32 54 48
14 - 0 18 18 21 18 44 36 71
15 - 0 0 0 0 0 0 0
16 - 25 16 58 15 20 19
17 - 0 70 16 17 17
18 - 22 19 19 43
19 - 15 18 17
20 - 23 55
21 - 44
22

The deviation from 1.0 is possibly due to weathering of the Table 3.
spilled oil. ProbabilitV of Tar Balls Originating From Argo Merchant Cargo
Intercomparison of Tar Balls. Infrared spectra of the 22 tar (In % Units)
balls listed in Table 1, were measured and the digitized
fingerprints stored in a computer data file. Many of the Argo Merchant Cargo
spectra had similar contours and some were almost Tar Ball Neat 3 Days- TO Days
identical as is shown in Figure 3. The spectrum of each Weathered Weathered
tar ball was compared with that of each of the others and
the probability that they came from the same source is 1 22 18 18
given in Table 2. The results are categorized according 2 18 25 18
to magnitude in Figure 4. 3 21 5 18
All samples from Jamestown, Martha's Vineyard and 4 19 0 18
Nantucket have probabilities >0.85 of being identical 5 20 0 18
except for sample 2 from Jamestown. This was a sample 6 20 7 is
of the outside layer of a tar ball, and the differences 7 19 19 18
reflect excessive weathering on the surface. In addition 8 20 22 18
to samples 1 -10 being from the same source, the proba- 9 21 23 18
bilities that samples 12, 16, 17 and 19 came from 10 21 20 18
this source are >0.5. Furthermore, tar balls 11 and 13 11 32 18 32
are virtually identical (P = 0.99). 12 23 18 27
it should be mentioned that the differences between 13 27 19 27
many of the tar balls could be due to weathering and that 14 18 18 18
this effect is more pronounced on the surface of the tar 15 0 0 0
balls. Many of the tar balls from Cape Cod were very 16 21 0 18
small and these could have been subjected to extensive 17 15 17 16
weathering. 18 18 16 18
19 18 25 18
Tar Balls and the Argo Merchant Cargo. The infrared 20 is 28 21
spectral fingerprint of the Argo Merchant oil is compared 21 19 18 22
to that of the Martha's Vineyard tar ball in Figure 5. The 22 22 18 28
general contours of the spectra are entirely different. We
treated each of the 22 tar balls as a spill sample and
determined the probability that each came from the

TAR BALL FROM
MARTHA'S VINEYARD a

W
0

U)
Z

Ir c

ARGO MERCHANT
CARGO

1400 900
I FREQUENCY, CM-'
Figure 6. Infrared spectra: a, Argo Merchant cargo, b, Grand
12bO 1000 800 0 Zenith oil, and c, Martha's Vineyard tar ball.
WAVENUMBER (CMI
Figure 5. Infrared spectra of tar ball from Martha's Vineyard also a possible source of the tar balls. The U.S. Coast
and Argo Merchant cargo. Guard R & D Center supplied us with a sample of the oil
loaded aboard the Grand Zenith. We measured infrared
spectra of the neat oil and of samples collected
periodically during one month of weathering at the URI
cargo. The results given in Table 3 show that only sample aquarium. The infrared spectrum of the Grand Zenith oil
No. 11 (Nauset Beach) had a probability of matching >0.3; is compared with the Argo Merchant cargo and the
most were <0.2. The spectrum of the cargo and the tar Martha's Vineyard tar ball in Figure 6.
balls (except for sample 15) were characteristic of No. 6 The probabilities obtained when comparing the
fuel oils. Thus, these finite probabilities reflect the fact spectra of the neat and 7 weathered samples to the tar
that the samples are the same type of oil. balls are given in Tables 4a and 4b. The probabilities
If the tar balls came from the Argo Merchant, they run as high as 0.83 when the neat oil is compared to tar
were "weathered" in the Atlantic for almost 2 months; balls Nos. 3 and 4; however, the highest Orobability (90%)
thus, we weathered some of the cargo oil at the URI was obtained from matching the 14 day weathered
aquarium for one month and periodically analyzed sample with tar ball No. 11 from Nauset Beach (see
samples. In most cases, the spectra of the tar balls and Figure 7 for spectra). Other tar balls found in this area
the weathered Argo Merchant tar balls became less (Nos. 13 and 14) have probabilities of -0.85 of matching
similar. The probabilities of matching forthe 3 and 10 day the weathered Grand Zenith oil.
samples are also given in Table 3. According to the probabilities given in Table 4, the
7@
77;@

Tar Balls and the Grand Zenith Oil. The tanker Grand tar balls can generally be placed into three categories:
Zenith disappeared somewhere off the New England coast i) those with high probabilities of matching the neat or
on the way from Teeside, England to Fall River, short term weathered samples, ii) those with high proba-
Massachusetts during January 1977; thus, its cargo was bilities of matching the long term weathered samples, and

Table 4a.
Probability of Tar Balls Originating From Grand Zenith Cargo
(In % Units)

Grand Zenith Cargo Weathered (Days)
Tar Ball 0 1 2 4

1 74 77 75 51
2 45 45 52 55
3 83 87 86 66
4 83 87 88 69 Uj
5 82 88 86 63 L) b
z
6 73 75 76 57
7 60 65 70 52
8 76 81 81 58
9 69 73 73 44
10 79 83 83 55
11 36 35 38 31 F-
12 38 38 42 41
13 32 33 35 30
14 18 18 18 18
15 0 0 0 0
16 75 78 77 40
17 33 37 40 44
18 14 16 16 20
19 71 75 74 81
20 16 16 16 15
21 18 is 20 19 _j
22 18 18 18 18
1400 FREQUENCY, CM 900

Figure 7. Infrared spectra: a, tar ball from Nauset Beach, and
Table Alb. b, Grand Zenith oil weathered 14 days.
Probability of Tar Balls Originating From Grand Zenith Cargo
(in % Units)

Grand Zenith Cargo Weathered (Days) iii) those with low probability matches. Samples 1, 3-10,
Tar Balls 7 14 21 28 16 and 19 fall in category i, samples 11 -14, 21 and 22 in
category ii, and the rest in iii.
1 73 55 26 27
2 58 47 33 34 Conclusions
3 77 50 22 24
4 75 49 20 21 The present study showed conclusively that the tar
5 76 47 20 21 balls found along the New England coast after the Argo
6 77 58 29 29 Merchant incident were not from the stricken tanker. Many
7 70 63 35 34 of the tar balls had similar infrared spectra and there is a
8 72 47 23 23 high probability that a number of these came from the
9 74 61 29 30 same source. Finally, there is a reasonably high proba-
10 78 54 25 26 bility that some of the tar balls originated from the stricken
11 55 90 86 87 tanker Grand Zenith. It should be noted that the tar balls
12 51 73 64 58 were compared with the oil loaded onto the Grand Zenith,
13 47 84 81 83 and not to the oil actually contained in the tanker. The
14 23 41 87 77 composition of the oil in the tanker could have been
15 0 0 0 0 slightly different due to residues in the tanker from
16 85 58 23 28 previous shipments.
17 32 24 18 18
18 17 19 18 21
19 59 37 18 20 Acknowledgments
20 18 22 37 30
21 27 46 61 63 This research was supported by the Environmental
22 22 40 58 64 Control Division, Department of Energy (Ccintract No.
E(l 1-1)-4047) and NOAA Contract 03-7-022-35123.

References
Ahmadjian, M., Baer, C.D., Lynch, P. F. and Brown, C. W., 1976
"Infrared Spectra of Petroleum Weathered Naturally and
Under Simulated Conditions", Environ. Sci. Technol. 8,
777-781.
Brown, C, W., Lynch, P. F. and Ahmadjian, M., 1976-a. "Infrared
Spectra of Petroleum - Data Base Formation and Applica-
tion to Real Spills", Proceedings, Workshop on Pattern
Recognition Applied to Oil Identification, Coronado, CA
(Nov. 11 -12), IEEE Cat. No. 76CHI 247-6C, 84-96.
Brown, C. W., Lynch, P.F. and Ahmadiian, M., 1975-b. "Identifica-
tion of Oil Slicks by Infrared Spectroscopy", U.S. Coast
Guard, Washington, Report No. CG-D-19-77.
Killeen, T. J. and Chien, Y T., 1976 "A Probability Model for
Matching Supects with Spills - or Did the Real Spitler Get
Away?" Proceedings, Workshop or) Pattern Recognition
Applied to Oil Identification, Coronado, CA (Nov, 11-12),
IEEE Cat. No. 76CH 1 247-6C, 66-72

Hydrocarbon Patterns in Some Marine
Biota and Sediments Following the Argo
Merchant Spill

William D. MacLeod, Jr., Marianne Y. Uyeda, Lawrence C. Thomas, and Donald W. Brown

NOAA National Analytical Facility
Northwest and Alaska Fisheries Center
2725 Montlake Boulevard East
Seattle, Washington

Abstract and aromatic hydrocarbons to elicit evidence of con-
Over 60 samples of marine biota or sediments tamination by the spilled Argo Merchant cargo. A sample
of cargooil collected from the wreck by Prof. J. Milgramof
collected in response to the Argo Merchant oil spill were M.I.T. (Grose and Mattson, 1977) served as the reference.
analyzed for hydrocarbons by glass capillary gas chroma- Petroleum and petroleum-based fuel oils contain a
tography (GC). High resolution GC patterns of the satu- broad range of compounds, mostly hydrocarbons. When
rated hydrocarbons extracted from the samples were such complex mixtures are released into the marine
compared with the corresponding pattern from the Argo environment, they are subject to a number of physical,
Merchant cargo. The stomach contents from one cod chemical, and biological processes known collectively
sample and one windowpane flounder sample gave GO as "weathering". Depending on many factors, these
patterns which compared closely to the cargo pattern. processes can alter the composition of the mixture of
Analogous comparisons of th e aromatic hydrocarbons compounds (distribution pattern) constituting a spilled oil,
confirmed only the correlation for the sample from the frequently beyond recognition in terms of the original
cod. When GC patterns failed to match closely, objective pattern (Clark and MacLeod, 1977). To some extentthese
pattern matching procedures were devised to rank the effects can be compensated for (Bentz, 1976), but they
patterns of the major sample hydrocarbons for their need to be understood much better before compositional
similarity to the reference pattern. Similarity between data on hydrocarbons from environmental samples can
sample and reference relative abundance patterns was be interpreted unambiguously with respect to sources of
expressed as the percent of the sample hydrocarbons oil contamination (Farrington and Medeiros, 1975; Malins,
(among those measured) whose individual relative abun- 1977). Despite such reservations, it is still possible to
dance fit the reference relative abundance within a derive inferences linking a suspected source of con-
specified tolerance. Calculations were based on the tamination with a sample extract, if finely discriminating
assumption of either (a) no weathering effects on the spectral or chromatographic characteristics of their
hydrocarbon distribution patterns, or (b) significant, but respective hydrocarbon fractions agree, As one example,
indeterminate, alteration of the distribution of the more Adlard et al. (11972) have demonstrated that high resolu-
volatile and water-soluble hydrocarbons. tion gas chromatography (GC) can provide finely detailed
characterizations of hydrocarbon fractions to differen-
Introduction tiate oils.
Establishing a relationship between hydrocarbons
The grounding of the tanker Argo Merchant on from an environmental sample and a suspected source of
Nantucket Shoals in December 1976, and its subsequent oil contamination becomes increasingly difficult as
breakup with a total loss of 7.7 million gallons of No. 6 environmental processes alter the distribution pattern of
fuel oil cargo, set in motion a number of scientific hydrocarbons from environmental samples. Since this is
activities noted by Grose and Mattson (11977). Their report a common problem, it is important to investigate detailed
describes sampling cruises by the National Marine pattern matching procedures for a better understanding
Fisheries Service (NMFS) and by the University of Rhode of the fate of hydrocarbons introduced into the marine
Island (URI) which procured samples of marine biota and environment, including their transformations throughout
sediments for the analyses discussed here. We have the food web. Oil pattern matching methodology based
analyzed more than 60 of these samples for saturated on comparisons of spectral or chromatographic proper-

70* 68* 66*

DELAWARE I ICRUISES
42' 42' ENDEAVOR 004 SEDIMENTS
a76-13, DEC. 22-23,1976
3.
% 4 -77-01, JAN. 4 -10,1977 AM-0 ARGO MERCHANT BOW
R
--- OIL SLICK BORDE AM-S ARGO MERCHANT STERN
6 ARGO MERCHANT OL
..'e, 10
0 CLEAR BOTTOM
31 - - - - - - - - - - - - - - - - - - - - - - A-34 A-33
C
- A IS -43b 0/
1 ..39 M
38
-------- O_,6 4t'
29
%,40 'A'40-4 01
24. t,, D-35 O-C-23
-7-0
0--G-42

-F-28
40* 40* 0-15-0
70* 68* 66*

Figure 1. Collection locations for samples of marine biota taken 69'
during cruises of the RIV Delaware // (Grose and Mattson, 1977). Figure 2. Collection locations for samples of sediments taken
during cruise 004 of the R/V Endeavor (Grose and Mattson, 1977).
ties has been reviewed by Bentz (11976). Recent develop-
ments have been reported from the workshop convened son, 1977). Samples of marine sediments in the vicinity
by Chien and Killeen (1976). Interestingly, little of this of the wreck were collected by URI on the cruise of the
work has exploited the power of high resolution (i.e., R/V Endeavor shown in Figure2 (Groseand Mattson, 1977).
capillary) GC, in spite of the superior separations demon- All samples were frozen until analyzed. In general,
strated on complex hydrocarbon mixtures a decade ago samples of biota were composited from three or more
(MacLeod, 1968; Modzeleski et al., 1968) or subse- individuals of a species. These samples were homoge-
quently (Adlard et al., 1972). nized, digested with alkali, solvent extracted, chromato-
In early oil pattern matching work utilizing a modified graphed on silica gel, analyzed by glass capillary GC, and
GC capillary, Zaf iriou (1973) employed a support-coated confirmed by mass spectrometry (MS) according to
open-tubular (SCOT) column to obtain abundance data published procedures (MacLeod et al., 1976). Sediment
on four alkane descriptors in his procedure for comparing samples also were analyzed according to these proce-
oil-rich samples. Clark and Jurs (1975) incorporated dures, except that they were extracted twice by methanol
additional descriptors from Zafiriou's chromatograms to and three times by methylene chloride/methanol (2/1)
test computerized pattern matching procedures for according to Brown et al. (1978). Hexamethylberizene
alkanes. The SCOT column is also used in Flanigan's was used as the GC internal standard (IS). Previous
(1976) pattern matching routines. studies have shown that losses of individual hydro-
Now that more high resolving glass capillary GC carbons are generally less than 30% in sample workup,
columns and related chromatographs are available and that relative standard deviations of ca. 20% can be
commercially, detailed compositional characterizations attained for abundance measurements for most hydro-
of complex hydrocarbon mixtures have become suffi- carbons (MacLeod et al., 1976).
ciently routine that these analyses can be readily auto- Gas chromatograms of the saturated and aromatic
mated (MacLeod et al., 1976, 1977). Hence, it is timely hydrocarbons from the samples were compared with the
that the wealth of compositional information obtainable corresponding reference chromatogram from the Argo
by glass capillary GC receives due consideration in oil Merchant cargo. Abundance levels were calculated for 24
matching research. major alkanes in the saturated hydrocarbon fraction. This
In this study, detailed GC patterns of the saturated included the n-alkanes from n-C 1 OH22 to n -C31 H64, Plus
and aromatic hydrocarbon fractions obtained from pristane (2,6,10,14-tetramethylpentadecane) and phy-
environmental samples were compared with reference tane (2,6,10,14-tetramethylhexadecane). Abundance
GC patterns from the Argo Merchant cargo. If the sample levels were also calculated for 18-21 major arenes
and reference high resolution GC patterns visually in the aromatic hydrocarbon fraction. After correcting for
corresponded closely in the fine details, a probable analytical background levels, the alkane and arene abun-
match was indicated. When sample and reference high dances were normalized relative to tetracosane
resolution GC patterns differed sufficiently to eliminate (n-C24H50) and phenanthrene (C14H10) abundances,
the possibility of an obvious match, the excellent resolu- respectively. The resulting major alkane or arene relative
tion still assured confidence in major hydrocarbon abun- abundance profile for a sample constitutes the relative
dance measurements and their relative freedom from distribution pattern of these major hydrocarbons from
interferences. Under these circumstances, we employed that sample.
two simple arithmetical procedures to rank the major The major hydrocarbon distribution patterns from
hydrocarbon distribution patterns according to their the samples were ranked for their similarity to the
similarity to the Argo Merchant reference pattern. reference. This was expressed as the percent of com-
pounds used in.the distribution pattern (excluding the
Methods and Materials normalizing compound) whose individual relative abun-
dance fell within a �1/3 tolerance of the corresponding
Samples of marine biota were collected during two reference relative abundance. In addition, the major
I --- 4_7 "@

cruises of the R/V Delaware I/ (Figure 1) using standard alkanes from the biota were analogously compared with
NMFS groundfish survey procedures (Grose and Matt- the cargo alkanes using 90% confidence intervals cal-

W 1 2 25

PRIS E
14 26

PHYT NE

is 29

Figure 3. High resolution gas chromatogram of saturated hydrocarbons extracted from a sample of the Argo Merchant cargo collectedby
Prof. J. Milgram (Grose and Mattson, 1977). Numbers denote normal alkane carbon chain lengths. 20mxO.25mm WCOT glass column
coated with SE-30. 2,'pilsplitless injection, split(10:1) after 12 seconds with 14psiheliumcarriergas. Isothermal forfirst5 minutes at40'C,
programmed from 40' to 2700C at a rate of 40CIminute. Internal standard (IS): hexamethylbenzene.

24
STANE
15 1 17 1 25

14 26

13 @"@A 4E 27

is
@ 4 A@4 @4@4 30
10

Figure 4. High resolution gas chromatogram of saturated hydrocarbons extractedfrom the stomach contents of codcollected on a cruise
of the RIV Delaware 11 (DE 77-01, station 29). Numbers denote normal alkane carbon chain lengths. Conditions same as for Figure 3.

X
36

30
11-J 11,

Figure 5. High resolution gas chroma togram of aroma tic hydrocarbons extracted from a sample of the Argo Merchant cargo collected by
'Prof J Milgram (Grose and Mattson, 1977). Conditions same as for Figure 3. Identified compounds confirmed by mass spectrometry-@

Figure 6. High resolution gas chromatogram of aromatic hydrocarbons extracted from the stomach contents of cod collected on a cruise
of the RIV Dela ware I/ (DE 77-01, station 29). Conditions same as for Figure 3. Identified compounds confirmed by mass spectrometry.

24
25
PRISTANE
26
22

21 ZT
2
29

19
29
Is
11

30
12 14 ME 31

Figure 7. High resolution gas chromatogram of saturated hydrocarbons extracted from the stomach contents of windowpane flounder
collected on a cruise of the RIV Delaware /I (DE 76-13, station 4). Numbers denote normal alkane carbon chain lengths. Conditions
same as for Figure 3.

is
17

PRISTANE 27
23

15 22
29
16
18 Is 21
14 30

13 31
12 kP"
3

13
2

Figure 8. High resolution gas chromatogram of saturated hydrocarbons extracted from the stomach contents of codco//ected on a cruise
of the RIV Delaware I/ (DE 77-01, station 38). Numbers denote normal alkane carbon chain lengths. Conditions same as for Figure 3.

23
25

22
26

21 27

is 20

29
20
30
PRISTA: P MNE 31
16 1@7 / _L a

Figure 9. High resolution gas chromatogram of saturated hydrocarbons extracted from the stomach contents of silverhake collected on a
cruise of the RIV Delaware ll(DE77-01, station 24). Numbers denote normalalkane carbon chain lengths. Conditions same as forl`igure3.

culated from quadruplicate analyses of the reference oil spilled Argo Merchant cargo, the evidence is not as
via the analytical procedure for the biota. strong as in the previous case.
The high resolution GO pattern of the saturated
hydrocarbons from another sampling of cod stomach
Results and Discussion contents (cruise DE 77-01, station 38) gave major alkane
High Resolution Gas Chromatography. In our analyses of levels comparable to the windowpane flounder above, but
the hydrocarbons from marine biota and sediments its resemblance to the reference GO pattern (Figure 8 vs.
collected in response to the Argo Merchant oil spill Figure 3) was poorer. The aromatic hydrocarbon GC re-
(Grose and Mattson, 1977), the most direct evidence of sponse was too weak for pattern comparison. In another
contamination by the Argo Merchant cargo was found in case of interest, the stomach contents of a silver hake
the GO patterns (gas chromatograms) of the hydro- sample (cruise DE 77-01, station 24) showed high levels
carbons from three samples of fish stomach contents. (10 ppm) of certain residual paraff ins in its GO pattern
The hydrocarbons from the stomach contents of the cod (Figure 9), although it is not clear how they could be re-
collected on cruise DE 77-01, station 29i correlated best lated to the Argo Merchant cargo. Six of the major cargo
with the hydrocarbons from the cargo reference sample. arenes were present at ppm levels, but the aromatic hydro-
Figure 3 shows the high resolution GO pattern of the carbon GO pattern bore no resemblance to Figure 5. None
saturated hydrocarbons from the reference. The corre- of the remaining 33 samples of biota or 25 samples of
sponding GO pattern from the cod stomach contents is sediment gave high resolution GO patterns of their satu-
shown in Figure 4. Both finely detailed GO patterns show rated or aromatic hydrocarbons which could be visually
extensive qualitative and semi-quantitative agreement in related to Figure 3 or Figure 5, respectively.
their distribution of compounds. A similar correspon- Major Hydrocarbon Pattern Ranking. As demonstrated
dence was found between the aromatic hydrocarbons of above, the high resolution gas chromatograms of the
this sample and the reference. Figure 5 shows the high saturated and aromatic hydrocarbon fractions from
resolution GO pattern of the Argo Merchant aromatic environmental samples can contain a wealth of informa-
hydrocarbons; its counterpart from the cod stomach tion useful for visually documenting detail by detail com-
contents appears in Figure 6. Both saturated and aromatic parisons with a known reference source of oil contamina-
hydrocarbons compare so well with those of the cargo tion. However, when the GO patterns under comparison
sample that it would be difficult to refute the contention fail to correspond well in a logical way with regard to
that the stomach contents of cod from cruise DE 77-01, conceivable environmental alterations (e.g. weathering),
station 29, had been contaminated with oil from the Argo interpretations on the basis of these visual, subjective
Merchant spill, even though it apparently was collected comparisons become difficult to articulate meaningfully.
outside the perimeter of the surface slick (Figure 1). In view of these uncertainties, some sort of objective
The saturated hydrocarbons from the stomach con- scale for ranking these less definitive comparisons might
tents of a sample of windowpane flounder collected on prove useful. In pursuing this issue in an elementary way,
cruise DE 76-13, station 4, gave a GO pattern (Figure 7) we elected to concentrate on sets of major hydrocarbon
which also corresponded well with the reference GO compounds since they can be measured readily and with
pattern (Figure 3). However, the level of individual major confidence by high resolution glass capillary GO. The
alkanes (10-100 ppb) was much lowerthan that of the cod major hydrocarbon abundance data so obtained was.
stomach (10-30 ppm) discussed above, and few of the converted to a common relative scale by normalizing this
major arenes could be detected and measured with data with respect to a "reliable" member of the set. The
certainty. Thus, while it would appear that the window- resulting set of relative abundances of the major alkanes
pane flounder could also have ingested some of the or arenes (listed according to position in the GO pattern)

20 2,0

16 16

1 2

08 06

04 04

0 6 8 30
10 12 14 r. P ISP222 24 22
10 12 4 r P 18 P20 22 24 26 26 30 7 .
R Y
IY S
S

Figure 10. Relative abundances of pristane, phytane, and Figure 11. Relative abundances of pristane, phytane, and
n-alkanes, normalized to the abundance of tetracosane in an n-alkanes, normalized to the abundance of tetracosane in an
extract of a sample from the Argo Merchant cargo. i indicates extract of a sample from the Argo Merchant cargo. i indicates
upper and lower bounds of �1/3 tolerance used for comparing upper and lower bounds of 90% confidence intervals used for
sample relative abundances to this pattern. comparing sample relative abundances to this pattern.

constitutes a distribution pattern of the set. reference relative abundance value based on the
The choice of a reliable normalizing compound was following considerations:
relatively straightforward in the case of the major alkanes. (a) The relative standard deviation of the analytical
Tetracosane (n-C24H50) is within the molecular weight procedure (MacLeod et al., 1976) averages �15 to
range for low susceptibility to evaporation and dissolution �20% for most environmental samples.
relative to more volatile alkanes, especially under the (b) In comparing the results of two analyses (e.g.,
wintry conditions of the Argo Merchant oil spill. Being an sample vs. reference alkanes), the overall uncer-
even-carbon numbered n-alkane, it is less likely to have tainty contribution is roughly additive, or �30 to
been contributed from immediate biological sources. �40%.
Finally, an examination of the high resolution gas chroma- (c) A tolerance span of �1/3 approximates �30 to
trograms revealed that it appeared to be free of inter- �40%.
ference by adjoining peaks. This would not be true for Application of this pattern ranking procedure to the
octacosane which may be merged with a branched or, sample arenes ended when it became obvious that the
cyclic hydrocarbon (MacLeod et al., 1976, 1977). measurable levels of the gas chromatographable arends
Phenanthrene was chosen for normalization among the found in most of the samples were insufficient for useful
arenes because it was the least volatile of the major comparisons. The remaining discussion therefore
arenes found in theArgo Merchant cargo; the others were centers on the major alkanes found in the cargo, viz., the
too volatile and water-soluble to merit serious n-alkanes from C1 OtO C31, plus the isoprenoidal alkanes,
consideration. pristane and phytane. The reference relative abundance
The distribution patterns of the alkane and arene pattern from the Argo Merchant cargo sample with its � 1/3
relative abundances were plotted in bar-graph form for tolerances is shown in Figure 10.
easy visual comparisons. Since exact fit of the relative The percentage of alkanes in the set whose relative
abundances between sample and reference was not abundances fit the reference pattern within the �1/3
expected, a tolerance span associated with the tolerance (Figure 10) is listed for the samples of biota in
reference relative abundance was employed to rank- Table 1 under "% Fit". Three of the 37 samples analyzed
order inexact fits of the sample relative abundance registered a 57-61 % fit. These results are from the
values. For simplicity, this ranking was expressed as the stomach contents of the fish which also gave the best
percentage of compounds within the set (excluding the visual GC pattern comparisons above. Two samples
normalizing compound) which were within a specified registered a 30-39% fit and 11 samples registered a
tolerance. 13-22% fit, including the heavily contaminated silver hake
Within practical limits, the order of ranking of the stomach contents (17% fit). The remaining 19 samples
sample relative abundance patterns with respect to the failed to indicate measurable alkane levels by our analyfl-
I

Argo Merchant reference pattern did not depend upon cal procedure.
the size of the tolerance spans. For example, the order of Since certain assumptions were made with the �1/3
ranking was similar whether �1/3 or �1/2 tolerances were tolerance criterion, an alternate method of setting the
used. We employed a �1/3 tolerance span around each tolerance span was investigated. To place the alkane

Table 1. Percent fit of alkane relative abundance patterns of Argo Merchant cargo vs. marine biota samples by the �113 tolerance and the
90% confidence intervals; median (chi-square test), and diversity index (t-test) results.

% Fit Median Diversity
Sample. Cruise Station 113 90% Chi-square (t-test)

Cod: stomach contents 77-01 29 57 70 + +
Haddock: flesh 77-01 8 22 13 +
Silver hake:
flesh 77-01 3 17 26
stomach contents 39 30
Yellowtail flounder:
flesh 77-01 3 22 22 +
Winter flounder:
stomach contents 77-01 3 22 13 +
Cod:
flesh 77-01 38 13 13 +
stomach contents 57 65 + +
Haddock:
stomach contents 77-01 27 30 26 +
Silver hake:
flesh 77-01 24 17 26 +
stomach contents 17 22 +
Winter flounder:
flesh 77-01 31 17 22 +
stomach contents 13 17 +
Windowpane flounder:
stomach contents 76-13 4 61 65 + +
Sea scallops 77-01 39 22 26 +
Lobster 76-13 6 13 13 +

pattern comparison on a firmer statistical basis, the were not shown to be different than those of the Argo
reference cargo oil was processed through the entire Merchant cargo (Table 1). These results are consistent,
analytical procedure for biota in quadruplicate. The 90% with the highest percent fit values obtained by the two
confidence intervals around the mean value of each tolerance criteria which provide statistical support for
alkane was established as an experimentally determined visual comparisons of the gas chromatograms discussed
tolerance span (Figure 11). above.
The cargo reference patterns and associated toler- The use of the reference patterns in Figures 10 and 11
ances produced by the two different procedures (Figures was based on the assumption of no differential weather-
10, 11) are similar but not identical. Consequently, minor ing eff ects on the sample major alkanes. Since this could
differences occurred in the percentage of alkanes in the not always be expected to be the case, the ranking of
samples which fit the reference alkanes within the given sample major alkane patterns with respect to the refer-
tolerance. Comparisons of the percent fit by the "�1/3" and ence was also calculated on the basis of the alkanes
by the "90% confidence" criteria are listed in Table 1. above n-C1 61-134 only (Figure 10) to allowforsomeof the
Despite differences of percent fit by the two criteria, more predictable effects of weathering (Clark and
the considerable similarities between the two data sets MacLeod, 1977). Interestingly, this allowance had little
are reassuring. This supports the use of the �1/3 tolerance effect on the percent fit values for the alkanes from the
as a simple way to approximate objective alkane pattern biota or on their rank-ordering.
comparisons among samples suspected of being con- Of 36 sediment samples screened for evidence of
taminated by spilled oil. On the other hand, if it is feasible Argo Merchant cargo by ultraviolet fluorescence (UVF)
to perform multiple analyses of the reference sample analysis by R. Jadamec of the U.S. Coast Guard (Grose
using the complete analytical procedure, the 90% con- and Mattson, 1977), 25 were forwarded frozen for GC and
fidence interval may provide a sounder tolerance criterion GC/MS analysis in our laboratory. Analysis of 10 sedi-
for pattern comparisons. ment samples whose estimated contamination by the
Two additional statistical operations were per- spilled oil exceeded 0.1 ppm by the UVF procedure
formed on the normalized alkane distributions from the (Grose and Mattson, 1977) failed to reveal any GC
biota vs. that of the cargo (Table 1). The first tested the patterns or major hydrocarbon relative abundance pat-
medians of the biota and cargo data sets for differences terns which could be readily related to theArgo Merchant
using a chi-square test at the 90% confidence level. The reference cargo sample. The remainder of the sediment
second tested the diversity indices of the biota and cargo samples did not afford measurable levels of major
data sets for differences using t-test tables at the 90% alkanes or arenes by our procedure.
confidence level. The medians and diversity indices of the
relative abundance patterns for the stomach contents of:
(a) Cod: cruise 77-01, station 29, Summary
(b) Cod: cruise 77-01, station 38, and
(c) Windowpane flounder: cruise 76-13, station 4 The saturated hydrocarbon fraction better demon-

strated evidence of possible contamination from theArgo Puget Sound and the Strait of Juan de Fuca. National
Merchant cargo than did the aromatic fraction. The best Oceanic and Atmospheric Administration Tech. Memo.
analytical evidence for cargo contamination was found in No. ERL MESA-8, Boulder, Colo.
the gas chromatograms of the hydrocarbons from the MacLeod, W. D., Jr., D. W. Brown, R. G. Jenkins, and L. S. Ramos.
stomach contents of two fish samples. The aromatic 1977. Intertidal sediment hydrocarbon levels at two sites
on the Strait of Juan de Fuca, p. 385-396. In D. A. Wolfe
fraction afforded less evidence of contamination by the (ed.), Fate and Effects of Petroleum Hydrocarbons in
cargo oil than did the saturated fraction, primarily Marine Organisms and Ecosystems. Pergamon, New York,
because the major cargo reference arenes failed to give N.Y.
measurable levels in most samples. Malins, D. C. 1977. p. 454-456. In D. A. Wolfe (ed.), Fate and
Effects of Petroleum Hydrocarbons in Marine Organisms
and Ecosystems. Pergamon, New York, N.Y.
Acknowledgments Modzeleski, V. E., W. D. MacLeod, Jr., and B. Nagy. 1968. A com-
bined gas chromatographic-mass spectrometric methods
This study was supported by funds from the Admin- for identifying n- and branched-chain alkanes in sedimen-
istrator's reserve of the National Oceanic and Atmos- tary rocks. Anal. Chem. 40:987-989.
pheric Administration and by the Bureau of Land Manage- Zafiriou, 0. C. 1973. Improved method for characterizing
ment. L. Scott Ramos, Patty G. Prohaska, and Joseph L. environmental hydrocarbons by gas chromatography.
Schwahn prepared the samples for analysis. Debra L. Anal. Chem. 45:952-956.
Safronek assisted in processing the analytical data.
Robert C. Clark, Jr., offered helpful comments, as did Prof.
James G. Quinn of URI, and Dr. John W. Farrington of
Woods Hole Oceanographic Institution.

References

Adlard, E. R., L. F. Creaser, and P. H. D. Matthews. 1972. Identifi-
cation of hydrocarbon pollutants on seas and beaches by
gas chromatography. Anal. Chem. 44:64-73.
Bentz, A. P. 1976. Oil spill identification. Anal. Chem. 48:454A-
472A.
Brown, D. W., L. S. Ramos, A. J. Friedman, and W. D. MacLeod, Jr.
1978. Analysis of trace levels of petroleum hydrocarbons
in marine sediments using a solvent/slurry extraction pro-
cedure. Proceedings of the 9th Materials Research Sym-
posium on Trace Organic Analysis: A New Frontier in
Analytical Chemistry. National Bureau of Standards,
Washington, D.C.
Chien, Y. T., and T. J. Killeen. 1976. Workshop on Pattern Recog-
nition Applied to Oil Identification. Institute of Electrical
and Electronic Engineers No. 76CH1247-6C, New York,
N.Y.
Clark, H. A., and P. C. Jurs. 1975. Qualitative determination of
petroleum sample type from gas chromatograms using
pattern recognition techniques. Anal. Chem. 47:374-378.
Clark, R. C., Jr., and W. D. MacLeod, Jr. 1977. Inputs, transport
mechanisms, and observed concentrations of petroleum
in the marine environment, p. 91-223. In D.C. Malins (ed.),
Effects of Petroleum on Arctic and Subarctic Marine Environ-
ments and Organisms. Academic Press, New York, N.Y.
Farrington, J. W., and G. C. Medeiros. 1975. Evaluation of some
methods of analysis for petroleum hydrocarbons in marine
organisms, p. 115-121. In Conference on Prevention and
Control of Oil Pollution. American Petroleum Institute,
Washington, D.C.
Flanigan, G. A. 1976. Ratioinq methods applied to gas chromato-
graphic data for oil identification, p. 162-173. In Y. T. Chien
and T. J. Killeen (eds.), Workshop on Pattern Recognition
Applied to Oil Identification. Institute of Electrical and
Electronics Engineers No. 76CH1 247-6C, New York, N.Y.
Grose, P. L., and J. S. Mattson. 1977. TheArgo Merchant Oil Spill,
a Preliminary Scientific Report. National Oceanic and
Atmospheric Administration, Environmental Research
Laboratories, Boulder, Colo.
MacLeod, W. D., Jr. 1968. Combined gas chromatography-mass
spectrometry of complex hydrocarbon trace residues in
sediments. J. Gas Chromatog. 6:591-594.
MacLeod, W. D., Jr., D. W. Brown, R. G. Jenkins, L. S. Ramos, and
V. D. Henry. 1976. A Pilot Study on the Design of a Petro-
leum Hydrocarbon Baseline Investigation for Northern

A Comparison of Argo Merchant Oil
and Sediment Hydrocarbons from
Nantucket Shoals

Eva J. Hoffman and James G. Quinn

Graduate School of Oceanography
University of Rhode Island
Kingston, Rhode Island

Abstract the Argo Merchant cargo oil. The hydrocarbons in the
Surface sediment samples collected from the tar particles, and those sediments collected in February
Nantucket Shoals Argo Merchant wreck site area in in the immediate vicinity of the wreck sites, matched well
February, 1977, were analyzed for hydrocarbon content with the Argo Merchant cargo. The hydrocarbons in
by gas chromatography. Levels of hydrocarbons (n-C14 the sediments collected 3 km from the wreck in February
to n-C34 range) greater than 20 p g HC/gm dry weight and sediments collected at the wreck site in July match
sediment (ppm) were found at three stations, and traces poorly with the cargo sample.
of hydrocarbon contamination (0.7to 1.1 ppm)werefound
at two other stations. Background levels in this area were
less than 0.5 ppm. These findings indicate that significant Introduction
hydrocarbon contamination extended at most 3-4 km
from the wreck site in a SE direction. Traces of contamina- On December 15, 1976, the tanker Argo
tion were found 3-5 km from the wreck in the E and S Merchant ran aground on Fishing Rip of Nantucket
directions. At the highly contaminated stations, the Shoals off the Massachusetts coast. Within one week she
majority of hydrocarbons were in the form of minute tar had broken into three parts, the stern and midsection
particles mixed with sand. Analysis of sediment grab remaining aground near the original point of impact and
subsections revealed no clear trend of hydrocarbon the bow section floating and eventually sinking
contamination as a function of depth in sections as deep 2.8 km SE of the stern section. During these events,
as 10 cm. Analyses of box core sections showed that approximately 28 x 103 metric tons of No. 6 fuel oil were
contamination extended at least to 8-13 cm in depth. The spilled into the ocean. In order to determine the degree to
stations which were found to contain significant hydro- which the sediments were contaminated in the vicinity of
carbon contamination in February were reoccupied on a the wreck site and Little Georges Bank, the University
cruise in July, 1977. Only the sediments at the bow of Rhode Island organized five cruises to the area. On
section of the wreck site showed contamination (0.6ppm) several of the earlier cruises which collected sediments
and these levels were significantly lower than found at over a 4000 km2 grid, the U.S. Coast Guard conducted
this site in February (4 to 122 ppm). on-board hydrocarbon screening using UV fluorescence
The high degree of physical activity on the shoals is spectroscopy, the results of which indicated that the
probably responsible for the areal patchiness and the sediments were contaminated with Argo Merchant
inhomogeneous mixing of hydrocarbons with depth in the oil only around the wreck site (Grose and Mattson,
sediments. Although chemical, physical and biological 1977). On the basis of these findings, a detailed sediment
weathering could be responsible for the observed survey of the wreck site area was conducted on R/V
decrease in hydrocarbon concentrations at the wreck Endeavor cruise 005, February 22-27, 1977. Several of
site, turbulent mixing on the shoals probably transported these stations were reoccupied in July 22-24,1977, five
the contaminated sediments out of the area or buried months later, on F/V Sideshow cruise 001.
them under clean sand. The objectives of this study were as follows:
The chromatographically resolved hydrocarbon (1) determination of the areal extent of surface sediment
components of the tar particles and sediments were contamination by the Argo Merchant oil; (2) measure-
statistically compared with the resolved components of ment of the depth of sediment contamination; and

I I Cargo and Slick Sampling. One cargo sample and two
69040 69020'W surface slick samples were collected by Dr. Jerome
Milgram of MIT who graciously provided subsamples of
these for analysis and information on the collection and
63. storage procedures (Milgram, personal communication,
.64 1977). The Argo Merchant cargo oil sample was
Fishing Rip' collected on December 19,1976. The cargo sample, from
I Shoals the full port tank #4, was taken by putting a glass jar into
5 TOD 5 8, 70 62 the tank, withdrawing the jar and closing it. The first slick
7 a (D61 sample (Milgram #11) was collected on December 19,
e 056 1976, by lowering a bucket from the side of the USCGC
41000'N - 55 72 04 Vigilant. The sample came from a distance of 2 to 4 km
0 ?1 3. from the ship and Milgram estimates that it had been on
the water for about one - hour. The second sample
48* 49* 500 ';51@;21 54 (Milgram #2) was collected on December 25, 1976, at
40-54'N, 68-38.1'W, from a large pancake, also by
47 lowering a bucket from the side of the USCGC Vigilant.
It had probably been on the water for about four or five
046 days. The cargo and slick samples were stored at room
- temperature in closed glass jars; after arrival in our lab,
40050' 045 they were stored at O'C until analyzed.
044 Sediment Analysis. Selected sediment samples from
EN-005 were chosen for analysis on the basis of two
criteria: (1) evidence of measurable or trace quantities of
petroleum based on the Coast Guard screening results
Figure 1. Station locations of EN-005 (February22-27,1977), and and/or (2) presence of minute tar particles which appeared
SS-001 (July 2-24, 1977). The "S" (Station 59) is the location on the top of the formalin surface of the biology sub-
of the Argo Merchant stem and mid section wrecks and the sample upon agitation of the sediment. Selection of
"B" (Station 70) is the location of the Argo Merchant bow SS-001 samples were based on (1) indications of con-
section. The circled dots are the stations from which the tamination found on previous cruises and (2) stations
sediments analyzed in this study were collected. The other adjacent to contaminated stations of earlier cruises.
stations were analyzed by the U.S. Coast Guard. Occasionally minute tar particles (1 to 2 mm) were
removed from the samples with tweezers prior to
analysis, and the tar particles and sediment were
analyzed separately.
After determination of the moisture content of a
(3) evaluation of physical, chemical and/or biological small portion of each sediment sample, 70-240 gms of
weathering of the oil as a function of time. wet sediment (50-200 gms dry weight) was weighed intoa
two I round bottom flask. Generally 60 p g of n-C20 and
Experimental Procedures 200 W g of n-C22 were added as internal standards for the
EN-005 samples, and 20 vg of n-C20 and 100 Ug of
Sediment Sampling. Sediment samples were collected n-C22 were added as internal standards for the SS-001
from the Nantucket Shoals Argo Merchant wreck samples. The samples were then saponified using
site area aboard RN Endeavor (EN-005, February 500-800 ml of a 70% 0.5 N KOH/methanol - 30% toluene
22-27, 1977) and aboard FN Sideshow (SS-001, mixture under reflux conditions for 2 hours. Enough water
July 22-24, 1977) (Figure 1). Surface sediment samples was added to the mixture (10 to 20% of the methanol) to
(0-15 cm) were collected with a Smith-McIntyre grab prevent transesterification. The sediment-solvent mixture
sampler. One-half of each grab sample was preserved was allowed to cool and then filtered using a pre-ignited
in formalin for biological archiving purposes. From the Whatman GF/C filter. The sediment and filter were
other half, used for chemical analysis, successive sub- washed with methanol, then with petroleum ether. The
samples were placed into individual one quart Mason jars. filtrate and combined washings were transferred to a
The depths noted in the results section are estimates of two t separatory funnel and separated into two phases
the different depths in the grab sampler from which the by adding an amount of distilled water equal to the amount
individual subsamples were taken. After a small portion of methanol. The non-saponifiable fraction (petroleum
(5 cc) of each subsample was removed for on-board ether-toluene phase) was separated from the saponifi-
screening by Coast Guard personnel, the remainder of able fraction (water-methanol phase) and the latter phase
each subsample was frozen until later analysis (Side- was extracted two additional times with petroleum ether.
show sediment samples were stored in an ice locker The petroleum ether extracts were combined and added
while at sea and subsequently frozen upon return to the to the non-saponifiable fraction and evaporated to
laboratory). Sediment cores (13-14 cm each) were dryness on a rotary evaporator at 35-400C. The residue
obtained on EN-005 with a box corer. Three subsamples was charged to a column in one ml of a 95% petroleum
V@14
;F
iS@
70
a

of each core were taken with plastic core liners. Two ether-5% toluene mixture. This column, prepared in a
subsamples were used in benthic flux experiments. The Pasteur pipette, contained (from bottom to top) a glass
third subsample, analyzed in this study, was frozen wool plug, 0.3 gm Cu powder (activated with 3N HCI), 1 gm
immediately after collection. silica gel (deactivated with 5% water), 1 gm alumina

Table 1. Total hydrocarbons in sel acted- sediment samples from Cruise EN-005. February 22-27, 1977.

Depth of
Station sediment Total hydrocarbons
(replicate) Latitude Longitude (cm) gglgm (dry wt sediment)

500) G* 40'57.1'N 69* 30. VW 0-1 <0. 1
50(2) G 40* 56.9'N 69'29.9'W 0-1 0.8
50(2) G 1-3 0.4
50(2) G 3-5 <0.1

560) G 40* 59.2'N 69* 27.0'W 0-1 1.2
56(3) G 40* 59.2'N 69* 27.3'W 0-1 <0.3
56(4) G 40'59.0'N 69* 29.0'W 0-1 21.5

57(l) G 41 * 02.0'N 69o 33.5'W 0-1 <0.1

590) G 41 * 02.5'N 69* 27.3'W 0-1 2.4t
590) G 1-3 0.5
59(l) G 3-5 <0.1 t

59(3) G 41'03.0'N 69* 27.5'VV 0-1 2.6t
59(3) G 1-3 <0. 1
59(3) G 3-5 <0.1

59(4) G 41'02.9'N 69* 27.2'W 0-1 0.3
59(4) G 1-3 0.1
59(4) G 3-5 <0.1

590) BC** 41'02.5'N 69'27.0'W 0-3 5.1
590) BC 3-8 1.3
590) BC 8-13 24.6

59(2) BC 41*02.6'N 69o 27.0'W 0-4 0.3
59(2) BC 4-9 0.8
59(2) BC 9-14 0.4

61 (2) G 41'02.6'N 69* 22.5'W 0-1 1.1
61(3) G 41 o 0 1.0'N 69* 23.0'W 0-1 0.7

70(l) G 41'01.8'N 69* 26.2'W 0-1 12.8
70(l) G 1-3 29.6
70(l) G 3-5 11.5
70(l) G >5 19.7

70(3) G 41 * 02.0'N 69o 26.6'W 0-1 10.2
70(3) G 1-3 4.0
70(3) G 3-5 5.6t

70(4) G 41 o 02.0'N 69* 26.5'W 0-1 118, 69.7, 35.7 t t
70(4) G 1-3 5.1
70(4) G 3-5 122

70(l) BC 41 o 02.0'N 69' 26.5'W 0-3 1.9
70(l) BC 3-8 2.7
70j1) BC 8-13 2.2

70(2) BC 41 * 0 1.9'N 69'26.3'W 0-3 2.7
70(2) BC 3-8 28.2
70(2) BC 8-13 37.5

*G = Smith-McIntyre grab samples
t tarball removed prior to analysis
* * BC = box core samples
tt triplicate analysis

(deactivated with 5% water), and another glass wool plug. site (now the location of the mid and stern sections); and
The sample was eluted with 15 ml of the 95% petroleum Station 70, the location of the bow section. Station 56 is
ether-5% toluene mixture. It was found that this column located 3.2 km SE of the bow section in the channel
procedure retained elemental sulfur, methyl esters, adjacent to Fishing Rip Shoal. The sediment at 59 and 70
methyl ketones and more polar organic compounds while was coarse sand while the sediment at 56 was gravel
eluting hydrocarbons including n -alkanes, and aromatics, mixed with crushed shells and mud. Background hydro-
such as phenanthrene, , pyrene and chrysene. The carbon levels in the area were less than 0.5 Mg/gm.Trace
petroleum ether-toluene mixture was then evaporated to levels of petroleum (0.7 to 1.1 Wg/grn) were found at
dryness and dissolved in methylene chloride. Each Stations 50 and 61. (Station 50 was 4.8 krn SW of the wreck;
sample was analyzed by gas-liquid chromatography Station 61 was 3.2 km E of the wreck.) Both of these
(GLC) with a Hewlett-Packard Model 5711 A or 5840A gas stations were located in the channel to the S and E of the
chromatograph equipped with dual flame ionization wreck and the presence of traces of petroleum hydro-
detectors. Most of the analyses were accomplished using carbons at these locations could either be from sources
dual packed columns (2m, 2.2 mm i.d. stainless steel) other than the Argo Merchant, or from the Argo
containing 10% SP-1 000 on Supelcoport, 80/100 mesh. A Merchant itself. Since the Coast Guard found no
few analyses were done using a glass capillary column evidence of Argo Merchant oil at the other stations
(15 m, 0.25 mm W., OV-1 01, Quadrex Corp.). The hydro- and, assuming that their results are correct, our findings
carbons measured eluted from the column between would indicate that significant sediment contamination in
n-C14 and n-C34 using temperature programming from February extended at most 3-4 km from the wreck site in a
90'C to 260'C at 80/minute for the packed column and SE direction. Traces of contamination from an unknown
600C to 240'C at 40/minute for the capillary column. The source could be found 3-5 km from the bow in the Eand SW
chromatograms were quantified by comparing the areas directions. At most, this would indicate that in February,
of the chromatograms with the areas of the n-C20 and/or a 10-15 kM2 area of the sediments were contaminated. It
n-C22 internal standards. is likely that only a small percentage of this area was
Procedural blanks were determined by carrying significantly contaminated, these areas being limited
solvents and internal standards through the analysis primarily to the immediate vicinity of the wreck site,
procedure in the absence of the sediments. The average especially around the bow section.
blank was 23 � 10 pg/sample. Concentrations were B. Depth of Contamination. An attempt was made to
reported for all sediment samples in which the blank
correction represented <50% of the blank plus sample determine the depth of sediment contamination by taking
value. For a 200 gm sediment sample, this would yield a successive subsamples out of Smith-McIntyre grab
detection limit of approximately 0.1 pprn (23 ug/200 gm). samples or by examination of box cores. Neither sampling
Whenever the blank correction was greater than 50% of device collected sand any deeper than 13-20 cm, and in
the total, the values were reported as less than the the case of gravel bottom stations, a maximum of only
detection limit for that specific sediment sample weight. 3-4 cm of sediment was collected. No clear trend as a
All the values reported in the tables have been corrected function of depth was noted in the EN-005 sediment
for the blank. samples. For example, at Stations 70(1)BC and 70(1)G,
the hydrocarbon concentrations were fairly constant with
Cargo, Slick, and Tar Particle Analysis. After weighing, depth, while at Stations 59(1)BC, 70(4)G, and 70(2)BC,
these samples were dissolved in the 95% petroleum the higher concentrations generally appear in the sub-
ether-5% toluene mixture and charged to the Cu-silica surface sediments, and at Stations 59(1)G and 59(3)G
gel-alumina column described previously. The hydro- the highest concentrations appear in the surface
carbon elutant was evaporated to dryness, dissolved in sediments (Table 1). The probable reasons for this
methylene chloride, and analyzed by GLC. variation are: patchiness due to inhomogeneous mixing
of minute tar particles into the sand, coupled with the
Results and Discussion highly active physical nature of the shoals. The tidal
currents in this area at times exceed 3 knots. Such
A. Areal Extent of Sediment Contamination. The locations current speeds will result in erosion and transportation of
where sediment was collected on EN-005 are given in coarse sand as found on the shoals and then deposition
Figure 1. Sediments from all of these stations were elsewhere as the current slackens (Heezen and Hollister,
prescreened by Coast Guard personnel. This study 1964). Then the tidal current reverses direction and the
presents detailed GLC analysis only on those samples process occurs again, this time in an opposite direction.
found to have at least some trace of oil by the Coast The depth to which this erosion-deposition cycle extends
Guard technique or those samples from which small tar with each half tidal cycle has not been studied on
particles floated out of the corresponding biological Nantucket Shoals but has been shown to involve at least
samples. A total of 45 subsamples from six EN-005 2 meters on Middle Ground (a shoal in Vineyard Sound,
stations and 15 subsamples from six SS-001 stations Massachusetts), and at times sediments as deep as 6
were analyzed. These latter stations are also shown in meters can be moved (Smith, 1969). In this present study,
Figure 1. it can only be concluded that contamination of some of
The present data (Table 1) indicate that in February, the sediment extends at least to8-13 cm in depth. Clearly,
1977, the sediments were significantly contaminated at the possibility exists that the contamination could be
3 stations (56, 59, 70 - significantly defined as greater significantly deeper than this.
than 20 gg HC/gm dry weight, a value ranging from 40
to >200 times higher than the background hydrocarbon C. Nature of Sediment Contamination. It was found that at
concentrations). Two of these stations were located at least some of the petroleum hydrocarbons in the sedi-
the wreck site locations: Station 59 at the original wreck ments were associated with minute tar particles. In order

Table 2. Comparison of total sediment hydrocarbon concentrations b
7 REGRESSION
with and without inclusion of tar particles (EN-005 sediments). 0 c LINE\
6-

5 -
Total hydrocarbons JJg1gm (dry wt sediment) Z Z
E' Z 4-
StatlRep/Depth* Sediment with Sediment with ;a
a
tar particles Tar tar particle
Y-1-0071-0.002
physically removed contribution contributions
included r -0.90

59(1)0-1 2.4 73.5 75.9 W 1 2 3 4 5 6
U)
59(1)3-5 <0. 1 48.5 48.6 Z ARGO MERCHANT Caroo Oil
0 Individual peaks (area percent)
OL
59(3)0-1 2.6 324 327 U)
W
70(3)3-5 5.6 29.1 34.7

0 n-C14
*all grab samples, depth in cm,
W
I.-
W
to find out the approximate percentage of the total n_C34
hydrocarbons these tar particles represented, particles
were physically isolated with spatula and tweezers from INCREASING TIME AND TEMPERATURE
four of the samples. The sediments (without their tar
particles) and the tar particles were analyzed separately- Figure 2. PartA. Chromatogram (packed column) oftarparticies
The result of this experiment is given in Table 2. In all separated from the sediment collected at Station 59, Grab 3,
cases, the tar particle contribution to the sediments was 0-1 cm depth, in February 1977. Part B. Graph of the areas of
the major contribution rather than any coating on the each resolved hydrocarbon component of this tar particle sample
versus the area of the corresponding peaks in the Argo Merchant
sand. It is also clear that the number and size of the tar cargo.
particles in any one sediment sample is going to greatly
affect the results. In order to minimize this problem, large
sediment weights (usually 100 to - 200 gms) were used
for each analysis in this study. In spite of this precaution,
replicate analyses from the same subsample indicate
variabilities of a factor of three. (See Station 70(4)0-1 in
Table 1 for replicate analyses data.) Since the uncertainty a
for the analytical procedure is no greater than 10%, the
sediment samples were inhomogeneous even when
taken from the same storage container. This is un-
doubtedly one of the reasons, perhaps the major one, why nC14
the Coast Guard on-board screening analytical results do
not agree in all cases with the results presented here.
Although our values are generally lower, there is some
correlation between the Coast Guard screening analysis
and our analysis (r = 0.69 for 20 samples) in spite of the UJ
patchiness. The patchiness due to inhomogeneous V) b
Z
mixing of the tar particles into the sand as a result of the 0
physical activity of the area led to even greater variations a-
U)
from grab sample to grab sample even within 50 meters of UJI
each other. At Station 70, for example, the upper sedi- cr
ments (0-3 cm depth) in replicate grab and core samples
contained 12.8,10.2,118,1.9 and 2.7 ug/gm, avariation
0
over two orders of magnitude. Thus the observed patchi- I
U
ness exists both on micro as well as a macro scale. UJI
D. Comparison of Slick, Sediment and Tar Particle Samples W
with the Argo Merchant Cargo Oil. The chromato- C
graphically resolved portion of the hydrocarbon com-
ponents of the tar particles, sediments and slick samples
were compared with the resolved components of the
Argo Merchant cargo oil, by a simple technique of
comparing areas of each resolved component. It should
be pointed out that the Argo Merchant carried two INCREASING TIME AND TEMPERATURE-
cargos, both No. 6 fuel oils, but only one of the cargos was
sampled at the scene. It was this cargo sample that was Figure 3. Part A. Chromatogram (packed column) of the Argo
used for matching purposes in this study. Although these Merchant cargo oil. Part B. Chromatogram (packed column)
L
'03
INE'5@"JN

Y., OO'._'.O.
ARGO @MERCHA C'ap9o Ol'lt'C
In 'V"Uo' peil'@
d 'area O'ce,
@-C.44 n_ 14

oils were reported to be nearly identical (Grose and ofa surface slicksample (Milgram *1) collectedon December 19,
Mattson, 1977), some of the following matching experi- 1976. Part C. Chromatogram (packed column) of a surface slick
ments could have been affected by differences in the sample (Milgram *2) collected on December 25, 1976.

Table 3. Comparison of resolved components in slick, sediment and tar particle samples with Argo Merchant cargo oil sample.

y vs. Argo Merchant No. of 5 ig.
y correlation coefficient area %s level

Argo Merchant duplicate
analyses same day 0.99 31 >0.995

Argo Merchant duplicate
analyses 2 days apart 0.97 38 >0.995

Argo Merchant duplicate
analyses - same day 0.93* 17 >0.995

Oil sample on water
one day (Milgram *1) 0.91 38 >0.995

Oil sample on water
five days (Milgram #2) 0.96 39 >0.995

Tar - station
59 (1) 0-1 (EN-005) 0.89 24 >0.995

Tar - station
59 (1) 3-5 (EN-005) 0.79* 17 >0.995

Tar - station
59 (3) 0-1 (EN-005) 0.90 40 >0.995

Tar - station
70 (3) 3-5 (EN-005) 0.44 25 0.975

Sediment and tar
station 70 (4) 3-5 (EN-005) 0.98 39 >0.995

Sediment without tar
station 70 (3) 3-5 (EN-005) 0.74 27 >0.995

Sediment - station
59 (1) BC 8-13 (EN-005) 0.86 19 >0.995

Sediment - station
56 (4) 0-1 (EN-005) 0.21 13 <0.95

Sediment - station
70 (1) 0-1 (SS-001) 0.18* 14 <0.95

*glass capillary mal .or peaks

resolved component pattern between the two cargos. in Figure 2. The area percents of each resolved peak in
With the help of the HP-5840 integrator, the area of the chromatogram (Part A of Figure 2) were plotted versus
each peak (or resolved component) was expressed in the corresponding peaks in the cargo sample (Part A of
area percent as per Equation (1): Figure 3) to yield the resulting graph given in Part B,
.A i Figure 2. Visually the chromatograms look similar and
A% =_ - 1000/0 linear least squares regression analysis of the individual
i fAi X area pe r-cents confirms a strong correlation (r = 0.90). An
example of a poorer match is given in Figure 4. Here the
where A%i is the area % of the individual resolved com - resolved pattern of the sample visually does not look like
ponent, and Ai is the area in arbitrary Units of each indivi- the cargo sample. The peaks are less pronounced and in
dual peak. The individual area percents of each peak of some places are absent altogether. Again the linear least
the standard (in this case, the Argo Merchant cargo squares regression analysis of the area percents of the
oil sample collected by J. Milgram) are then plotted versus resolved species indicates a poor correlation (r = 0.44). A
the corresponding area percents (using peaks having the summary of these examples and other samples com-
same retention times) of the sample. All the reported pared to the Argo Merchant cargo is given in Table 3.
peaks were used in the case of packed column chromato- Occasionally, a glass capillary column was used for the
grams; however, only the major peaks (with area per- matching experiments. An example of the Argo
centages greater than 1 %) were included when glass Merchant oil chromatogram using glass capillary gas
capillary GILC was used. An example of the result is given chromatography is given in Figure 5, part A.

9 - b 1425 weathering in slick samples up to 5 days. This is particu-
larly interesting, since the Argo Merchant cargo con-
tained about 20% of a cutter stock of light oil added to the
PERFECT FIT/
7 -
6 - No, 6 oil (Grose and Mattson, 1977). Apparently, there had
not been substantial losses of these lighter components
5 -
tE 10
a R SRESS N during the 5 days it was on the surface of the water.
4
LINE
Undoubtedly the cold atmospheric temperature impeded
Z the evaporative losses. (The mean temperature for these
Y-0.929X+0.22 5 days was -3'C, National Weather Service, Warwick, R.I.)
-0.44 In addition, the cold weather may have reduced the dis-
S
Z solution of the more potentially soluble components from
0 1 2 3 4 5 6 the slick into the water column.
CL
U)I ARGO MERCHANT Cargo Oil
LLI Three of the four tar particle samples also match
X Individual peaks (area percent) well with the cargo oil (r= 0.79 to 0.90, Table 3). The other
Ir
0 tar particle sample (sample 70 (3) 3-5, also shown in
Figure 4), apparently had undergone some weathering, or
W n- C 14 n-C34 was from a different source. The sediments collected
around the wreck site matched well with the cargo
although the sediment with the visible tar particles
INCREASING TIME AND TEMPERATURE removed did not match as well as the other sediments
Figure 4. Part A. Chromatogram (packed column) of tarparticles from the wreck site. The hydrocarbons in the sediments
separated from the sediment collected at Station 70, Grab 31 collected 3 km from the wreck site in February (Station 56
3-5 cm depth, in February 1977. Part B. Graph of the areas of (4)0-1, EN-005) did not match with the cargo oil; the
each resolved hydrocarbon componentof this tar particle sample chromatogram showed that either extensive weathering
versus the area of the corresponding peaks in the Argo Merchant of the resolved peaks had taken place (if the hydro-
cargo. carbons were Argo Merchant derived), or the hydro-
carbons at this station were a weathered product of some
other petroleum input. The hydrocarbons found in the
sediments in July (Station 70 (1), SS-001) also did not
match with Argo Merchant oil and the glass capillary
chromatogram of these hydrocarbons is given in Figure 5,
Part B. Visually there is little similarity between the
resolved hydrocarbon pattern of the Argo Merchant
nC.4 cargo and this July sample. The July sediment pattern
nC14 does, however, resemble patterns of tar lumps found on
Bermuda beaches reported by Butler et al. (11973),
patterns of tar particles >1 cm diameter in size in the
W N. Atlantic reported by Wade et al. (11976), and patterns
U) nCio nC22 b
Z
for hydrocarbons in surface microlayers in coastal waters
south of Martha's Vineyard in February, 1977 (Boehm,
1977). It can therefore only be concluded that the source
of -the hydrocarbons in the vicinity of the Argo
Merchant wreck site in July is unknown. Clearly more
0
information concerning the source of the hydrocarbons
W
0- found in these environmental samples where poor
W rtl@14
0 matches-were obtained could be learned if specific persis-
nC14
tent compounds unique to the cargo could be found and
studied in the samples.
INCREASING TIME AND TEMPERATURE E. Variation with Time. On July 22-24,1977, five months
Figure 5. Part A. Chromatogram (glass capillary column) of the after the FIN Endeavor cruise 005 in February, and
Argo Merchant cargo oil. Part B. Chromatogram (glass seven months after the initial spill, further sediment
capillary column) of the sediment collected at Station 70(1), samples from the wreck site areas were collected aboard
0-1 cm depth, in July, 1977. (The n-C20 and n-,C22 are internal F/V Sideshow. Samples were collected for chemical
standards added for quantitation purposes.) analysis and biological archiving at all the previously
contaminated stations found on EN-005, with the excep-
In order to demonstrate the correlations obtained tion of Stations 56 and 57 (where the July samples con-
using duplicate injections, Argo Merchant cargo oil tained only cobbles and shell fragments). The hydro-
was analyzed using different injection conditions (as carbon analytical results are given in Table 4. The
noted on Table 3). All of these procedures yielded strong samples all contained 0.6 jA g HC/gm or less. Only at one
correlations (r > 0.90) for Argo Merchant oil versus station (Station 70 (1)) were traces of petroleum hydro-
Argo Merchant oil. The chromatograms of the cargo carbons found. An example of the chromatograms for this
-C
Znc
14

and surface slick oil samples are given in Figure 3. station is given in Figure 5B. Note that the chromatogram
Visually there is a strong resemblance and the correla- does not compare well with the Argo Merchant oil
tions of the resolved area percents also confirms this cargo chromatogram (Figure 5A), and there is no statisti-
strong correlation (Table 3). There is little evidence of cal correlation for.the resolved components (r = 0.1.8,

Table 4. Total hydrocarbons in selected sediment samples from Cruise SS-001. July 22-24, 1977.

Depth of
Station sediment Total hydrocarbons
(replicate) Latitude Longitude (cm) JAgIgm (dry wt sediment)

50(2) 40' 56.9'N 69* 30.1'W 0-1 0.1

590) 41 * 02.5'N 69* 27.8'W 0-1 0.2
59(2) 0-1 <0.1
59(3) 0-1 <0.2

61 (1) 41*01.5'N 69* 24.3'W 0-1 0.4

700) 41 * 02.0'N 69* 26.8'W 0-1 0.5
70(l) 1-3 0.6
70(l) 3-5 0.6
70(l) 5-10 0.4

70(2) 41 * 02.0'N 69* 26.8'W 0-1 0.2
70(3) 41 * 02.0'N 69* 26.8'W 0-1 <0.2

72 (2) 40* 59.2'W 69* 27.7'W 0-1 0.2

76 0) 41 * 02.0'N 69* 26.5'W 0-1 0.2
76(2) 41 * 02.4'N 69* 26.8'W 0-1 <0.2
76(3) 41 * 02.7'N 69* 27.7'W 0-1 0.4

*all grab samples

Table 3). Since this sediment sample was collected at the based on such samples when daily fluctuations in
Argo Merchant bow section wreck site station, it is sediments of two meters are possible.
tempting but highly unlikely to conclude that this is Argo In summary, the hydrocarbon concentrations in the
Merchant oil after 7 months of physical and chemical surface sediments in the vicinity of the Argo Mer-
weathering and/or biological degradation. This is unlikely chant wreck site in July, 1977, were significantly lower
since the n-alkanes which appear predominantly in the than found on the earlier cruise in February. At only one
sample are usually the first to degrade with time relative station was evidence of petroleum hydrocarbon contami-
to the complex mixture of cycloparaffins, aromatics and nation found, and these hydrocarbons did not match the
naphtheno-aromatics. These hydrocarbons could have Argo Merchant cargo oil.
been from non-cargo sources of oil from the Argo F. Mass Balance Considerations. A rough estimate of the
Merchant (i.e. hydraulic fluids, etc.) or from non-Argo percentage of spilled oil which became incorporated in
Merchant sources altogether. With the exception of the sediments can be made if a series of assumptions are
this station (Station 70(l)), the rest of the samples gave no used. Generally these assumptions tend to maximize the
evidence of petroleum hydrocarbons, weathered or amount of oil in the sediments: (1) Assume an average
otherwise, above the normal background levels. concentration of hydrocarbons in the vicinity of the wreck
Since the fishing vessel had a draft of only 2 meters, (Stations 59 and 70) in February, 1977, is 16 ppm
it was possible to collect sediment samples on top of the (arithmetic mean � the standard deviation was 16 � 30
shoal, between the bow and stern section locations, a ppm); (2) Assume a linear rate of weathering or dispersal
feat not possible with R/V Endeavor (draft of over 5 with time, such that 16 ppm in February decreased to an
meters). Stations 76(l), 76(2), and 76(3) represent a average of 0.3 ppm in July in a linear decrease (extrapola-
transect across the shoal from the bow site to the stern tion of this date back to December, 1976, produces an
site. It was felt that sediments at these stations might average minimum value hydrocarbon concentration of 22
have been heavily contaminated with oil as the bow ppm); (3) Assume an affected area of 15 kM2; (4) Assume
section drifted over them before it was sunk (Grose and a depth of contamination of 15 cm, and a constant
Mattson, 1977). In July, no evidence of such contamina- concentration of hydrocarbons with depth; and (5) The
tion was found. Again because of the turbulence of average sediment density is approximately 3 gm/ce.
the area, the oil could have been degraded, Using these assumptions, it was calculated that a
buried or transported away from the shoal, especially maximum of 1.5 x 102 metric tons of oil was incorporated
since seven months had passed from the time of the in the sediments, or 0.5% of the total oil spilled. It is felt
original spill. that this amount represents a maximum estimate.This
No evidence of either transportation to other calculation would therefore indicate that surface
adjacent stations (50, 61, 72) or burial at the wreck site sediments in thevicinityof the wreck site were notamajor
was found. Again, since we obtained sediments only as sink for the oil in the case of the Argo Merchant
deep as 10 cm, it is unwarranted to make conclusions oil spill.

Acknowledgments
This work was supported by NOAA Contract
03-7-022-35123, by EPA Grant R80547701 0, and by the
National Sea Grant Program (04715844088). We wish to
thank Sheldon Pratt, Robert Brown and Keith Cooper for
aid in the collection of samples; William Hahn, Ted
Bentinnen, and Jim Hannon, marine technicians, for aid in
deployment of collection equipment; Captain Herb
Bennett and the crew of the R/V Endeavor for the safe
journey to the wreck in February in spite of the unpleasant
conditions; Captain Paul Brayton and crew of F/V
Sideshow who gave up fishing time to transport us to
the shoals in July; Terry Wade and Edward Van Vleet for
aid in the analysis; Richard Jadamec and Scott Fortier
(USCG, Groton) for the use of their preliminary screening
data; and Jerome Milgram for the collection and donation
of the cargo and surface slick samples used in this study.
We also thank EPA-Narragansett for the use of the
Smith-McIntyre grab samplers.

References

Boehm, P. D. 1977. Hydrocarbon Chemistry Quarterly Progress
Report, Chapter 5.1 In New England OCS Environmental
Benchmark, 4th quarterly summary report, Contract AA
550-CT6-51, Energy Resources Co., p. V1 -V52.
Butler, J. N., B. F. Morris, and J. Sass. 1973. Pelagic tar from
Bermuda and Sargasso Sea. Bermuda Biological Station,
Special Publication No. 10. 346 pp.
Grose, P. L., and J. S. Mattson, eds. 1977. The Argo Merchant
Oil Spill: A preliminary scientific report. NOAA Special
Report. NOAA-ERL, Boulder, Colorado. 322 pp.
Heezen, B. C., and C. D. Hollister. 1964. Deep sea current
evidences from abyssal sediments. Marine Geology
1: 141-174.
Smith, J. D. 1969. Geomorphology of a sand ridge. J. Geol.
77: 39-55.
Wade, T. L., J. G. Quinn, W. T. Lee, and C. W. Brown. 1976.
Source and distribution of hydrocarbons in surface waters
of the Sargasso Sea. Proceedings of Sources, Effects,
and Sinks of Hydrocarbons in the Aquatic Environment.
American Institute of Biological Sciences, American
University, Washington, D.C. pp. 271-286.

Biological Studies

Kenneth Sherman, Chairman

Summary of Biological Studies

The discussions of biological effects of Argo oil on adjacent areas showed no evidence of reduction in the
marine populations were wide ranging and not as sharply )opulation densities of the more important fish and
focused as the physical and chemical sessions. Prior to shellfish stocks.
the Argo spill no large-scale effort to assess oil-related The limited shorHerm impact was attributed to the
biological damage had been undertaken in the open mitigating ecological circumstances during the spill:
waters of the continental shelf off the northeast coast. 1) biological productivity was at an annual low, 2) no large
The concerns expressed by the Symposium participants scale deposition of oil was observed in the sediments,
emphasized the difficulties in attributing changes in the and 3) the high velocity wind and currents of the season
abundance and condition of marine populations to oil as carried the oil, limited in its distribution largely to the
compared to changes from other causes. Attempts made surface, rapidly offshore while it was in the process of
to isolate oil induced mortalities or sublethal effects need evaporation, emulsification, and dissipation in the tur-
to be viewed against the background of natural fluctua- bulent upper layers of the water column.
tions and the complexities of multispecies interactions at The longer term ecological effects will be more
different trophic levels and over a range of spatial and difficult to assess against the background of natural
temporal scales. Studies to deal with these problems fluctuations in population abundance. Time-series
need to be in place and operational long before any acute observations in the spill area are continuing jointly by the
spill event. Northeast Fisheries Center of the National Marine
The results brought together at the Symposium, Fisheries Service and the Manomet Bird Observatory.
although somewhat fragmentary, were sufficient to indi- Damage assessment of acute spills of oil and other
cate that the impact of Argo oil on the populations of potential toxicants needs to be considered against base-
Nantucket Shoals was minimal based on the short-term lines of population densities and their physiological
assessment studies carried out during the previous 12 condition. The existing baseline for marine populations
months. Supporting evidence for this conclusion was off the east coast requires some augmentation, particu-
found in both the population studies and the physiological larlywith respect to "physiological eff ects" and "probable
effects studies done at the tissue and organism levels for onshore effects." In addition, to ensure adequate scien-
molluscs, fish, birds, zooplankto6, and benthic crusta- tific response to acute spills, an organizational matrix is
ceans. In the histopathological, biochemical, and required that considers national, regional, and local
physiological studies the recovery from initial observa- expertise of scientists and managers in the development
tions of sublethal impact to "normal" conditions was of regional and local response plans.
reported for each of the groups of organisms examined, Several specific recommendations for dealing with
except for marine birds. The bird mortalities caused by oil, acute spill events on the continental shelf were made
however, were few and the impact on bird stocks was not during the biological workshop session:
considered significant. Genetic damage was observed in 1. New methods and systems need to be developed for
developing cod and pollock embryos, but the effect was combining laboratory, in situ experiments, and popula-
localized and it is unlikely to have a significant effect on tions assessment studies, in order to reduce the biases
subsequent year-class recruitment of these stocks. involved in extrapolations from small sample sizes to
Results of laboratory petroleum hydrocarbon exposure population level inferences.
studies of developing cod embryos indicated that con- 2. A significant series of as yet unanalyzed biological
centrations of 250 ppb were lethal. Although this level of samples and/or data pertinent to long-term studies is
oil was detected, the area of concentration was limited to available particularly for benthos. Effort should be under-
the immediate vicinity of the Argo wreck, and is likely to taken to support the completion. of these studies.
have affected a relatively small fraction of the Nantucket 3. More consideration should be given in the future to
Shoals cod stocks. observations of primary production; this principal eco-
Trophic pathways of Argo oil were observed among system component was not adequately examined in the
the zooplankton and benthos leading directly to important Argo spill assessment.
fish stocks. Further comprehensive stomach analyses of 4. Future biological studies should make greater use of
fish revealed a minor incidence of "oiled" prey in the risk-analyses projections in the planning and conducting
digestive tracts of fish. Less than five percent of the fish of population assessments and laboratory studies of
examined showed anytraces ofArgo oil in tissue samples acute spills.
collected within 30 days of the spill. On subsequent 5. Concern was expressed over: 1) the persistence of tar
surveys of bottom fish no Argo oil traces were found in clumps in the surface waters and their effects on biologi-
fish, shellfish, or zooplankton samples. Commercial cal populations; and 2) the persistent municipal and
catches and bottom trawl surveys of the spill zone and industrial input of petroleum hydrocarbons and other

toxic substances and their effects on the marine eco-
system. Additional research eff ort in the Northeast should
be directed to deal with these problems.
6. The need was expressed, and a recommendation
endorsed by the participants, for developing national
standards for acceptable levels of petroleum hydro-
carbons in fish and shellfish.
Considerable effort is now underway in the North-
east by NOAA and EPA to develop a regional response
plan that will be implemented in the event of a large-scale
spill of oil or other toxic substances, as a direct result of
the Argo Merchant incident. Although our knowledge of
the effects of petroleum hydrocarbons on marine popula-
tions of the outer continental shelf remains rather incom-
plete, the Argo incident is serving as a catalyst to over-
come existing deficiencies.

Kenneth Sherman, Chairman
Biological Studies Session

Microscopic Observations on Vertebrates
and Inverteb tates Collected Near the
Argo Merchant Oil Spill

Thomas K. Sawyer

National Marine Fisheries Service
Northeast Fisheries Center
Oxford Laboratory
Oxford, Maryland

Abstract
Fish, molluscs, crustaceans, sea urchins, and starfish were identified, dissected for histological processing,
were collected from control stations and from stations and embedded in paraffin blocks. Sections were cut at
which were selected because of their proximity to theArgo 6 pm and stained with Harris's hematoxylin-eosin solu-
Merchant oil spill. Histopathological findings that could be tions or by the Feulgen-reaction. Stained sections were
attributed solely to exposure to petroleum were not seen examined with routine bright-field microscopic proce-
in any of the tissues examined. Although collections dures and photographed with Kodak Plus-X film. Speci-
probably were made too soon after the oil spill for mens taken from control stations and oil-spill stations
potential tissue damage to have occurred, the micro- were compared to determine whether pathological
scopic findings provide baseline data for future studies. conditions were present that could be attributed 'to
Several fish species had edematous gills, detached petroleum hydrocarbon exposure.
epithelium, or hyperplasia of the olfactory epithelium.
Molluscs, sea urchins, and starfish tissues were un-
remarkable. Hermit crabs had sessile ciliate protozoa Results
attached to the antennae and numerous granulocytic Fish, molluscs, crustaceans, sea urchins, and star-
hemocytes in the hemolymph. Additional studies on fish that were processed for histological study are
hemolymph of hermit crabs are necessary to determine summarized in Tables 1 and 2. Molluscs were remarkably
whether abundant granulocytes are characteristic of free of tissue pathology of any kind. Several fish
this host. (alewives, winter flounder) had edematous gills and
Early in 1977 the Pathobiology Investigation staff of detached epithelium, and the olfactory epithelium of
the Oxford Laboratory was asked to participate in a @ellowtail flounder appeared to be hyperplastic in areas
multi-disciplined study of the marine organisms that might of non-sensory and indifferent epithelium. Arnmodytes
have been affected by the Argo Merchant oil spill. It was larvae had ocular lesions and malformations or lack of
realized from the outset that the spill was too recent to pigmentation of the eye. None of the conditions observed
assure that any observed pathology could be related in fish could be attributed directly to the effects of
directly to oil. However, it was thought that the study spilled oil.
would provide an opportunity to obtain background Sea urchins and starfish did not have obvious signs
information for future follow-up studies. The following of pathology or stress and all specimens examined
report summarizes observations that were made on a appeared normal. Probable ciliate protozoa that were
diverse group of vertebrate and invertebrate species. poorlyfixed and stained in histologic sections were noted
in the digestive tracts of several sea urchins. Live material
Methods was not available for study. One of the lobsters had
several cellular nodules in the gill, pyknotic nuclei, and an
All animals were collected with an otter trawl, encysted larval trematode. Hermit crabs had ciliates
sorted, and preserved in neutral formalin. Specimens (sessile suctorians and loricate forms) attached to their

Table 1. Summary of species of fish and molluscs
examined for histopathology.

Species Control Station Oil Station

FISH
Winter Flounder, 15
Pseudopleuronectes
15 12
americanus

Yellowtail Flounder,
Limanda ferruginea 15 7

2
Arnmodytes sp. (larvae) 6

Totals 30 25

MOLLUSCS
Scallop, Placopectin
magellanicus 15 25

Squid,
illex illecebrosa 1

Squid, Loligo sp. 1 8

Whelk, Buccinum sp. 5

Clam, Artica islandica 1
4
Clam, Mytilus edulis 1 MAW:
Totals 17 40 Figures 1-4. Pholomicrographs of stained sections of various
tissues of the hermit crab, Pagurus pollicaris. 1. Loricate sessile
ciliates on gill cuticle. Harris hematoxylin-eosin, 640 x 2. Cellular
response in eyestalk, note black degenerate nuclei. Harris
hematoxylin-eosin, 640 x 3. Focal area of cellular degeneration
localizedin sinus of the hepatopancreas. Feulgen-reaction, 640 x
Table 2. Summary of species of crustaceans, sea urchins, and starfish 4. Large eosinophilic hemocytes in gill filament Harris
examined for histopathology. hematoxylin-eosin, 1600 x

gills or antennae (Figure 1); foci of tissue necrosis in the
Species Control Station Oil Station intestine, stomach, or eye stalk (Figure 2); necrotic areas in
the hepatopancreas (Figure 3); and abundant granulocytic
CRUSTACEANS hemocytes (Figure 4). Although the hermit crabs had the
Hermit Crabs, greatest variety of tissue abnormalities, there were no
Pagurus pollicaris 6 12 specific differences in animals from the various collection
sites.
Rock Crabs,
Cancer irroratus - 4
Discussion
Lobsters, Homarus None of the microscopic findings from animals
americanus 2
- - collected during the oil spill cruises could be related to
Totals 6 18 the presence or absence of petroleum hydrocarbons.
Abnormalities in fish and hermit crabs were remarkable
SEA URCHINS and provided indications of the types of conditions that
Strongylocentrotus would be worthy of follow-up studies on long-term effects
droehbachiensis Q) 9 of oiled ocean-sediments. Specific effects of oil on wild-
Totals 9 caught fish are difficult to measure because migratory
habits may prevent associations from being made
STARFISH1 between control and stressed collecting stations.
Asteria forbsei 2 5 Bottom-dwelling crustaceans such as the hermit crab
Totals 2 5 may be suitable since they are non-swimming animals
1 Starfish total includes one Henricia sp. (?) from an oil station. and probably do not routinely move great distances.
Similarly, starfish and sea urchins may be suitable for

long-term studies on bottom-dwelling animals. The which should be avoided in planning future research.
numbers of starfish and sea urchins were too small to be First, we must know beforehand whether the oil is of high
of value in the present study and their duration of expo- or low toxi 'city and the extent to which the oil has dis-
sure to oil probably was too short to initiate any detect- persed or remained in the vicinity of the spill. Secondly,
able tissue response. Preliminary findings with hermit a highly co-ordinated team response is needed to assure
crabs indicated that they may have remarkably large that hydrocarbon analyses are made of the bottom
numbers of granular hemocytes. Follow-up studies to sediment and water column at each sampling station.
determine whether or not exposure to oil has a rapid Ideally, each animal selected for pathological examina-
effect on crustacean hemocytes may prove useful for tion should be retained for follow-up hydrocarbon
assessing some of the immediate effects of spilled oil on analysis. Stations that are selected for intensive study
marine invertebrate species. Our studies show that in should be sampled at regular intervals to follow pro-
general, new information on the normal histology of fish, gressive change or lack of change in preselected
molluscs, crustaceans, etc., is essential to a better animal species.
understanding of their subtle changes in response to The critical need for precise measurements on the
environmental pollutants. effects of oil spills on aquatic resources is as much in
evidence today as it was over 50 years ago. Lane and
Bauer (1925) wrote: "It would appear, therefore, that oil
Considerations for Future Research pollution has considerable effect upon the edible quali-
ties of aquatic animals and may aff ect the migratory habits
Stalked ciliates on the antennae and gills of hermit of fish; it is detrimental to shellfish by reason of destroying
crabs collected in control and oil-spill stations provided the larval forms and rendering adult molluscs unfit for
indirect evidence that the full impact of potential harm food. With regard to waterfowl, it appears to be a cause
caused by petroleum had not occurred at the time of the of considerable destruction, rendering the birds helpless
study. Sessile ciliate protozoans are known to reproduce through its mechanical action on the feathers. It has been
by budding and shedding free-swimming ciliated larvae. stated that when birds are not actually killed as a result
The larvae in turn become attached to new substrates of contact with the oil they are rendered unfit for food
where they develop to the stalked adult stage. Langlois due to the oil taint."
(1975) studied the effects of dissolved organic matter,
phenols, and carbohydrates on substrate selection by
motile telotroch larvae of Vorticella marina and found that Acknowledgments
certain algal exudates exert a significant influence on the
distribution and settlement of Vorticella on aquatic Histological observations on molluscs were made
substrates. New studies on the effects of oiled crusta- by Mr. Fred Kern; on fish by Mr. Martin Newman, Dr. Joel
cean exoskeletons on settling rates of larval ciliates Bodammer, and Dr. Robert Murchelano. Crustaceans,
could be useful for investigating the fate and effects of oil sea urchins and starfish were dissected and prepared for
in the environment. Jones and Rogers (1968) listed 7 histological processing by Ms. Sharon A. MacLean.
species of ciliate protozoa from the intestinal tracts of 4 Mr. John Ziskowski was responsible for the collection,
species of American sea urchins, and Beers (1961) dis- preservation and station data pertaining to all of the
cussed the adaptation of Euplotes balteatus to commen- animals that were examined by the laboratory staff.
sal.life in Strongy1ocentrotus droehbachiensis. Recent
studies by Andrews and Floodgate (1974) showed that References
free-living ciliates such as Euplotes and Uronema
ingested oil residues while feeding on bacteria adhering Andrews, A. R., and G. D. Floodgate. 1974. Some observations
to partially degraded oil globules. Published studies on on the interactions of marine protozoa and crude oil
sea urchins and ciliates suggest that benthic animals residues. Marine Biol. 25: 7 - 12.
might ingest oiled food organisms with subsequent Barry, M., and P. P. Yevitch. 1975. The ecological, chemical and
uptake of the oil by intestinal protozoa. The limited histopathological evaluation of an oil spill site. Part 111.
observations on sessile and endocommensal ciliate Histopathological studies. Marine Poll. Bull. 6: 171-173.
protozoa that were made in the present study indicate Beers, C. D. 1961. Is the ciliate Euplotes balteatus adapting to
commensal life in the sea urchin Strongylocentrotus
that they deserve further investigations under careful and drobachiensis? J. Parasitol. 47: 478.
well-planned field and laboratory conditions. Dow, R. L., and J. W. Hurst, Jr. 1975. The ecological, chemical and
The effect of spilled oil on mortalities among benthic histopathological evaluation of an oil spill site. Part 1.
animal species is difficult to assess at sea. In contrast, it Ecological studies. Marine Poll. Bull. 6: 164-166.
was estimated that 85 percent of the marketable clams, Jones, I., and T. E. Rogers. 1968. Studies on the endocommensal
Mya arenaria, were killed following an oil spill in Long ciliate fauna of Caribbean sea urchins. Biol. Bull. 135:
Cove, Searsport, Maine (Dow and Hurst, 1975). Neverthe- 514-519.
less, it should be possible to study the population Lane, F W., and A. D. Bauer. 1925. Effect of pollution on
dynamics of certain animal species indigenous to shoals marine and wild life. Append. V to the Report of the
Commissioner of Fisheries for 1925. B. F. Doc. No. 995.
and ridges in the vicinity of known deepwater oil spill Langlois, G. A. 1975. Effect of algal exudates on substratum
locations. Barry and Yevitch (1975) discovered a high selection by motile telotrochs of the marine peritrich
incidence of gonadal-tumors in Mya arenaria that were ciliate Vorticella marina. J. Protozool. 22: 115-123.
contaminated by the Long Cove oil spill. They continued
their sampling program subsequent to the spill and found
a 9 - 18 percent incidence of tumors 31/2 years later.
It is apparent that our efforts to respond to the Argo
Merchant oil spill suffered from several serious limitations

Histopathological Analyses of Benthic
Organisms From the Vicinity of the Argo
Merchant Wreck

Robert S. Brown and Keith R. Cooper

Marine Pathology Laboratory
Department of Animal Pathology
University of Rhode Island
Kingston, Rhode Island

Abstract Rhode Island cruises to the area of the Argo Merchant
wreck; four on the R/V Endeavor and one on the F/V
A variety of benthic species were collected on two Sideshow. The authors visited the wrecksite on the last
cruises to the Argo Merchant and examined histo- two cruises in February and July, 1977. The purpose of
pathologically. the fourth cruise was to determine, two months after the
On the first cruise, two months after the spill, only spill, the areal extent of sediment, phytoplankton, zoo-
two crabs were collected. A Cancer crab was found dead plankton and macrobenthos contamination with Argo
with a thick deposit of Argo oil coating the remnant gut. A Merchant oil and to deploy bottom drifters. The authors
hermit crab was found moribund. One Modiolus had cooperated with Dr. Eva Hoffman and Sheldon Pratt in the
mantle lesions consisting of round, raised calcified collection of sediments and with Renata Polak and Audrey
nodules several mm in diameter adjacent to patches of Fillion in the collection of zooplankton. Scott Fortier, USCG,
Argo oil deposited on the internal shell,surface. One performed the spectrofluorometric analysis of samples of
starfish of 250 examined had tarballs in the buccal cavity. Argo oil. The purpose of the July cruise was to determine
In contrast, on the second cruise, seven months the extent of persistence ofArgo oil in the area after seven
following the spill, no oil was seen in eight Cancer and ten months. Again, in cooperation with Dr. Eva Hoffman and
hermit crabs, 44 starfish and 5 sea cucumbers collected Sheldon Pratt, most of the February sites were
alive. One Modiolus, also visibly uncontaminated, had re-sampled.
extensive calcium nodule formation on an adductor The purpose of this paper is to present the histo-
muscle and surrounding mantle. The findings suggest the pathological findings and conclusions based on these
effects of the Argo oil were, for the most part, within the findings on the impact of theArgo Merchant oil spill on the
physiological toleration limits of the macrobenthos, and the macrobenthos dredged from the vicinity of the wreck.
overall impact of the oil spill was minor.

Methods
Introduction Benthic organisms were collected by towing a
The wreck of the Argo Merchant had all the makings scallop dredge for approximately 10 min. on the bottom.
of a major ecological disaster. As the ship broke up on Five dredge samples were collected on February 26,
December 21, 1976, approximately 71/2million gallons of 1977, which was two months after theArgo Merchant spill
oil were released into an area of ocean near Georges took place. Again on July 24, 1977 four dredge samples
Bank, one of the world's most productive fishing grounds. were collected from the same vicinity to measure residual
The well known toxic effects of oil on marine organisms, effects seven months after the spill. The precise locations
especially when spilled into marine environments from and depths of the areas are shown in Table 1. In general,
damaged vessels (all references listed), made it impera- the dredges were made along a southwest to northeast
tive to determine the areal extent of oil contamination and line in the Fishing Rip of the Nantucket Shoals in front of
its potential effect on the biota. near the wreck site. the Argo Merchant bow.
For these purposes, there were five University of All animals were examined grossly for any abnor-

Table 1. Location of the Dredge Samples

Dredge Sample Date Time Dep th M Loran C LatitudelLongitude

February 1 start 2/23/77 2200 40 37609.9 41'03.7' N
70091.3 69'23,8'W
stop 2/23/77 2238 42 37596.4 41'04.9' N
70087.2 69'23.0'W

2 start 2/24/77 1035 46 37661.8 40'59.5' N
70104.7 69*27.0'W
stop 2/24/77 1103 46 37655.8 41 *00.2' N
70102.4 69*26.8' W

3 start 2/24/77 1608 42 37699.3 41'00.1' N
70095.6 69*32.2' W
stop 2/24/77 1630 38 37701.3 41'00.5' N
70093.5 69'33.0'W

4 start 2/26/77 1008 44 37663.3 40*59.5' N
70105.9 69*26.9'W
stop 2/26/77 1034 44 37657.1 40*59.6' N
70105.0 69'26.4'W

5 start 2/26/77 2306 40 37637.1 41*02.1' N
70094.9 69026.0'W
stop 2/26/77 2330 24 37632.0 41*02.9' N
70091.5 69'26.0'W

Loran A
July 1 7/24/77 1100 40 31-15-1026 4loO4.6' N
11-13-3695 69o22.4'W
31-14-6383

2 7/24/77 1130 38 31-15-1035 41'02.2'N
11-13-3720 69*25.2' W
31-14-6370

3 7/24/77 1200 38 11-13-3730 41*00.7' N
31-14-6360 60*26.1'W

4 7/24/77 1230 40 11-13-3743 4W59.0' N
31-15-1049 69*27.8'W
31-14-6347

malities or oil fouling before preservation. Horse mussels macrobenthos were stored in plastic bags on ice on
were carefully shucked on board and the shell liquor board ship and treated with fixative the following day in
examined for the presence of oil through either a visible the laboratory. Archived samples of all animals were
sheen or odor. The gonads, colored either red (male) or rinsed of the initial fixative and stored in 70% ethanol. For
yellow (female), were examined for extent of develop- histopathological analysis, tissue samples were rinsed
ment. A bissecting incision was made in each starfish arm overnight in tapwater, dehydrated in an alcohol-xylene
to enhance examination and fixation. All macrobenthos series and embedded in paraffin. Six micron thick
were preserved in 10% formalin in seawater. Details of sections were prepared, stained with hematoxylin-eosin
collection and examination of zooplankton samples are and observed microscopically for lesions.
presented by the junior author in another paper in this
symposium (Polak et al.). Zooplankton samples were
divided so that examination of fresh animals could be Results
made on board and preserved specimens could be made
in the laboratory. For comparative purposes, zooplankton There was a diversity of organisms brought up in the
were preserved in either 10% formalin in seawater alone, dredge on both the February and July cruises, and the
3% glutaraldehyde in seawater alone, 1% osmium predominant macrobenthos; in the area were starfish,
tetroxide in 0.1 m collidine buffer alone, or both the followed by horse mussels, which were often associated
formalin and glutaraldehyde treated specimens were with sponges, tunicates and worms. The fauna are dis-
post-fixed with the osmiurn tetroxide solution. In July, cussed in more detail by Pratt (in these proceedings).

............
.................. .

s
1,17
PP.

z
.........
z

cancer crob stomach

............

Figure 1. Cancer crab found dead with stomach remnant coated 28LO 380 3@10 3519 A80
with Argo oil. WAVELENGTH, nm
Figure 4. Comparison of UV spectrolluorometric analyses of
Cancer crab stomach and Argo Merchant No. 6 standard oil. The
stomach was grossly contaminated with Argo oil in excess of
117 ppm.

40@ vm "Y

IS,

Figure 2. Forceps placed through Cancer crab mouth parts
extended into Argo-oil coated stomach. @J
Figure 5. Appendages of a hermit crab coated with an oil-like
substance.

On the February cruise to the Argo, 250 star-
fish were collected and grossly examined; 33 animals
V
mined histopathologically. Of these, only
were exa
one animal showed any evidence of oil contamination,
which appeared in the form of several discrete black tarry
masses in the buccal cavity. All animals collected were
alive and showed no external oil contamination, nor any
4
behavioral indication that they were unhealthy.
Only one rock crab, Cancer borealis, was col-
lected. This animal was found dead with its carapace only
partially attached to its body. No organs were present
Y
except a large, approximately 21/2 cm wide blackened
pouch (Figure 1). Forceps inserted through the mouth
entered into this pouch which suggests the black pouch
was the remnant of the stomach (Figure 2). When dissected
Q -,,;r out, the entire stomach remnant was observed to be
coated with a blackish-brown tar, the consistency of roof
tar(Figbre3). One square millimeter sample was macerated
in hexane. The spectrof luorometric analysis of the extract
Figure 3. Cancer crab stomach remnant dissected out to show (see technique in Polak et al., these proceedings)
black tarry appearance. indicated that the Cancer crab stomach was heavily con-

......................
@tcirsd-d 8.2 pp.

hermit crab

280 300 310 359 4@10
WAVELENGTH,

Figure 6. Comparison of UV spectrofluorometric analyses of
16
hermit crab digestive tract and Argo Merchant No. 6 standard oil.
The peaks from the hermit crab roughly match the first two peaks
of Argo oil, 8.2 ppm, but the oil constituent of the third peak
se,
was absent.

Z 11..d.rd 3.5ppm

Figure 8. Round, raised, white, calcified nodules on mantle of
............. mussel shell horse mussel. Approximately 5X
..............
............
collected on the February cruise and 16 examined histo-
pathologically. Rainbow-like sheens, characteristic of oil
280 3@6 3@10 359 480 contamination, were found on the shell liquor of three
WAVELENGTH, nm mussels. One of these animals had approximately 20 dis-
Figure 7. Comparison of UV spectrolluorometric analyses of creet black tarry spots, 1-3 mm in diameter, on the inside
mussel shell oil spot scrapings and Argo Merchant No. 6 standard of its shell. Some of these spots appeared to be over-
oil. 7he oil spots apparently consisted of Argo o# at a concentration layered by a thin, nacre-like material. Several spots
slightly less than 3.5 ppm. were scraped off with a metal spatula, extracted in
hexane and examined spectrofluorometrically. As shown
in Figure 7, the wavelength "fingerprint" of the scrapings
taminated withArgo Merchant No. 6 oil, at aconcentration from the shells of the mussel were almost identical to
in excess of 117 pprn (Figure 4). Argo Merchant No. 6 standard oil at 3.5 ppm, suggesting
Only one hermit crab, Pagurus longicarpus, was these spots were remnants of Argo Merchant oil.
collected in February. This animal was alive, but In the places where the overlying mantle of the
moribund, at the time of collection. It fell out of its shell mussel lay in contact with the spots of oil were found
when picked up from the dredge net. It was in- discreet, white, round raised nodules, 1 to several mm in
capable of lifting any of its limbs, and only weak move- diameter (Figure 8). These were rock-hard to the touch of a
ments of its antennae and eye stalks were observed. The metal spatula and required a great deal of pressing force
appendages of the hermit crab were thinly coated with a to crush them. Upon gross examinatiory the nodules
black, oil-like material. There were approximately 1 cm appeared to be homogeneous and to contain nothing more
wide patches of black material near the mouth parts and than a white mineralized material. These were suspected
on the appendages nearest the mouth parts (Figure 5). to be calcifications which are part of the nacrezation proc-
Upon dissection, the digestive tract contained a brown fluid ess that shellfish often produce in response to injury. The
which had an oily-waxy feel. Upon spectrofluorometric nodules dissolved in 40% formic acid, stained blue with
examination, shown in Figure 6, the first two characteris- hematoxylin and eosin, and appeared to be homo-
tic peaks of the Argo oil standard at 8.2 ppm match the geneous upon histopathologic examination (Figures 9 and
peaks of the hermit crab. However, the third large peak at 10). There was no evidence that these nodules contained
360 nm was not present in the hermit crab sample. parasites. The nodules were surrounded by an eosino-
Thirty-two horse mussels, Modiolus modiolus, were philic material that, upon examination at 100OX, was

Figure 9. Cross section of a nodule found on horse mussel Ao@
mantle, partially decalcified in 40% formic acid. Note absence of
parasite and presence of cellular inflammatory response.
Approximately 60X, hematoxylin and eosin.

J1
*4W

Figure 12. Perivascular inflammation. Approximately 250X,
hematoxylin and eosin.

Figure 10. Completely decalcified horse mussel mantle nodule shown to be accumulations of hemocytes (Figure 11).
with associated inflammatory response. Approximately 1OX, These accumulations of hemocytes in other mussels and in
hematoxylin and eosin. higher organisms are regarded to be inflammatory reac-
tions, that is, an acute infiltration of hemocytes in response
to an injury. Inflammatory reactionswere alsofound in other
parts of the body not associated with the oil, especially
around blood vessels (Figure 12). Eggs and sperm were
A A
present in all mussels and there was no gonadal necrosis
found upon microscopic examination (Figure 13). Food was
present in the alimentary tract of all horse mussels
examined, suggesting no impairment of the feeding
mechanism (Figure 13). No histopathological abnormalities
were observed in seven sea scallops collected almost
due east of the Argo Merchant wreck from an area which
was not believed to be impacted by the oil spill.
IV
A second trip was made to the area of the Argo
Merchant oil spill wreck site on July 24, 1977. Again, the
purpose of this trip was to collect sediment and benthic
organisms for comparison with the February 26 samples.
Ar An oil slick of undetermined origin was present in the
- *i4l
WN
Ell-
vicinity of the Argo stern, but no oil was found in sedi-
z ments (Hoffman and Quinn, these proceedings) or on
zooplankt6n (Polak et al., these proceedings).
Figure 11. High power view of inflammatory response showing Over 400 starfish were collected alive and apparent-
hemocytes. Approximately 1 OOOX, hematoxylin and eosin. ly in good physical condition. No gross or microscopic

100

Argo Merchant oil contamination in the immediate area of
the wreck in both sediment and zooplankton as deter-
71
mined by spectrofluorometric analyses of fresh speci-
"Y mens on board. In February, 1977, sediment from 30 sites
were taken and 26% contained Argo oil (Hoffman and
Quinn, these proceedings). Of 22 zooplankton sampling
sites, 32% had a trace amount of Argo oil and 64% were
heavily contaminated (Polak et al., these proceedings).
Our findings of oil contamination in two crabs and several
V 21, mussels also indicated Argo oil reached the bottom and
persisted in the area of the wreck.
It is interesting to note that the hermit crab found in
February, although contaminated with an oil, did not have
the characteristic Argo Merchant No. 6 UV absorption
spectrum. The characteristic third peak of Argo oil was
not present in the hermit crab. This suggests that the
hermit crab may have been exposed to either the Argo
Figure 13. Horse mussel gonad is replete with eggs and intestine cutter stock, which was the lighter oil used to dilute the
contains food. Approximately 160X, hematoxylin and eosin. No. 6, or it may mean that the hermit crab was able to
partially metabolize some of the Argo No. 6 oil, or the
lesions were discovered upon histopathological exami- hermit crab may have been contaminated by some other
nation of 44 starfish. oil of unknown origin present in the area.
Of the 24 horse mussels collected, 12 were exam- There were marked differences in our findings in
ined histopathologically. Approximately 60% of the February and July. In February, there was visible oil
adductor muscle of one horse mussel was covered with the contamination of zooplankton, two crabs, and several
same type of round, raised, white, apparently calcified mussels, while there was no visible oil contamination of
nodules as described above. However, the organism's those species collected in July. In February, one mussel
shell was intact and closed upon collection, suggesting was found with discreet calcified, inflamed lesions. In
the lesion did not severely affect the function of the July, one mussel was found with calcification, but not
adductor muscle. Also, upon histopathologic examina- inflammation. In February, there was evidence of oil
tion, these calcified nodules were not accompanied by impact on the crab population. One crab was found dead
any inflammatory response, suggesting that this lesion and one crab was moribund and both were contaminated
was a response to an injury that had occurred earlier in with oil. In July, the findings did not indicate any crab mor-
the life of the animal and that had since healed. It is not tality in the area. The eighteen crabs collected were clean
known whether this lesion was formed in response to and apparently healthy, as compared to finding only two
exposure to Argo Merchant oil, although it is certainly crabs which were severely affected in February. It is not
possible, as it was similar to the one lesion found in known whether the increased number of crustaceans found
February that was apparently the result of the Argo in July reflect sampling variability, seasonal changes in
Merchant oil. Again, as in February, there were no abnor- the populations, or whether they reflect a return of these
malities found in the reproductive organs of the horse organisms which might have fled in response to the Argo
mussels collected in July. Merchant oil spill.
In contrast to the February samples, the July In both February and July there was no microscopic
samples contained a greater number of crustaceans evidence of impairment of gonadal development. This
does not rule out the possibility of mutagenic or terato-
Eight rock crabs and 10 hermit crabs were collected in genic effects, which apparently could have occurred to
July as compared to only one of each in February. These
organisms were collected alive and appeared to be in fish eggs exposed to Argo oil (Longwell, these
good physiological condition, which contrasted to the proceedings).
organisms found in February that were dead or moribund. Based on our observations and investigations on
Also in July there were 5 sea cucumbers collected where two visits to the Argo wrecksite, the authors are led to
there were none collected in February. The area where believe the level of oil toxicants that reached the bottom
the scallops were collected in February was not visited were, for the most part, within the physiological toleration
in July. limits of the macrobenthos.
The apparent minor effects of the Argo oil spill can
be explained in terms of established toxiocological princi-
Discussion ples. That is, the toxicity of a particular substance to an
organism is dependent upon the amount and duration of
Argo oil was found to be persistent in the environ- exposure.
ment of the Argo wreck for at least two months after the The immense volume of ocean can be considered to
spill. The oil had an impact on the macrobenthos, but less have acted as an extremely large diluent of the potentially
so in July than in February. Although the effects will never toxic oil, thus reducing the amount of oil that reached the
be entirely known, and although there may possibly macrobenthos. Also, the ocean currents are in a constant
develop chronic, long term effects, there is reason to state of flux in the shoal area. The tidal flow rates through
believe the effects of the Argo on the macrobenthos the Nantucket Shoals are great (Pratt, these pro-
were minor. ceedings), and there were storms with high seas following
Even though collections were made more than two the spill that further acted to disperse the oil, and reduce
months after the wreck itself, there was still evidence of the time of exposure to it. Therefore, it is likely that only

101

a small portion of the Argo oil ever reached th-e bottom, of Blumer, M., H. L. Sanders, J. F. Grassle and G. R. Hampson.
which a smaller portion still would have impacted the 1971. A Small Oil Spill. Environment, March, 13, 2-12.
macrobenthos. The results obtained by Hoffman and Blumer, M. and J. Sass. 1972. The West Falmouth Oil Spill, Data
Quinn (these proceedings) indicated that little oil actually available in November, 1971. 11. Chemistry. Woods Hole
was found in the bottom sediments supporting macro- Oceanographic Institution Technical Report No. 72-19.
benthos. Our own results indicated that oil reached the Blumer, M., J. M. Hunt, J. Atema and L..Stein. 1973. Interaction
between marine organisms and oil pollution. Project
bottom, but affected only six animals of hundreds col- # 18050 and 18080 EBN, Report Number EPA-R3-73-042,
lected. It can also be speculated that the effects were prepared for the Off ice of Research and Monitoring, U.S.
minor because the majority of animals were able to Environmental Protection Agency, Washington, DC
escape exposure to what oil did reach the bottom. The 20460. 97 pp.
crustaceans could have moved out of the spill area and Boesch, D. F., C. H. Hershner and J. H. Milgram. 1974. Oil spills
the mussels could have remained closed until the oil was and the marine environment. Ballinger Publ. Co., Cam-
removed by tidal action. This could explain why so few bridge, Mass. 114 pp.
affected animals were found and why there were many Carthy, J D. and D. R. Arthur. 1968. The biological effects of oil
more crustaceans present in July. In thefinal analysis, it is pollution on littoral communities. Field Studies Council, 9
Deversox County, Strand, London, W.C. 2. 198 pp.
impossible to reach a conclusion as to the magnitude of Hoffman, E. J. and J. G. Quinn. 1978. A comparison of Argo Mer-
the effect of the Argo Merchant oil spill. Only a very small chant oil and sediment hydrocarbons from, Nantucket
portion of a very large area that was impacted was Shoals. These Proceedings.
sampled, and our samplings did not begin until two Longwell, A. C. 1977. A genetic look at fish eggs and oil. Oceanus
months after the spi 11. This, of course, was due to the large 20:4:46-58.
number of scientists and observers interested in studying National Academy of Sciences. 1975. Petroleum in the Marine
the spill and the limited number of research vessels and Environment. Report of a workshop held under the
available space on them to accommodate everybody. auspices of the Ocean Affairs Board, National Academy
When interpreting our results, keep in mind that there is of Sciences.
Polak, R., A. Fillion, S. Fortier, K. Cooper and J. Laniel. 1978.
no information on the status of the benthos at the time of Observations on Argo Merchant oil in zooplankton of
the spill or immediately after, when acute toxic effects Nantucket Shoals. These Proceedings.
would have most likely occurred. In the case of future oil Pratt, S. D. 1978. Interactions between petroleum and benthic
spills, it is recommended that there be sufficient contin- fauna at the Argo Merchant spill site. These Proceedings.
gency planning to get as many investigators to the scene Sanders, H. 1977. The West Falmouth oil spill - Florida, 1969.
as quickly as possible so that the degree of effect can be Oceanus 20:4:15-25.
precisely determined. Stegeman, J. J. 1974. Hydrocarbons from shellfish chronically
exposed to low levels of fuel oil. In Pollution and Physiol-
ogy of Marine Organisms. Academic Press, N.Y. pp. 329-
Acknowledgments 347.
Stegeman, J. J. and D. J. Sabo. 1974. Uptake and release of
This paper is contribution #1788 of the Rhode petroleum hydrocarbons by marine organisms and some
Island Agricultural Experiment Station sponsored in part metabolic implications. Proc. Estuarine Res. Fed., Outer
by funds from Sea Grant. Continental Shelf Conference, December, 1974.
Stegeman, J. J. 1977. Fate and effects of oil in marine animals.
Oceanus 20:4:59-66.
References Vandermeulen, J. H. 1977. The Chedabucto Bay spill - Arrow,
1970. Oceanus 20:4:31-39.
Anderson, J. W. 1975. Laboratory studies on the eff ects of oil on
marine organisms: an overview. American Petroleum
Institute, Division of Environmental Affairs, publication
No. 4249.
Anonymous. 1973. Oil spill, Long Island Sound, March 21,1972.
Environmental effects. Final report. Order No. 001, Con-
tract No. 68-01-0044, Division of Oil and Hazardous
Materials, Office of Water Program Operations, Environ-
mental Protection Agency, Washington, DC 20460.
135 pp.
Atema, J. 1977. The effects of oil on lobsters. Oceanus 20A
67-73.
Barry, M. and P. P. Yevitch. 1975. The ecological, chemical and
histopathological evaluation of an oil spill site. Ill. Histo-
pathological studies. Mar. Pollution Bull. 6:11:171-173.
Bender, M. E., J. L. Hyland and T. K. Duncan. 1974. Effect of an oil
spill on benthic animals in the lower York River, Virginia.
NBS Spec. Publ. 409, Marine Pollution Monitoring (Petro-
leum) Proceedings of a Symposium and Workshop held at
NBS, Gaithersburg, MD, May 13-17,1974, pp. 257-259.
Blumer, M., G. Souza and J. Sass. 1970. Hydrocarbon Pollution
of Edible Shellfish by an Oil Spill. Marine Biology. 5,
195-202.
Blumer, M., G. Souza and J. Sass. 1970. Hydrocarbon Pollution
of Edible Shellfish by an Oil Spill. Report to the Board of
Selectmen, Town of Falmouth, Mass., Unpublished Manu-
script, Woods Hole Oceanographic Institution, Refer-
ences *70-1.

102

Some Physiological Effects of the Argo
Merchant Oil Spill on Several Marine
Teleosts and Bivalve Molluscs

Frederick P. Thurberg, Edith Gould, and Margaret A. Dawson

National Marine Fisheries Service
Northeast Fisheries Center
Milford Laboratory
Milford, Connecticut

Abstract studies (Emery, 1972; Moore and Dwyer, 1974; Anderson,
1977). By contrast, measurements performed on animals
Subsequent to the oil spill from the tanker Argo taken from their natural habitat following an oil spill are
Merchant, two cruises provided animals for physiological rare, and grounding of the tanker Argo Merchant with its
and biochemical testing. Blood samples were taken from subsequent release of 7.7 million gallons of No. 6 fuel oil
a variety of teleost species, and although the sample into the surrounding waters provided an unusual
number for most species was too small forvalid statistical opportunity for a series of such measurements.
analysis, there did appear to be a disruption of serum ions This report, therefore, attempts to compare some
in winter flounder, Pseudopleuronectes arriericanus, aspects of the physiology and biochemistry of several
yellowtail flounder, Limanda ferruginea, and haddock, bivalve mollusc and finfish species taken from oil-
Melanogrammus aeglefinus. Serum osmolality, sodium, impacted areas to those of animals taken from adjacent
and potassium values were variously depressed in fish areas untouched by the oil spill. We were able to collect
collected from oil-impacted areas, as compared to fish only limited numbers of any one species, however, and we
from clean or unimpacted areas. present our data here with correspondingly limited
Ocean scallops, Placopecten magellanicus, and statistical confidence. Nevertheless, because post oil-
horse mussels, Modiolus modiolus, collected during the spill physiological data, however incomplete, are rare,
first cruise from oil-impacted areas had depressed gill- and because the data presented here do suggest some
tissue oxygen consumption, but normal values were possibly vaiuable indicators of metabolic stress, we feel
recorded from scallops collected during a second cruise that the information will be a useful supplement to the
six weeks later. Serum sodium and calcium levels of existing oil-exposure literature and may form a base for
scallops collected from oil-impacted areas during the future studies of a similar nature.
second cruise were elevated, as compared to scallops A number of physiological processes have been
collected from clean areas. measured in recent years as indicators of sublethal
Malic dehydrogenase activity of scallop muscle was stress. Among these, the parameter that has received the
significantly decreased in scallops from oil-impacted greatest attention in the literature is the rate of oxygen
areas. Lactate oxidation was also significantly lower in consumption. A review by Anderson (1977) recently
these animals, although pyruvate reduction, catalyzed by emphasized the usefulness of respiratory measurements
the same enzyme, remained the same. Both observations in oil pollution studies. Measurements of respiration have
suggest a possible weakening of the ability to shift to also been recommended by researchers of the NSF-
anaerobiosis. IDOE Biological Effects Program as a valuable indicator
of pollutant stress (Giam, 1977). Hematological effects of
Introduction pollutants have received less emphasis; however,
several recent publications have demonstrated the use-
The eff ects of petroleum hydrocarbons on a variety fulness of this type of measurement in physiological
of marine animals have been studied extensively by studies related to environmental pollution (Calabrese
means of controlled laboratory exposures. Several et al., 1975; Dawson, in press).
recent publications provide extensive reviews on these Enzymes involved in carbohydrate metabolism are

103

said to constitute the largest fraction of enzymes affected For biochemical testing, the adductor muscles were
by acute oil exposure (Heitz et al., 1974). The posterior excised from the scallops that were transported live to the
adductor muscle of the blue mussel, Mytilus edulis, for laboratory, then kept frozen (-29*C) until analysis. Other
example, was depleted of glycogen stores after the scallops were frozen whole at sea just after collection and
animal's experimental exposure to oil, a loss of energy transferred frozen to the laboratory. They were thawed
that is normally held for use in gametogenesis, and that is just enough to enable excision of the adductor muscles,
untapped even during prolonged starvation (Dunning and which were then packaged and kept frozen under the
Major, 1974). The object of our biochemical testing in this same conditions. For enzyme measurements, the muscle
study, therefore, was to measure the relative activities of samples were made 1:1, w/v, with an iced 0.88 M sucrose
regulatory glycolytic enzymes, as well as representative solution containing dithiothreitol (1 mM). The preparations
enzymes of the tricarboxylic acid cycle, the pentose were vigorously ground in a glass homogenizer with
shunt, and nitrogen metabolism. Most closely examined 20-30 mg of 25- p m glass powder, and the resulting paste
in bivalve adductor muscle were both forward and reverse centrifuged for 1 hr at 4'C and 17,000 x g. After centrifu-
reactions of malate dehydrogenase (E.C. 1.1.1.37; MDH) gation, the clear supernates (2X prep) were transferred
and D-lactate dehydrogenase(E.C. 1.1.1.28; LDH). Induc- with Pasteur pipettes to cold test tubes and kept on ice
tion of glycolytic enzymes generally reflects a mobiliza- throughout the analysis. For the malate oxidation (MDH)
tion of energy reserves, and induction of pentose shunt and both LDH reactions, the preparations were diluted
enzymes reflects increased biosynthetic rates, both of 1:4 for a final 10-fold dilution (10X prep), and for oxalo-
which might reasonably be expected, at least initially, in acetate reduction (MDH), the tissue preparations were
sublethally stressed animals. During anaerobiosis in diluted 1:49 for a final 100-fold dilution (10OX prep).
invertebrates, MDH and LDH activities are respectively Preparations of teleost tissues were: pooled brain
induced and repressed (Hochachka and Somero, 1973), samples, 1:9 in H20, w/v; pooled kidney samples, 1:9 in
a pattern that has also been observed in the tissues of H20, w/v; and gonad samples, 1:4 in 0.88 M sucrose -
lobsters, Homarus americanus, stressed by low salinity 1 mM DTT*, w/v. Centrifugation was at 4C and 28,000 x g
(Gould, unpub.) for 60 min for kidney and brain preparations, and at 40,000
x g for 45 min for gonad preparations. Tissue pools were
from 2 animals each.
Methods All solutions were made with doubly glass-distilled
Teleost and molluscan samples for this study were water, and all reaction rates were read at 340 nm with a
collected during the 4-10 January 1977 cruise of the double-beam ratio-recording sp6ctrophotometer, cham-
NOAA R/V Delaware // and the 18-26 February 1977 ber temperature 25'C, in cuvettes having a 1 0-mm path-
Polish R/V Wieczno cruise. Samples were taken from length. Reaction volume was 3.00 ml. Reaction rates were
both oil-impacted and adjacent clean areas. Ocean followed on a linear-log potentiometer recorder and
scallops (P/acopecten magellanicus) and horse mussels calculated from the fastest portion of the curve.
(Modiolus modiolus) were collected with a Digby dredge For the assays, reactant concentrations (mM) were:
and maintained alive on the ship in running seawater. (1) oxaloacetate reduction (MDH): buffer, pH 9.0,90
They were transported to the Milford Laboratory on ice glycine and 0.9 -disodium EDTA* dihydrate; 0.15 reduced
and then transferred to running ambient seawater (about NAD*; 1.0 oxaloacetic acid, cis-enol form; and 0.10 ml
26-28 ppt salinity, 2-5C) for up to one week prior to 10OX enzyme prep to start;
testing. (2) malate oxidation (MDH): same buffer as for (1);
For oxygen-consumption studies, a single gill was 0.30 NAD; 6.7 L-malic acid,. neutralized with KOH; and
dissected from each bivalve and placed in a 15-ml 0.10 ml 1OX enzyme prep to start;
Warburg-type flask. Each flask contained 5 ml of sea- (3) pyruvate reduction (LDH): 90 phosphate buffer,
water at a salinity of 27 ppt. Oxygen consumption of each pH 7.5; 0. 10 reduced NAD; 10 sodium pyruvate; and 0. 10
gill was monitored over a 4-hr period in a Gilson ml 1OX enzyme prep to start;
differential respirometer at 100C. Oxygen consumption (4) lactate oxidation (LDH): 90 Tris* buffer, pH 8.0;
rates were calculated as microliters of oxygen consumed 0.30 NAD; 133.3 DL-lactic acid, lithium salt; and 0.10 ml
per hour per gram, dry weight, of gill tissue (M t 02/hr/g), 1OX enzyme prep to start;
corrected to microliters of dry gas at standard tempera- (5) malic enzyme (ME) (E.C. 1.1.1.40):81.3 HEPES*
ture and pressure. buffer, pH 8.0; 1 .0 MnC12; 0.4 NADP*; 0.20 ml 2X enzyme
For hematological measurements, blood was col- prep; and 6.7 L-malic acid, neutralized with KOH, to startt
lected from scallop 'adductor muscles using a 3-ml ME reaction. ME activity was corrected for endogenous
syringe and a 20-gauge needle. Each blood sample was NADP reduction by subtracting the reaction rate
centrifuged for5 min at 12,000x g andthe serum removed measured before addition of the malate substrate.
for measurements of sodium, potassium, calcium, and All data were analyzed statistically by using the
osmolality. Sodium, potassium, and calcium were Student's 'T' test.
analyzed with a Coleman Model 51 flame photometer.
Serum osmolality was measured on an Advanced Model Results
3L osmometer using 0.2-ml samples. Teleost blood
samples were taken by cardiac puncture from live fish The mean gill-tissue oxygen consumption rate of
immediately after trawl capture. Again, a 3-ml syringe with scallops collected from sites within the spill area during
a 20-gauge needle was used and the sample centrifuged the January cruise was 594 Wt 02/hr/g(n= 7) (Figure 1).
at 12,000 x g. The samples were then frozen for later This value is considerably lower than the 675 A t02/hr/g
analysis at the Milford Laboratory. Serum sodium, (n = 5) rate found in scallops from adjacent unimpacted
potassium, calcium, and osmolality were measured using areas. The respiratory rates of scallops collected from
the same instrumentation as for the scallop serum. oil-contaminated areas in February had a mean value of

104

SCALLOP GILL mean value of 596 A I 02/hr/g of 6 mussels collected
from an adjacent clean area. No mussels were collected
700- during the February cruise.
Serum samples taken from scallops from oil-
0) impacted areas during the February cruise showed some
serum ion disruption. Both serum sodium and calcium
600- levels were significantly elevated, whereas osmolality
ZL and potassium were unaffected (Table 1).
MDH activity was significantly depressed in the
adductor muscle of scallops frozen whole at sea after
Z 500
0 being taken from oil-impacted areas, in comparison to the
0 Z Z scallops from clean areas (Table 2). Lactate oxidation
CIA < Ic was also lower, but pyruvate reduction, catalyzed by the
0 _J ui _J UJ same enzyme (LDH), remained unchanged. Malic enzyme
J
Q activity was also significantly depressed. For scallops
brought back live to the laboratory and dissected there for
analysis, however, there were no significant differences
between animals from.oil-impacted areas and those from
clean areas, when the same activities in the same tissue
JAN FEB were compared (Table 3). The only observation of statisti-
cal significance in these animals was that lactate oxida-
Figure 1. Gill-tissue oxygen consumption values of sea scallops, tion increased 20- to 30-fold over animals frozen whole at
Placopecten magellanicus, taken from oil-impacted areas (n = 7 sea (627 and 469, as compared with 36 and 14 in Table 2),
and 9) and adjacent clean areas (n = 5 and 9) in January and
February, 1977. while pyruvate reduction decreased by almost half (83, as
compared with 141 and 159). As a result, the PR/LO
values averaged 0.2, in contrast to 3.9 and 14.3 for
651 At 02/hr/g (n = 9) and were similar to the January animals frozen immediately at sea after being taken from
control values from unimpacted areas. The February clean or from oil-impacted areas, respectively.
Control site animals had a mean rate of 654 Of the teleost species tested biochemically, most
A t 02/hr/g (n = 9). The standard errors for those four belonged to one of three families: the pleuronectids, the
groups were similar, ranging from 28-34. Only 2 mussels clupeids, and the gadoids. The first group was repre-
were collected from an oil-contaminated area during the sented by yellowtail flounder (Limanda ferruginea) and
January cruise. The mean oxygen-consumption rate of winter flounder (Pseudopleuronectes americanus); 6
gills from these bivalves was 451, a value lower than the animals each were taken from clean areas and 6 each
from oil-impacted areas. Pools were from 2 animals each
for kidney and for brain. Gonads were prepared indivi-
Table 1. Serum measurements made on ocean scallops, Placopecten dually, because of the variables of sex and degree of
magellanicus, from oil-impacted areas and adjacent clean areas. gonad maturation. The clupeids were represented by the
0smolality values are in mOsm/kg; others are in meq/1. alewife (Pomolobus pseudoharengus), 3 fish from clean
and 6 fish from oil-impacted areas, and the blueback
Tes t Clean N: 18 Oil N:26 P (Pomolobus aestivalis), 1 specimen only, from a clean
area. The third group was represented by cod (Gadus
osmolality 819 SE 2 818 SE 2 NS morhua), 4 fish from clean areas and 1 from an impacted
Sodium 365 2 380 2 P<.001 area, and haddock (Melanogrammus aeglefinus), 1 from a
Potassium 12.5 0.3 12.5 0.3 NIS clean and 2 from an impacted area. Single specimens
Calcium 16.5 0.2 17.5 0.2 P<.05 were also taken of other species representing other
families. The brain, kidney, and gonad preparations were

Table 2. Sea scallops, Placopecten magellanicus, collected from clean and from oil-impacted areas during first cruise of Delaware // in
January, 1977, frozen whole at sea, and thawed for dissection and analysis in April, 1977. Units of enzyme activity: reduction,
11moles NADH oxidized/min/mg protein; oxidation, Mmoles NAD reduced/min/mg protein; malic enzyme, Mmoles NADP
reduced/min/mg protein.

Clean MW Oil-impacted W 10)
Enzyme 7 SE (range) SE (range) P

MDH: oxaloacetate reduction 3128 423 M62-5014) ills 92 (825-1651) <0.001
MDH: malate oxidation 59 15 (34-131) 18 2 (10-29) <0.05
OAR/MO 58.5 5 (38.3-70.3) 69.2 8.8 (32.8-127.0)
LDH: pyruvate reduction 141 24 (67-233) 159 11 (119-237)
LDH: lactate oxidation 36 9 (24-71) 14 2 (5-24) <0.0 1
PR/LO 3.9 0.9 (1.5-6.4) 14.3 2.1 (5.7-23.8) <0.01
MDH (OAR)/LDH (PR) 26.1 5.7 (11.2-46M 7.3 0.7 (4-11) <0.001
Malic enzyme 23 4.3 (10.6-38.5) 11.7 1.1 (5.5-17.8) <0.01

105

Table 3. Scallops, Placopecten magellanicus, collected from clean and from oil-impacted areas during first cruise of Delaware if in
January, 1977. They were maintained alive on the ship in running seawater (3-4 days), returned to the laboratory on ice (5 hr),
and transferred to running ambient seawater for up to 1 wk before dissection and analysis. Units of activity are as in Table 1.

Clean Oil-impacted
Enzyme activity (N) X SE (range) (N) X SE (range)

LDH: pyruvate reduction (13) 83 24 15-271) (6) 83 12 43-131)
LDH: lactate oxidation 7) 627 63 (414-913) (5) 469 31 (416-582)

Table 4. Serum measurements made on yellowtail flounder, Table 6. Serum measurements made on haddock, Melanogrammus
Limanda ferruginea, and winter flounder, Pseudopleuronectes aeglefinus, from oil-impacted areas and adjacent clean areas.
americanus, from oil-impacted areas and adjacent clean areas. 0smolality values are in mOms/kg; others are in meq/1.
Osmolality values are in mOsm/kg; others are in meq/1.

Test Clean N:6 Oil N: 14 P
Yellowtail flounder serum
Tes t Clean N:7 oil N: 17 P Osmolality 453 SE 31 385 SE 5 P<.05
Sodium 200 15 179 3 P<.05
osmolality 473 SE 34 425 SE 11 P<.05 Potassium 8.31 1.20 7.36 0.38 P<.05
Sodium 208 12 193 4 P<.05 Calcium 3.92 0.92 3.83 0.50 NS
Potassium 7.26 1.10 7.07 0.73 NIS
Calcium 6.30 0.38 5.97 0.30 NS
Winter flounder serum each tissue-specific enzyme seemed to fall within the
same range for each teleost family (gadoids, clupeids,
Clean N:5 ioil N:5 P pleuronectids) over all collection sites.
Sodium 193 SE 7 205 SE 2 NS Results of the teleost serum studies are presented
Potassium 7.08 0.41 3.97 0.13 P<.001 in Tables 4, 5, and 6. Sodium was depressed in yellow-
Calcium 5.74 0.40 5 0.38 NS tail flounder from oil-impacted areas, as was osmolality;
winter flounder sodium was elevated, but osmolality was
not measured (Table 4). No differences in osmolality,
sodium, or potassium were detected in either alewives or
herring when fish from oil-impacted and clean areas were
Table 5. Serum measurements made on herring, Clupea harengus, compared (Table 5). The limited number of haddock
and alewife, Pomolobuspseudoharengus, from oil-impacted areas tested showed depressed serum osmolality, sodium, and
and adjacent clean areas. 0smolality values are in mOsm/kg; potassium levels in fish collected from oil-impacted
others are in meq/1. areas, as compared to fish from an adjacent unimpacted
area (Table 6).

Herring serum
Tes t Clean N:9 oil N:7 P Discussion
osmolality 530 555 NS The respiratory results of this study give some
Sodium 245 241 NIS indication of stress due to oil exposure. The seven
Potassium 7.44 9.45 NS scallops from oil-impacted areas showed considerable
Calcium 6.51 6.66 NS depression of gill-tissue oxygen consumption when com-
pared to scallops from adjacent clean areas. The small
Alewife serum number of bivalves sampled precluded meaningful
Clean N:6 oil N:9 P statistical analysis, but does give an indication of oil-
related respiratory stress. Other investigations have
osmolality 527 497 NS reported respiratory alterations in various marine
Sodium 230 228 NS animals, including bivalve molluscs, after exposure to oil
Potassium 9.24 7.97 NS or oil fractions (Brocksen and Bailey, 1973; Avolizi and
Nuwayhid, 1974; Gilfillan, 1975; Percy, 1977). Scallops
collected one month later showed no signs of respiratory
stress. They may have recovered from oil exposure or
variously examined for glucose phosphate isomerase may never have been exposed even though they were
(E.C.5.3-1-9), pyruvatekinase(E.C.2.7.1.40),glucose-6- collected from areaswhere an oil slick had been reported.
phosphate dehydrogenase (E.C.1.1.1.49), isocitrate A recovery seems more probable; other investigators have
clehydrogenase (E.C.1.1.1.42), and aspartate amino- reported respiratory recovery in marine animals following
transferase (E.C. 2.6.1.1), as well as for MDH and LDH. their return to clean water after oil exposure (Brocksen
Although there were insufficient data for each species for and Bailey, 1973; Anderson et al., 1974). These scallops
statistical comparison of clean and oil-impacted areas, had elevated plasma sodium and calcium levels, an

106

observation suggesting that they had indeed been seawater aboard ship until brought to shore (2-3 days),
exposed to the oil. then transported on ice to the Milford Laboratory (5 hrs),
Because bivalves in general are osmoconformers, it where the animals were placed in flowing seawater for up
is not surprising that oil exposure did not affect serum to one week before dissection. It is a plausible assump-
osmolality in scallops. Anderson and Anderson (1975) tion that the animals at least partially cleared themselves
reported similar results for the American oyster using a during the shipboard holding period, and that they shifted
short-term laboratory exposure to oil. Although total toward anaerobiosis during their transport to the labora-
osmolality did not change in our oil-impacted animals, tory. There was a greater than 20-fold increase in lactate
there were changes in some of the components. Serum oxidation in these animals, as compared with the animals
sodium rose from 365 meq/1 to 380 meq/1 in animals frozen at sea, and a drop in pyruvate reduction, both
from the oil-impacted area. Calcium also increased in oil- mechanisms preventing the dead-end accumulation of
impacted animals to 17.6 meq/1, compared to 16.9 lactic acid during anoxic glycolygis. It is our belief that
meq/1 in controls. data from animals frozen at sea more accurately repre-
In view of the suggestion by Sabo and Stegeman sent the biochemical state of these animals in their
(1977) that oil may alter membrane structure, teleost native habitat, and our future sampling procedures will be
osmo- and ion-regulation would appear to be among the based on this observation.
more reasonable areas of study for future oil-spill No differences could be detected in enzyme activi-
research. However, information on how oil exposure ties of kidney, brain, or gonad taken from fish collected
affects these processes in any animal is extremely from clean areas, as compared with those of fish taken
limited. The fish studied exhibited a species-variable from oil-impacted areas. Each tissue-specific enzyme fell
response to oil exposure. Two of them, the herring and within the same range for each teleost family (gadoids,
the alewife, had no significant oil-induced changes in any clupeids, pleuronectids) over all collection sites,
of the variables tested. In the species that showed a however, and the information will be used to form the
change, serum osmolality, sodium, or potassium went nucleus of a baseline data bank. In this respect, it will be
from a point below that of the surrounding seawater to an important to have consistent sampling of a selected
even lower point. In fishes tested for serum calcium, there number of representative species at different seasons of
was no change. The possibility of utilizing calcium stored the year, together with information on water temperature,
in bone and other tissues may make calcium regulation in salinity, and depth of catch.
fish less susceptible to disruption by oil exposure than The foregoing results indicate that bivalves and
sodium and potassium regulation (Fleming, 1967; teleosts from an area perturbed by an oil spill experience
Fleming, Brehe, and Hanson, 1973). a variety of physiological alterations. Although it would be
The bivalve adductor muscle is a sensitive indicator scientifically unsound to draw any definite conclusion or to
of anaerobiosis - more so than mantle tissue (Livingstone, make any general statements or assumptions from such
1976) - and lends itself especially well to biochemical limited data, it would be equally unwise to ignore the
examination. Of the bivalves collected, only the scallop indications of stress that show in these data. Similar
was taken in sufficient numbers for statistical analysis. testing must be conducted after future oil spills before
Some were frozen whole at sea immediately after collec- definitive conclusions can be made. The Northeast
tion, and others were brought back to the laboratory for Fisheries Center of the NMFS is initiating a program of
analysis, a circumstance that allowed a practical physiological monitoring to coordinate and carryout such
comparison of methods. studies. This program, called Ocean Pulse, is designed to
The lowered activities of MDH and ME observed in monitor and assess the health of the ocean's living
the adductor muscle of sea-frozen scallops taken from resources on the continental shelf of the northeast
oil-impacted areas, as compared with those taken from Atlantic (Pearce, 1977; Sherman, 1977). Such an effort
"clean" areas (Table 3), could be attributed to any of will coordinate short-term assessment studies in
several possible factors: 1) inhibition of the animal's response to a pollution event such as the Argo Merchant
capacity to maintain a normal metabolism (MDH repres- oil spill, and provide for long-term physiological monitor-
sion); 2) a slowed feeding rate (low ME), either in concert ing of a variety of representative marine species. The
with or as a consequence of a lowered metabolic rate, results obtained in this study are valuable as "test
another sign of which was the depressed rate of gill- phase" data and will be used as such in the Ocean Pulse
tissue oxygen consumption; or even 3) aweakening of the program planning and execution.
ability to shift to anaerobiosis (MDH and lactate oxida-
tion down, but pyruvate reduction unchanged), in defense Note: Use of trade names is to facilitate description and
against ingestion of toxicants. Hammen (1969) has does not imply endorsement by the National Marine
pointed out that the PR/LO ratio indicates the probability Fisheries Service, NOAA.
of lactic acid production, which must be shut down during The authors thank Miss Rita S. Ri ,ccio for her critical
anaerobiosis. We hesitate, however, to draw any firm reading and typing of this manuscript.
conclusions. Although the data are statistically valid, the
sample number is small, and until these observations can
be confirmed by future measurements under similar
circumstances, no great significance should be attached
to them.
More useful at this point, perhaps, is the comparison *Abbreviations used are: EIDTA, ethylenediarnine tetraacetic
of the data obtained from animals frozen whole at sea after acid; NAD, nicotinamide adenine dinucleotide; NADP, nicotin-
capture (Table 2) with the data obtained from scallops amide adenine dinucleotide phosphate; Tris, tris (hydroxy-
returned live to the laboratory for dissection and analysis methyl) aminomethane; IDTT, dithiothreitol; HEPES, (n-2 hydro-
(Table 3). The lafter group of scallops was held in clean xyethyl)-piperazine (n'-2-ethane) sulfonic acid.

107

References Pearce, J. B. 1977. A report on a new environment assess-
ment and monitoring program, Ocean Pulse. ICES, C.M.
Anderson, J. W. 1977. Responses to sublethal levels of petro- 1977/E:65.
leum hydrocarbons: are they sensitive indicators and do Percy, J. A. 1977. Eff ects of dispersed crude oil upon the respira-
they correlate with tissue contamination? p. 95-114. In tory metabolism of an arctic marine amphipod, Onisimus
D. A. Wolfe (ed.), Fate and effects of petroleum hydro- (Boekisimus) affinis. p. 192-200. In D. A. Wolfe (ed.), Fate
carbons in marine organisms and ecosystems. Pergamon and effects of petroleum hydrocarbons in marine
Press. organisms and ecosystems. Pergamon Press.
Anderson, J. W., J. M. Neff, B. A. Cox, H. E. Tatem, and G. M. Sabo, D. J., and J. J. Stegeman. 1977, Some metabolic effects
Hightower. 1974. The effects of oil on estuarine animals: of petroleum hydrocarbons in marine fish. p. 279-287. In
toxicity, uptake and depuration, respiration. p. 285-310. In F. J. Vernberg, A. Calabrese, F. P. Thurberg, and W. B.
F. J. Vernberg and W. B. Vernberg (eds.), Pollution and Vernberg (eds.), Physiological responses of marine biota
physiology of marine animals. Academic Press. to pollutants. Academic Press.
Anderson, R. D., and J. W. Anderson. 1975. Effects of salinity and Sherman, K. S. 1977. The Argo Merchant oil spill and fishery
selected petroleum hydrocarbons on the osmotic and resources of Nantucket Shoals: a preliminary assessment
chloride regulation of the American oyster, Crassostrea of impact. ICES, C.M. 1977/E:58. p. 1-39.
virginica. Physiol. Zool. 48:420-430.
Avolizi, R. J., and M. Nuwayhid. 1974. Effects of crude oil and
dispersants on bivalves. Mar. Pollut. Bull. 5:149-152.
Brocksen, R. W., and H. T. Bailey. 1973. Respiratorv resr)onse of
juvenile chinook salmon and striped bass exposed to
benzene, a water-soluble component of crude oil. Proc.
Conf. on Prevention of Oil Spills. A.P.I., E.P.A., U.S.C.S.,
Washington, D.C.
Calabrese, A., F. P. Thurberg, M. A. Dawson, and D. R. Wenzloff,
1975. Sublethal physiological stress induced by cadmium
and mercury in the winter flounder, Pseudopleuronectes
americanus. p. 15-21. In J. H. Koeman and J. J. T. W. A.
Strik (eds.), Sublethal effects of toxic chemicals on
aquatic animals. Elsevier.
Dawson, M. A. In press. Hematological effects of long-term
mercury exposure and subsequent periods of recovery on
the winter flounder, Pseudopleuronectes americanus. In,
W. B. Vernberg, A. Calabrese, F. P. Thurberg, and F. Jl@
Vernberg (eds.), Marine pollution - functional responses.
Academic Press.
Dunning, A., and C. W. Major. 1974. The effect of cold seawater
extracts of oil fractions upon the blue mussel, Mytilus
edulis. p. 349-366. In F. J. Vernberg and W. B. Vernberg
(eds.), Pollution and physiology of marine organisms.
Academic Press.
Emery, A. R. 1972. A review of the literature of oil pollution with
particular reference to the Canadian Great Lakes. Res.
Information Paper (Fisheries) No. 40. Ministry of Nat.
Resources, Ontario.
Fleming, W. R. 1967. Calcium metabolism of teleosts. Am. Zool.
7:835-842.
Fleming, W. R., J. Brehe, and R. Hanson. 1973. Some compli-
cating factors in the study of calcium metabolism of
teleosts. Am. Zool. 13:793-797.
Giam, C. S. led.). 1977. Pollution effects on marine organisms.
Lexington Books.
Gilfillan, E. S. 1975. Decreaseof netcarbonflux intwospeciesof
mussels caused by extracts of crude oil. Mar. Biol.
29:53-57.
Hammen, C. S. 1969. Lactate and succinate oxidoreductases in
marine invertebrates. Mar. Biol. 4:233-238.
Heitz, J. R., L. Lewis, J. Chambers, and J. D. Yarbrough. 1974.
The acute effects of empire mix crude oil on enzymes in
oysters, shrimp and mullet. p. 311-328. In F. J. Vernberg
and W. B. Vernberg (eds.), Pollution and physiology of
marine organisms. Academic Press.
Hochachka, P. W., and G. N. Somero. 1973. Strategies of bio-
chemical adaptation. Saunders.
Livingstone, D. R. 1976, Some kinetic and regulatory properties
of the cytoplasmic L-malate dehydrogenases from the
posterior adductor muscle and mantle tissues of the
common mussel, Mytilus edulis. Biochem. Soc. Trans.
4:447-451.
Moore, S. F., and R. L. Dwyer. 1974. Effects of oil on marine
organisms: a critical assessment of published data. Water
Res. 8:819-827.

108

Observations on Argo Merchant Oil in
Zooplankton of Nantucket Shoals

R. Polak and A. Filion
McGill University
Montreal, Canada

S. Fortier and J. Lanier
U.S. Coast Guard
Groton, Connecticut

K. Cooper
University of Rhode Island
Kingston, Rhode Island

Abstract Histopathological examinations of hematoxylin and
eosin stained organisms from February samples indi-
Zooplankton samples were taken in February, July cated an oil-like material present in the area of the gut
and December 1977, on Nantucket Shoals, following the of some crustacean zooplankton. Fluorescence micro-
Argo Merchant spill. scopy was used in an attempt to identify the oil-like
Visual observations of apparent oil in zooplankton globules.
were followed by Ultra Violet Spectrofluorometric analy-
sis to establish relative levels of contamination. Of 22
sites sampled in February, zooplankton samples from 14
showed significant concentrations of oil. Frozen samples Introduction
collected in February, July and December 1977 were also Following the wreck of the Argo Merchant tanker on
analyzed using the spectrofluorometric method. Samples Nantucket Shoals(Figure 1), onDecember 15,1976, and the
taken in February, and preserved in formalin, were also subsequent release of approximately 7.7 million gallons
examined for comparison with the frozen samples. of No. 6 fuel oil, a survey was carried out, February
21 to 25, 1977, to investigate the possibility of persis-
tent contamination of zooplankton in the area of the spill.
Observations on the marine communities imme-
diately after the spill (Dec. 22-24) were conducted by the
personnel of the Northeast Fisheries Center of the
National Marine Fisheries Service. At that time, the oil
slick was still in the area of the wreck, and at 6 of the 11
stations sampled, copepods fouled by an oil-like material
were observed. The contamination was classified by
NMFS as:
a) external; smudges on the exoskeleton.
b) mandibular; particles adhering to feeding appdn-
dages or tar stains on mandibles.
c) apparent oil particles which had been ingested
and were present in the gut.
During the Endeavor 005 cruise, February 21-25,
1977, we sampled the zooplankton at 22 stations around
the wreck. On July 22, 1977, 6 of the February stations
were repeated and a new station was added. On December
15, 1977, another 4 of the February stations were
Figure 1. Nantucket Shoals, site of the wreck. sampled by NMFS at our request.

109

69* 20'W 50

2
(L 40,
.64 IL
008 Z
65 0 30

ARGO STERN .7 12 25
.5 ARGO BOW
Z
41*N LU 20
(Y72 'a 0
Z
0
5 01 L)
148 49 0 /.50 9
,2 54
10
47 82

-45 50 56 61 70 59 71 73
STATION NO.
Figure 3. Comparison of oil concentrations in plankton vs. net,
.333 mm mesh.
.411
Figure 2. Chart of sites Sampled. so

Methods a- 40
During 21-25 February, oblique tows were taken first
with 0.5 meter nets of 0 and 6 mesh size, and later in the Z
.cruise with large 61 cm bongos of corresponding mesh 0
size. In July and December, double oblique tows were 30,
made with paired 61 cm bongos and the same mesh <
sizes. The zooplankton collected was examined super- Cr
ficially under a dissecting microscope. A number (-5-15 25
of organisms were removed for subsequent histological Z
W
examination. From the remainder of the net contents, 0 20
samples of approximately 1 cm3 were taken for oil Z
analysis. These samples were composed primarily of 0
zooplankton, but necessarily included particulate matter, 0
water and some phytoplankton. Oil present was ex- _j
tracted using 5 mls of spec-grade n-hexane. The hexane - 10
supernatant was analyzed on a Farrand Mark 1 Fluores- 0 82
cence Spectrophotometer, using syhchronous-scanning
techniques. The techniques, previously described by
Gordon and Keiser (1974, 1976), involve scanning both 3.5
excitation and emission monochromators, simultaneously 0 -0
with a constant offset ( AN ) between them, in this case
25 nm. Quantitative estimates of oil concentrations T,1 1 -1
present in the zooplankton samples were made by com-
@RN
ARGO e

paring their spectra with those of Argo Merchant stan- 50 56 61 70 59 71 73
dards. Because of its sensitivity and speed, fluorescence STATION NO.
spectroscopy is generally recognized as a reliable Figure 4. Comparison of oil concentrations in plankton vs. net,
method for monitoring oil in the marine environment. .505 mm mesh.

110

Table 1. Summary of Stations Sampled and Analyses Performed.
Oil concentrations in ppm.
*indicates presence of lighter oil
+ indicates samples which were examined histologically
- indicates no oil was found
blank indicates analysis was not performed

February 1977 July 1977 December 1977
Station Mesh No. Fresh Frozen Histo. Frozen Histo. Frozen Histo.

44 0 -
45 0 <.24
47 0 1.2-3.5 + +
48 0
49 0 +
50 6 3.5-8.2 + +
0 1.2
57 0 75-100 +
64 6
0
61 6 .4-1.2
0 <.24
62 6
0 +
54 6 5-7.5 16-25
0 3.5 +
56 6 25-50
0 .24-.72 +
53 6 >117
0 >117
52 0 .72-1.2
63 6 25-50 >1 17
0 50-75 2.7-8.3
58 6 8.2-25 7.2
0 8.2-25 +
56 333
505
61 333 8.2-25 + +
505 25-50
70 333 +
505 .72-1.2
59 333 25-50 +
71 333 3.5-8.2 +
505 25-50
73 333 8.2 +
505 8.2
El + +
Control 333 +
505
72 505 +

Duplicates of most samples were preserved for later organisms were examined visually, and fluorescent spec-
examination. These were stored in n-hexane-cleaned tra were analyzed using 366 nm as the excitation wave-
dark glass bottles and frozen immediately to avoid micro- length and scanning the emitted light from 400 to 750 nm.
bial degradation of any oil present. Samples of known Argo Merchant oil were examined as
Zooplankton prepared for light histology were fixed standards. This method had previously been used to
in 10% buffered formaldehyde, 3% gluteraldehyde or 1 % identify and quantitate crude oil taken up by protozoans
osmium tetroxide. Tissues were dehydrated in alcohol (Lanier, 1977).
and xylene, embedded in paraffin, sectioned at 6 micro-
meters and stained with hematoxylin and eosin. Results
Externally fouled organisms were also examined
with a fluorescence microscope to which a Farrand Figure 2 charts the locations of the stations
Microscope Spectrum Analyzer was attached. The sampled. The results of chemical (UV) and histological

Table 2. Comparison of Oil Concentrations in Plankton vs. Nets.
- indicates no oil was found

Station Date Plank ton Nets
Mesh No. Oil (ppm) Oil (ppm)

50 26/2/77 333 OIL STANDARD
505 .72-1.2 50 pp..
56 26/2/77 333
505 - ZOOPLANKTON
61 26/2/77 333 8.2-25 -
505 25-50 .24-32
70 26/2/77 333
505 .72-1.2
59 26/2/77 333 25-50 1.2-3.5
505 .72-1.2
71 27/2/77 333 .35-8.2 -
505 25-50 3.5 Figu re 5. Fluorescence spectra of hexane extracts of net me terial
73 27/2/77 333 8.2 - and zooplankton compared with Argo Merchant oil standard.
505 8.2

Table 3. % Contamination of Organisms Found in Formalin
Preserved Samples.

% Fouled Oil Level in Fresh
Station Organisms Samples (Ppm)

73 5 8.2
62 0 0
57 4 75-100
-50
A
25
71 25
70 2 .72-1.2
-25
61 36 8.2

examination of fresh and frozen samples taken in
February, July and December are presented in Table 1. It
should be noted that oil concentrations observed in fresh
February samples ranged from less than 0.24 ppm to Figure 6. Gammarid amphipods showing varying contamination,
greater than 117 ppm in 14 of the 22 stations sampled. external and internal, as well as clean amphipods.
The single July sample showed no trace of Argo oil, while
one of the four samples from December did show a low
concentration of Argo Merchant oil. The sites selected for the concentrations of oil associated with the net material
December, however, were the least contaminated of the were 8 to 100 fold less than that associated with the net
stations sampled in February. At some stations, indi- content. Figure 5 compares spectrofluorometric tracings
cated by an asterisk in Table 1, we detected the presence of the hexane extracts of net material and zooplankton,
of light oil. with a known standard of Argo Merchant oil.
The possibility that these data were the result of Histological examination of both fresh and fixed
cumulative contamination of the nets was investigated. samples revealed that some zooplankters were conta-
After several tows were taken with number 0 and 6 nets, minated with oil. The amphipods in Figure 6 illustrate the
we changed to clean bongo nets of comparable mesh types of fouling of the zooplankton. Amphipods A and C
size. Strips of netting of the same material and mesh size had external ly-fou led carapaces. Amphipod B appears to
were attached in the caudal end of the new nets. After have internal contamination, while amphipod D is
each tow, a swatch was taken from these strips and apparently clean. When formal in-preserved zooplankton
analyzed for presence of oil. The results of chemical samples were examined nine months later, oil could still
analysis of net versus zooplankton are summarized in be observed on and in the animals. At five stations with
Table 2 and Figures 3 and 4. The stations are listed in the samples showing contamination in February, the inci-
chronological order in which they were sampled. Cumula- dence of fouled individuals varied from 1 to 25% of the
tive contamination would be seen as an increasing con- total sample (Table 3). To establish the nature of the
centration of oil on the nets from station to station. Ten fouling material, a sample of several contaminated
of the fourteen net swatches showed noArgo oil present. organisms from the 5 stations were mascerated and
At those stations where net contamination was apparent, extracted in hexane. A synchronous-scanning spectro-

112

fluorometric analysis of the hexane extract is shown in
Figure 7, and compared with an Argo Merchant oil
standard.
The organisms were also examined using fluores
cence microscopy and a microscope spectrum analyzer.
Argo Merchant oil standards, however, emitted no
OIL STANDAR@@,, ORGANISM measurable fluorescence spectrum when excited at 366
5OPPM nm and scanned from 400 to 750 rim. Visual examination
of the fouled organisms using fluorescence microscopy
revealed the entire organism fluorescing with the excep-
tion of the black globules suspected of being Argo
Merchant oil.
Histopathological examination of the osmium fixed
specimens collected in February showed that an oil-
.like material was present not only on the external sur-
faces of the animals but also in the area of the gut
(Figure 8). Neither the zooplankton specimens collected
from the February control stations northe July specimens
Figure 7. Fluorescence spectra of hexane extracts of 6 organisms showed a comparable oil-like material in the gut or on
preserved in formalin compared to the fluorescence spectra of the carapace (Figure 9).
an Argo Merchant oil standard.

@r :Vq FHM e,
gl F r @n w *i
Kqr@_ff RANY-RAITIF Discussion
-5
Estimates of oil concentrations in the water column
H
vary with station, time and the mesh-size of the net. The
variation from station to station in February can be seen
in Table 2. The same station, when sampled on 2 days,
using the same mesh-size, yielded different results. The
station 61 sample, on February 23, had an oil concentra.
tion of less than 0.24 ppm, while on February 26 the same
sample from the same mesh-size net had an oil concen-
tration of 25-50 ppm. Different concentrations were
U obtained from the 2 mesh sizes used at the same station
on the same day. At station 56 on February 24, the *0 net
yielded a concentration of 0.24-0.72 ppm while the #6 net
-50 ppm. This variability in the
had a concentration of 25
observed concentrations appears typical of the uneven
distribution of oil in the surface water, the water column,

V-
sediments, beaches, organisms and tissues. Because of
this uneven distribution, large numbers of samples are
essential to assess the degree of contamination of an
Figure 8. Longitudinal section of a fouled copepod from station area or of organisms.
56, showing globules of an oil-like material. (Mag. 10x) Where differences in concentration were found in
samples from different mesh sizes, the smaller mesh
(333 mm, no. 6) generally yielded higher concentrations.
nM
This is consistent with the idea that a smaller mesh would
retain more of the particulate matter found in the water
Z, i
column.
Argo Merchant oil concentrations as high as 117
ppm were found in the water column of the area two
& months after the spill. As we are dealing with a shallow
and turbulent region, this high concentration may be due
im to the resuspension of sediments and oil into the water
column, thus maintaining its availability to the zooplank-
ton community. The dynamic nature of the shoals (high
turbulence, shifting sediments and changing currents)
contributes to the dispersion of oil.
Al
A-0
Although the bulk of the oil apparently shifted from
the site of the wreck, there appears to have been a
V gradient of decreasing oil concentration with distance
from the wreck beyond this shift. The highest concentra-
0 @11 @[email protected]
SO

tions of oil observed were not found immediately adjacent
Al"
'J`77NIMPPI,
to the wreck. For example, station 70 had an oil concen-
Figure 9. Longitudinal section of a "clean"copepod, from station tration of 0.72-1.2 ppm while station 57 had a concentra-
73, without globules of oil-like material. (Mag. 10x.) tion of oil of 75-100 ppm. Examination of the results for

113

the south-west transect, shown in Figure 2, indicates the tography showed a complex mixture of hydrocarbons
presence of a gradient. Station 73 had a concentration of similar to that present in petroleum and found in the
8.2 ppm; station 50:3.5-8.2 ppm; station 47:1.2-3.5 ppm; emulsion and surface film samples of the area. The field
station 45: 0; and station 44: 0 (Table 2). study suggested that the species of zooplankton investi-
When the frozen samples were analyzed chemically gated can store and concentrate pollutant hydrocarbons
nine months after they were taken, no correlation was which were present in the surface waters. I
found between concentrations in fresh and frozen Stegeman (1977) points out that the metabolism of
samples. This may be attributed to several factors: certain petroleum hydrocarbons, which may be relatively
a) the storage temperature may not have been low inert, can result in the production of highly toxic and
enough to inhibit microbial activity, carcinogenic derivatives.
b) repeated thawing and freezing of material may A component of the observed concentrations may
have accelerated degradation of the oil, be due to the fractions of oil dissolved in sea water. The
c) dewaxing of oil present, during long storage. amount of water soluble fractions seems to vary de-
Optimum storage temperature should be -30*C, and the pending on the oil involved. Anderson et al. (11974) exam-
maximum storage period should be less than 2 months ined water soluble fractions of No. 2 fuel oil, Bunker C
(Engelhart, 1977). The weathering of oil standards stored residual oil, Kuwait crude and South Louisiana crude. The
along with the samples should be investigated for two crude oils gave more concentrated water soluble frac-
comparison. tions than did the two refined oils. Frankenfeld (1973)
Formaldehyde preservation apparently does not investigated the weathering of No. 2 fuel, Bunker C and
greatly affect the fluorescence spectrum of oil. Fouled Venezuelan crude under simulated natural conditions.
organisms preserved in formaldehyde yielded spectra After a week of weathering in laboratory simulators, the
similar to that of Argo Merchant oil, when extracted with dissolved fractions of No. 2 were approximately 3.5 times
n-hexane and observed with the synchronous-scanning those found for the heavier oil. After one month, 1.4 ppm
fluorescence spectrophotometer. Whether the concentra- of Louisiana crude and 3 ppm of No. 2 fuel remained in
tion of oil remains unchanged through time remains to be solution. This finding disputes reports that light oils
explored. disappear almost entirely due to evaporation shortly after
The total oil present in our samples is most likely due being introduced in the marine environment.
to combinations of the following factors: Because of the turbulent conditions in the area of
1) oil in the gut the Argo' Merchant wreck, some component of the oil
2) oil in body tissues concentrations observed may be due to adhering emul-
3) water soluble fractions sions of oil and tar. If a relatively thick oil film undergoes
4) adhering emulsions of oil and tar. violent agitation at the air-sea interface, even gasoline
Oil in the gut may be due to oil coated material or oil may form a temporary emulsion (Blackman et al., 1973).
particles that are directly ingested by zooplankton. Rough seas tend to create emulsions of the oil in water
Parker et al. (1971) demonstrated the presence of "con- phase. Berridge et al. (1968) studied the formation and
siderable quantities" of oil in the guts of copepods and stability of water in oil emulsions and found that they were
barnacle larvae after exposing them to a fine suspension stable for periods exceeding 100 days. According to
of oil of 2-10 ppm for 18 hours at 10*C. Conover (1971) Berridge, water in oil emulsions do not require the
reported zooplankton ingesting oil droplets of Bunker C addition of external dispersing agents but form naturally
after the Arrow spill. The oil had no observable effect on on a dynamic sea surface.
the organisms and apparently passed through the intes- Small tar balls were noted during the examination of
tines essentially unchanged. Histological examination of the material. These could be the cause of the external
our February samples indicated oil-like droplets in the fouling of the carapaces. It is conceivable that a residual
guts of the copepods and amphipods collected from tar ball is responsible for the presence of oil at station 45
heavily contaminated areas of station 56. in December 1977 (less than 0.24 ppm). However, the
Oil may be transferred from the gut to the body fraction indicated by the spectrum is lighter than that
tissues. Oil may also reach the body tissues by uptake normally found in tar balls.
through other body surfaces. Lee (1975) described the
uptake, metabolism, storage and discharge of petroleum Conclusions
hydrocarbons by selected marine zooplankton and
benthos. Both paraffinic and polycyclic aromatic hydro- The study indicated that spectrofluorometric
carbons were added to seawater containing various analysis of zooplankton samples can be used to deter-
species of zooplankton collected off California, British mine the geograhic extent of oil in the water column and
Columbia and in the Arctic, including the copepods the persistence of oil residues following an oil spill.
Calanus plumchrus, C. hyperboreus, C. helgolandicus and Fluorescence spectroscopy allows for an almost positive
the amphipod Parathernisto pacifica. The uptake of dis- identification of the oil involved, as well as preliminary
solved hydrocarbons was linear for 24 hours, with no quantitative assessment of the oil levels, provided that
further increase in stored hydrocarbons after that time. spectra of the oil standards are available.
Most ingested hydrocarbons were metabolized and dis- This method was used to detect Argo Merchant oil
charged by the various crustacean species, although a in the water column in the area of the wreck two months
small percentage (lessthan 1 % of ingested hydrocarbons) after the spill and, in one case, twelve months after the
was retained by all species even after a "long depuration spill. Some of this oil was associated with zooplankton.
time". The above phenomenon was also observed by The oil was observed adhering to the carapaces of zoo-
Morris (1974) in the field. He found high levels of hydro- plankton and apparently ingested by the organisms.
carbons in the lipids of near surface samples of zooplank- Samples preserved for several months in formalde-
ton of the eastern Mediterranean. Thin layer chroma- hyde contained identifiable Argo Merchant oil, indicating

114

that formaldehyde may be an acceptable oil preservative
for spectrofluorometric analysis through further investiga-
tion is necessary.
It was not,possible to identify the oil-like globules
in the guts of crustacean zooplankton using fluorescence
microscopy, although this technique may be useful for
oils with higher concentrations of aromatic fractions.
Although these techniques demonstrate the
presence of oil, in order to detect the effects of persistent
low level petroleum concentrations in the marine environ-
ment, studies of the effects on zooplankton of water
soluble fractions, oil contamination of food, plants and
lower trophic levels should accompany field observa-
tions.

References

Anderson, J. W., et al. 1974. Characteristics of dispersions and
water-soluble extracts of crude and refined oils and their
toxicity to estuarine crustaceans and fish. Mar. Biol. 27:
75-88.
Berridge, S. A., M. T. Thew and A. G. Loriston-Clarke. 1968.
Formation and stability'of emulsions of water in crude
petroleum and similar stocks. J. Inst. Petrol. 54.
Blackman, R. A. A., et al. 1973. TheDona Marika oil spill. Mar. Poll.
Bull. 4(12):181-2.
Conover, R. J. 1971. Some relations between zooplankton and
BunkerCoil in Chedabucto Bay, following the wreck of the
tanker Arrow. J. Fish. Res. Bd. Canada 29:1327-30.
Englehart, R. 1977. Personal communication.
Frankenfeld, J. W. 1973. Factors governing the fate of oil at sea;
variations in the amount and type of dissolved or dis-
persed material during the weathering process. In
American Petroleum Institute, Proceedings of Joint Con-
ference on Prevention and Control of Oil Spills, Washing-
ton, D.C., March 13-15,1973. Pp. 485-95.
Gordon, D. C. Jr., and P. D. Keiser. 1974. Estimation of petro-
leum hydrocarbons in seawater by fluorescence spec-
troscopy: improved sampling and analytical sampling and
analytical methods. Tech. Rep. Fish. Res. Bd. Canada 481.
Gordon, D. C. Jr., P. D. Keiser and J. Dale. 1974. Estimates, using
fluorescence spectroscopy of the present state of petro-
leum hydrocarbon contamination in the water column of
the Northwest Atlantic Ocean. Mar. Chem. 2:251-61.
Gordon, D. C. Jr., P. D. Keiser, W. R. Hardstaff and D. G. Aldous.
1976. Fate of crude oil spilled on seawater contained in
outdoor tanks. Environ. Sci. Tech. 10:580.
Jadamec, J. R. 1978. Water soluble fractions of Argo Merchant
oil. In Proceedings of the Symposium "in the Wake of the
Argo Merchant."
Lanier, J. 1977. Unpublished manuscript.
Lee, R. F. 1975. Fate of petroleum hydrocarbons in marine zoo-
plankton. Proceedings of the 1975 Conference on Preven-
tion and Control of Oil Pollution.
Morris, R. J. 1974. Lipid composition of surface films and zoo-
plankton from the eastern Mediterranean. Mar. Poll. Bull.
5:105-9.
Parker, C. A., M. Freegarde and C. G. Hatchard. 1971. The effects
of some chemical and biological factors on the degradation
of crude oil at sea. In Water Pollution by Oil, edited by
P. Hepple, London, Institute of Petroleum, pp. 237-44.
Stegeman, J.J. 1977. Fate and effects of oil in marine animals.
Oceanus 20(4).

115

Field and Laboratory Measurements of
Stress Responses at the Chromosome
and Cell Levels in Planktonic Fish Eggs
and the Oil Problem

A. Crosby Longwell

National Marine Fisheries Service
Northeast Fisheries Center
Milford Laboratory
Milford, Connecticut

Abstract pollock eggs. However, only 4% of the laboratory-
Abnormalities in chromosome makeup at the early spawned cod eggs were dead or moribund.
At a station within the oil slick, pollock embryos were
embryo stage may be at least one of the most sensitive grossly malformed in 18% of the eggs, and at the periphery
practical indicators of the sublethal effects of marine of the slick, 9%were grossly abnormal. No embryosof cod
pollutants on reproduction in fish. Cytogenetics - the showed abnormalities, nor were there any abnormal
study of chromosomes and their divisions as they affect pollock embryos at any stations more distant from
heredity - has been applied to fish rarely in the past. It is the slick.
only recently that reliable methodology has been devel-
oped for conducting cytological and cytogenetic studies
on fish eggs from plankton samples taken at sea. These Ifitroduction
methods are even more easily applied to cytological-
cytogenetic analyses of fish eggs used in laboratory When fish eggs are spawned at sea they are
experimentation on oil and other contaminants. This arrested at the metaphase or telophase stage of the second
methodology, first developed and used in conjunction meiotic division of their chromosomes. On fertilization they
with a study of the eggs of Atlantic mackerel, Scomber
scombrus, in the New York Bight, is described.
Using the same procedures as for mackerel
eggs, microscopic examinations were made of the dis-
sected embryos of 79 cod and 162 pollock eggs taken
from surface waters in the vicinity of the Argo Merchant
spill. Seventy-five cod embryos from eggs of a laboratory
spawning of aquarium-held fish were also examined. This
study was greatly limited by the small number of eggs
available. Cod and pollock eggs were scarce at the
cleaner stations. This made precise station and
species comparisons impossible, though some could
be made, using the combined estimates of cyto-
logical mortality and cytogenetic moribundity.
Eggs at all stations showed some oil contamination
of the chorion, although fewer cod than pollock eggs were
fouled. A higher mortality of pollock eggs was Figure 1. Just before the 2-cell A tiantic mackerel embryo of Fig. 2
also apparent. Overall, about 20% of the cod was formed, it underwent its first mitotic division, as seen in this
eggs were dead or moribund, compared with 46% of the photomicrograph of mackerel cleavage.

116

------- duced by this process can form free radicals which react
with the hereditary DNA to produce mutations. On photo-
oxidation, other toxic components already present, such
as phenols, increase over a period of days. Phenols can
be mild chromosome-breaking agents (Hollaender,
1971 a and b).
Certainly, the membranes and jelly coat of fish eggs
afford some considerable measure of protection against
pollutants and, once hatched, larvae may exhibit in-
ased physiological sensitivity to some pollutants.
cre
@-A
Such protection though is most likely temporarily
diminished when perivitelline fluid formation takes place
after fertilization and the pollutant enters the egg with
imbibed water. A pollutant which damages or disor-
M
ganizes the egg membranes will modify their perme-
Figure 2. Pelagic Atlantic mackerel eggs as they occur in surface ability. Aromatic compounds of oil appear to alter the
waters. Egg to the lefthas completed its first cleavage division and surface properties of cell membranes, as studied by the
has a 2-cell embryo (see arrow). Body to bottom of the egg is the spin-labeling technique, while paraff in hydrocarbons do
oil globule characteristic of some species eggs. Egg to the tight not (Roubal and Collier, 1975). However, studies specifi-
has a 4-cell embryo at arrow. Actual egg size about 1 mm. As cally on the membranes of fish eggs have not been
viewed under low magnification with dissecting microscope. conducted.
The sensitivity of the early embryo stages of any
species to chromosome damaging mutagens, and to
complete this genetic-sensitive division. The well- cytotoxins which can disrupt the orderly division of the
established genetic vulnerability of gametogenesis is fur- chromosomes, lies, in large measure, in the necessity of
ther increased as the fertilized egg enters early cleavage the embryo for a normal complement of unaltered
(Figure 1) (Solberg, 1938; Muller, 1959; Murakami, 1971). chromosomes for normal development to proceed
The genetic vulnerability of fish eggs to induced mutation (Sonnenblick, 1940; Whiting, 1945; Muller, 1954a and b;
appears second only to the most sensitive mammalian Cave and Brown, 1957; von Borstel and Rekemeyer,
egg (Donaldson and Foster, 1957; Polikarpov, 1966; 1959; von Borstel, 1960; Epstein et al., 1970; Bateman
Purdom and Woodhead, 1973). Aquatic environmentalists and Epstein, 1971). Some measure of natural, back-
who investigate contaminant effects on reproduction via ground gametic wastage in fish could be the resultof such
non-genetic approaches conclude that reproduction natural environmental variables as light, temperature,
seems to be one of the most sensitive measures of salinity and oxygen acting on the highly sensitive meiotic
chronic sublethal effects clearly meaningful in nature and early egg stages. Chromosome errors are almost
(Sprague, 1971; Rosenthal and Alderdice, 1976). invariably lethal if they occur before the gastrula stage.
Many of the important commercial species of fish Unlike physiological effects, there can be no recovery.
have planktonic eggs that develop in the upper portion of After gastrulation of the embryo, lethality will depend on
the water column, often at the air-water interface (Figure 2). the number of affected cells and the destiny of their cell
Not only is the ocean surface microlayer directly exposed lineages in development. The correlation with lethality,
to atmospheric pollutants, it also appears to concentrate however, remains strong. Mutations occurring during the
heavy metals and chlorinated and petroleum hydrocarbons period of major organogenesis lead to developmental
(Duce et al., 1972; Macintyre, 1974). It bears the brunt of abnormalities of systems not fully developed in fish until
oil spills. As immobile occupants of surface waters, the post-hatching period.
buo.yant fish eggs will inevitably be exposed to whatever Generally, genetics has thus far had little impact on
noxious components escape from oil slicks and oil drop- fishery biology. Similarly, fish have played only a minor
lets in this vicinity. There is some evidence that fish may role in the development of genetics. The value of any kind
actively avoid oil spills. However, eggs can drift into spills, of genetic monitoring of fish has not hitherto been con-
and choice of alternate spawning beds could result in poor sidered by groups addressing marine baseline and
development because of unfavorable or less than optimal monitoring approaches (Longwell, 1975; McIntyre,
natural environmental conditions of these other grounds. 1976). Still, in a review of measurement of pollutant
Further, spills may occur after spawning, while embryo toxicity to fish, Sprague (1971) comments that tests for
development is already under way, as in the case of the reproductive damage are rightfully regarded as the most
Argo Merchant. important in determining "safe" concentrations.
The low-boiling-point aromatic hydrocarbons of oil The explosion of interest in vertebrate cytogenetics,
are soluble in water. They also are highly soluble in lipid that is, the study of chromosomes as they affect heredity
material, as present in fish eggs. Benzene, the most of the cell, the individual and population, was confined to
abundant such chemical compound in crude oil, has mammals. Fish were largely by-passed even though they
proved mutagenic in a number of tests on different comprise the largest of all vertebrate groups. Roberts
organisms. Polynuclear aromatic hydrocarbons can act (1967) reviewed the status of chromosome cytology of
as both carcinogens and mutagens (Hollaender, 1971 a the Osteichthyes. Investigating the effects of incor-
and b). At least when No. 2 oil is exposed to irradiance as porated radionuclides, the Russians more than any other
it would receive from sunlight, photooxidation produces group have conducted experimental cytogenetic studies
compounds more toxic and persistent, and more water on fish eggs (AEC-TR-6940, 1968; AEC-TR-7418,1971;
soluble than those present before irradiation (Scheier and AEC-TR-7299, 1972). Although Polikarpov (1966) notes
Gominger, 1976; Larson et al., 1977). Peroxides pro- the importance of direct studies on the neuston, the

117

richest biosphere in the world, these Russian studies
were still limited to the spawn of experimental, laboratory-
treated fish.
In 1974 a cruise of the sailing vessel Westward of
SEA (Sailing Education Association) provided a large col-
lection of Atlantic mackerel (Scomber scombrus) eggs
from 40 stations over the variously polluted New York
Bight. Using this egg collection, appropriate methodology
was developed for cytogenetic study of fish eggs col-
lected at sea in plankton samples. This methodology
was then applied to a study of these eggs. First efforts in
this regard and methods were described in a NOAA
Technical Memorandum (Longwell, 1976). Since
then, more than 15,000 eggs, mostly from plankton
samples, have been processed and examined cyto-
genetically. Most of these have been mackerel.* How-
ever, about a dozen other species, demersal and plank-
tonic, have also been processed, enough to establish that
just about any species of egg might be so studied.**
Results of mackerel work and improved methodology are
now being prepared for publication in a NOAA Technical
Memorandum and elsewhere.
Procedures for Cytogenetic and Cytological Study of Fish
Embryos. Plankton samples are collected at sea with
either neuston or bongo nets towed at appropriate levels
of the water column similarto standard tows forassessing
zooplankton. Tow time and speed are adjusted to mini- Figure 3. Planktonic Atlantic mackerel egg at very early morula
mize any damage to fish eggs. Efforts are made to stage of development, and embryo dissected off its egg. Arrow
estimate water flow over the eggs in the tow process, points to embryo still in the egg. Characteristic oil droplet is to
Immediately on being brought aboard the vessel, plank- bottom of the egg. Body to upper right is morula-stage embryo
ton is fixed in a 1:10 dilution of buffered histological grade dissected off its egg in preparation for examination of its cells and
formalin. At-sea physical factors that could affect egg chromosome divisions. Actual egg size about 1 mm. As viewed
viability, such as temperature and salinity, are measured under low magnification with a. dissecting microscope.
at each sample station.
Once fish eggs are picked out of the fixed plankton
in the laboratory, they are identified to genus and species.
They are then sorted by developmental stage. Eggs at
each developmental stage are carefully scrutinized under
low magnification to appraise overall egg and embryo
condition. At this time abnormalities of gross develop-
ment are noted.
Under a low-power microscope, the embryo is dis-
sected from the egg with a needle (Figures 3 and 4). After
post-fixation in acetic acid, the fragile embryo is stained
and squashed into a monolayer of cells on a microscope
slide (Figures 5 and 6). For cytogenetic work, the staining
medium is the standard aceto-orcein to which proprionic
acid is added. Cells and their dividing chromosomes are
viewed under high-resolution, high-power, light micro-
scope optics.
Depending on developmental stage, the fish
embryos lend themselves more or less well to the collec-
tion of different kinds of data on their chromosomes and
mitoses (to be described fully in a forthcoming NOAA
Technical Memorandum).
In mackerel from the New York Bight a large per-
centage of observed early cleavage embryos was
dividing irregularly at the chromosome level or showed

*Mackerel studies and methods development have been Figure 4. Plankt onic Atlantic mackerel eggs at the later tail-free
supported by the Marine Ecosystems Analysis Program stage of embryo development - the stage of a large portion of the
of NOAA. pollock eggs sampled at the time of the Argo Merchant spill. Arrow
points to a tail-free embryo partly encircling its eggs. At top and at
"Supported by the NEFC and by contracts with Ocean bottom of the photomicrograph are two such embryos dissected
Survey of NOAA for study of the Toxic Chemical off their eggs. Actual egg size about 1 mm. As viewed under low
Disposal Site 106. magnification with dissecting microscope.

118

T_
Stat;on
-30
*11,@,,'@,-,@@ll,@',@,@@,",,"'-@ @7
ITaIN-Free
,.,$tage
Mv
, Bl-_ih"_
Figure 5. Four tail-free embryos dissected off planktonic fish eggs
squashed onto microscope slide, stained, and ready for study of
their cells and mitotic divisions. Near actual size.

Figure 8. Normal mitotic telophase at gastrula stage in mackerel
embryo from a planktonic egg. Arrows point to one of the two
daughter groups of just divided chromosomes. Light microscope,
100 X objective.

-zr

Figure 6. Monolayer of gastrular cells of whiting (Merluccius
bilinearis) embryo prepared for cytological-cvtogenetic study.
Several normal mitotic configurations are obvious. Light micro-
scope, 63 X objective.

Figures 9 and 10. Two abnormal mitotic telophases at gastrula
stage from planktonic mackerel eggs. Note c romosomes
lagging between, and bridging the two groups of daughter
Figure 7. Monolayer of gastrular cells of whiting (Merluccius chromosomes and to the back of one group. Such telophase
bilinearis) embryo with abnormal chromosome configurations irregularities are indicative of chromosome damage, and result in
unable to complete their mitoses. Light microscope, 63 X irregular distribution of chromosome material to daughter cells.
objective. Light microscope, 100 X objective.

119

7711191 69-30, 6roo,
70-30, 7'G.M.
I I
41'30@ - 41-30'
'J V, OELAWARE 11 - CRUISE DE 76-13
m
ARGO MERCHANT OIL SPILL
No
OEC. 22-24, 1976
let.

41-00- 41-00,
9
2 10 !1

J, 70-30' 70-W
Figure 13. Station locations for Delaware I/ cruise. The solid line
Figure 11. Grossly abnormal mitotic telophase in mackerel indicates oiled area.
embryo from a planktonic egg. Extremely abnormal stickiness of
the chromosomes has caused them to lose their distinctness and
prevents their completing this mitotic division. Light microscope, differ widely in number of mitoses partly because of Wide
100 X objective. variation in cell number, the number of mitoses in gastru-
lating embryos is readily recognizable as an indicator of
embryo well-being. The chromosome configurations in
the tail-bud and tail7free stages become smaller, but telo-
phase bridging is still readily scorable. Mitotic index
remains a good indicator of embryo condition throughout
embryo development once influencing factors like tem-
perature are taken into consideration. See Figures 8-11.
As development proceeds, characteristic patterns
of cell differentiation appear in the various stage
embryos. Aberrations of some of these patterns, a portion
of which is no doubt related to abnormal chromosome
events, are readily discernible even in monolayered
preparations. Most common among these are: abnor-
mally, prematurely differentiated nuclei in blastular and
with too many cells of too small
size; the retention of one or more large, undivided early-
gastrular embryos, often

cleavage cells in otherwise later-stage embryos; and the
dramatic de-differentiation of the cells of the early or
later-stage embryos. All of these phenomena are readily
observable in the course of examining embryo mitoses,
and records are kept on such observations.
A recent methods development at the Milford
laboratory has made it possible to characterize individual
Figure 12. Chromosomes from a tail-bud Atlantic mackerel egg chromosomes of the planktonic fish eggs using theiryolk-
picked out of a routine plankton sample. For karyotyping chromo- sac membranes and also the meiotic configurations of
somes, animal and plant cells are usually pre-treated with prespawned eggs stripped from fish in the field (in pre-
colchicine before fixation, but this may not be necessary for most paration for publication). It is now possible to'detect in-
work on these fish eggs. The egg involved here was not treated duced abnormalities of chromosome numbers and form,
with colchicine. Light microscope, 100 X objective. as was done on mammalian cells in laboratory tissue cul-
ture and to do this on field collections of fish eggs. See
Figure 12.
early signs of cell death or nuclear lysis. Often the mitoses All of this methodology is, of course, applicable to
essential to continued development had ceased altogether. fish eggs utilized in laboratory experiments.
Any of the ensuing developmental stages might also show The Argo Merchant Study
division arrest. At the subsequent blastula and gastrula
stages mitotic telophase configurations are good indica- Field Sampling of Fish Eggs in the Argo Merchant Spill. A
tors of the regularity of chromosome division, and of cyto- portion of the eggs from the ichthyoplankton samples
toxic effects on the chromosomes and translocation of taken during the Northeast Fisheries Center's Delaware //
broken chromosomes that lead to chromosome bridging. December 22-24 cruise to the site of the Argo Merchant
Blastular and less often gastrular embryos with totally dis- spill (Figure 13) was sent to the Milford laboratory for
organized mitoses in nearly all their cells are not un- microscopic study of their embryos. Plankton was col-
common (Figure 7). At the gastrular stage abnormalities of lected at cruise Stations 4-9 with standard oblique tows of
the mitoses drop dramatically, most abnormal embryos the water column with bongo nets, and of the water sur-
,W,6

being unable to gastrulate. Unlike earlier stages, which face with neuston nets. No tar-like surface oil was

120

Figure 14. Pollock eggs sampled a t edge of the Argo Merchant oil
slick Tail-bud and tail-free embryo stages. Egg at upper left and
egg atlowerrighthave their outermembrane contaminated with a
tar-like oil. Arrows point to some of the oil masses. The uncon-
taminated egg at upper right has a malformed embryo. The un-
-contaminated egg at lower left is collapsed and also has an Figure 15. Surface of an oil-contaminatedpollock egg from edge
abnormal embryo. Actual size of pollock eggs around 1 mm; of of the Argo Merchantsfick. Arrow at Jeff points to one clump of oil.
cod eggs around 1.5 mm. A clean portion of the outer egg membrane, its pores barely
visible at this magnification, shows to the left, above and below
this arrow. Arrow on rightpoints to an antenna of a copepodstuck
observed at cruise Stations 4, 5, or 6, and the bongo net in a mass of the oil adhering to egg membrane. Copepods are
samples of the water column were clean. Samples taken other components of surface waters observed to be fouled with
at the surface with the neuston nets, however, contained Argo Merchant oil. Scanning electron micrograph. About 500 X.
specks of tar. Stations 7 and 8 were thick "pancake"-like
oil slicks that resembled wrinkled black cloth floating on
the water surface, and the neuston nets became however, did not appear any larger, generally, than at the
saturated with oil. No fish eggs were collected at Station other stations, and fewer of the cod eggs were fouled
7. Station 9 was at the periphery of the thick slick. This than were the pollock (60% of 60 in one estimate). There
station had extremely high numbers of zooplankton and might also have been less contamination of the cod at the
total biomass. Eggs and "pancakes" alike must have - other stations. Scanning electron microscopy of the cod
changed locations prior to the sampling but, presumably, and pollock eggs sampled from the spill vicinity reveals a
not to such a degree as to nullify completely the sampling different pore structure in the two species. See Figures
strategy. 17 and 18.
Fish eggs were examined and identified at the Cytological-Cytogenetic Study of Cod and Pollock Eggs
Narragansett laboratory and Sandy Hook laboratory of from the Argo Merchant Spill Area. Microscopic examina-
the National Marine Fisheries Service. Only cod (Gadus
morhua) and pollock (Pollachius virens) eggs were tions were made of the dissected embryos of 79 cod and
present in the samples. Pollock eggs were most numer- 162 pollock eggs taken from surface waters in the spill
ous within and adjacent to the thick floating slicks. Cod vicinity of the Argo Merchant Also examined were 75 cod
eggs occurred largely about the periphery of the spill embryos from eggs of a laboratory spawning of aquarium-
area. held fish at the NMFS Narragansett laboratory.
Total numbers of eggs available for study were
Oil Contamination of the Outer Membrane of the Fish Eggs drastically limited. Cod and pollock eggs were both
Sampled from Surface Water. Eggs at all sample stations infrequent at the cleaner stations (Table 1). Of the cod
showed some oil contamination of their outer membrane, eggs 63% were earlier than the tail-bud stage with the very
the chorion. Oil droplets and tar adhered to roughly half of earliest stages well-represented. Pollock eggs were
all fish eggs examined (both species, all stations). Almost divided equally between the later tail-bud embryo and tail-
all the pollock eggs at Station 9, just outside the free embryo stages. All this made precise station and
pancake"-like slicks, had their outer membrane fouled species comparisons impossible. Even so, using com-
with a tar-like oil (Figures 14-16). In one estimate this was bined estimates of cytological mortality and cytogenetic
94% of 49 pollock eggs. Pollock eggs at Station 9 were all moribundity, some comparisons could be made.
quantitatively more fouled than at other stations. The Embryos which had ceased to undergo mitosis were
particles of oil adhering to cod eggs at this station, categorized as moribund. Cod and pollock eggs in the

121

U@k

Figure 18. Portion of surface of seemingly clean cod egg from
27
Argo Merchant spill vicinity, showing pore pattern of outer egg
membrane ,the chorion. Scanning electron micrograph. About
Figure 16. Portion of surface of oil-contaminated pollock egg of 10,000 X
Fig. 15 at greater magnification. Upper arrow points to one of
many oil droplets. Lower arrow points to one of the membrane
pores. Scanning electron micrograph. About 5000 X dead and moribund categories showed a combination of
cytological abnormality of the embryo's cells or of the
nuclear configurations coupled with division arrest. For
both the cod and pollock eggs, there was an extremely
wide within-station variation in embryo mitotic index for
the same developmental stage (Table 2). Many embryos
still exhibiting a low number of mitoses may have been on
their way to complete division arrest.
A higher mortality of pollock over cod was apparent
even though the cod eggs were at earlier, more sensitive
development stages at which higher natural mortality
might be expected. Totaled over all the stations, about
20% of the collected cod eggs were dead or moribund,
compared with 46% of the pollock eggs. As noted above,
fewer cod eggs were fouled and less so than the pollock.
Only 4% of the samples of cod eggs spawned in the
laboratory were dead or moribund at about the same
assortment of developmental stages as the samples from
the vicinity of the oil spill. If this cod control is represen-
w
tative, then the cod' ere experiencing higher than usual
mortality for cod in the spill vicinity but still less mortality
than the pollock.
At Station 8 in the "pancake"-like slick, pollock
embryos were also grossly malformed in 18% of the eggs
(Figure 14). At Station 9, the periphery of the slick, gross
malformations occurred in 9% of the eggs. None were
observed in any of the cod embryos or in any of the pol-
lock embryos at any of the stations more distant from the
heavy slicks.
Pollock mortality was lower at Station 8 in the
Figure 17. Still greater magnification of a portion of surface of pancake"-like slick than in' the more membrane-
seemingly clean pollock egg from Argo Merchant spill vicinity, contaminated eggs of Station 9 at the slick periphery.
showing pore pattern of outer egg membrane, the chorion. Those Station 9 embryos (15%) which had not entirely
Scanning electron micrograph. About 10, 000 X ceased to undergo cell division had very few mitoses

122

Table 1. Cytofogical-cytogenetic assays of mortality and moribundity of cod and pollock eggs from the vicinity of the oil spill

Delaware
Cruise DE 76-13 Total no. No, eggs No. eggs dead No. eggs with
Station numbers eggs viable or moribund malformed
embryos

Station 4
Cod 14 13 1 0
Pollock - - - -

Station 5
Cod 6 3 3 0
Pollock 11 0 11 0

Station 6
Cod 3 3 0 0
Pollock 3 0 3 0

Station 8
Cod 1 1 0 0
Pollock 105 86 19 19

Station 9
Cod 55 43 12 0
Pollock 43 1 42 4
Laboratory
spawning cod 75 72 3 0

(Table 2). Station 9 was probably in the "sheen" fed by the Table 2. Total numbers of mitotic telophases in pollock embryos
"pancake", which could have been more toxic than the at Stations 8 and 9
waters just beneath the thick slick from where Station 8
eggs must have come.
The non-dividing pollock eggs of Station 9 were Delaware Telophases (actual number or estimate)
characterized by rather pycnotic nuclei, which had the Cruise DE 76-13
appearance of chromosomes arrested at prometaphase Station numbers 03-14 15-25 �50 �75 -100 - +200
of mitosis. The cells of these embryos had a clearly non- Station 8 6 7 8 27 8 28
differentiated appearance resembling the cells of an
earfier developmental stage more than those they were Station 9 356 0 0 0 0
expected to contain on the basis of their more advanced
developmental stage. This suggests cellular de-differen-
tiation in response to stress.
The chromosomes of the yolk-sac membranes of
these cod and pollock eggs were not examined for direct at the NMFS Narragansett laboratory, however, failed to
evidence of breakage as the methodology was not yet result in any membrane contamination (W. KUhnhold, per-
developed. sonal communication). Oil has never been observed
adhering to any of thousands of other planktonic fish eggs
Discussion observed at the Milford laboratory over a period of a few
years, with the recent exception of Atlantic mackerel
External Fouling of the Pollock Eggs from the Argo eggs at one station in slope waters off New Jersey.
Merchant Spill. This appears to be the first time oil drop- External fouling of the outer pelagic egg membrane,
lets have been reported adhering to fish eggs, either if not proved an artifact, is important. Not only could it be
taken from the vicinity of an oil spill or in laboratory experi- obvious evidence of oil exposure, but it would also in-
ments on oil toxicity. Crude oil has been reported, crease the risk of egg mortality. Although it does not seem
as adhering to most of 22 species of coral studied, very probable, the likelihood that the oil was adf ixed to the
causing tissue death at the sites of adherence fish eggs (and copepod surfaces) only during the fixation-
(FAO Report, 1977). Copepods sampled on Nantucket preservation in formalin should be explored in the labora-
Shoals at the same stations as the oil-contaminated cod tory; likewise, the possibility that real surface contamina-
and pollock eggs showed external mandibular and inter- tion is readily observable only after fixation of the eggs
nal contamination with oil, as has been reported in the with the resultant change in their refractive index. Net
case of other major spills (see proceedings of this contamination is another factor to consider. The net used
symposium). Contaminating oil was of the type carried by at Station 9 was fouled collecting samples at Stations 7
the Argo Merchant Exposure of cod eggs to No. 6 fuel oil and 8. This station, interestingly, was marked by the higher

123

than usual moribundity and abnormality. It is needed because ideal conditions for such estimates can
difficult, however, to suppose that eggs readily con- not be reasonably met for most commercial fish in either
taminated while collecting nets were towed through laboratory or field alone.
the water would not also be contaminated, at least to Effects of oil on the chromosome divisions of
some degree, by mixing of oil and eggs in the water developing fish embryos have not been investigated in
column. Also, the contaminated net yielded samples at the laboratory. To some degree, damage to the fish
Station 8 less fouled than at Station 9. (No eggs collected zygotes at the chromosome and cellular levels is
at Station 7.) Another possibility is that oil becomes probably reflected in the abnormality and increased
attached to the chorion only at certain times: on mortality of fish embryos reported on their experimental
spawning of the eggs and, again, after fertilization when exposure to oil or oil fractions (Mironov, 1968, 1969,
membrane changes occur which might make oil contami- 1972; Kuhnhold, 1972, 1974, and in this symposium;
nation more-or-less likely. Struhsaker et al., 1974). The genetics group at the Milford
Heavy fouling of the Station 9 pollock eggs cannot laboratory of the Northeast Fisheries Center will be col-
be attributed simplyto mortality and ensuing deterioration laborating with W. Kuhnhold of the University of Keil, FRG,
of the membranes as these eggs were almost all only in and the Environmental Protection Agency, Narragansett
the moribund class, as opposed to other pollock eggs Laboratory, in a joint study on the toxicity of oil to cod
which were less fouled but already showing signs of eggs in laboratory culture, one facet of which will be
cytological death. cytogenetic.
Any real difference in membrane contamination of Cytogenetic examination of fish embryos could fur-
the cod and pollock may be related to different depths at ther be of use in appraisals of salinity and temperature
which these fish may spawn in the water column and the effects on embryonic development. These data would also
rate at which their eggs rise to the surface. Of course, be useful in establishing baselines. The developing embryo
duration of exposure may also be a factor. The cod eggs is limited in its functions largely to cell and chromosome
were at an earlier stage of development than were the division and cell differentiation. It might be expected then
pollock eggs. Another factor would be the species varia- that the embryo's mitoses and other cytological
tion in the chorion structure. The Argo Merchant study is phenomena might, aside from appraisals of cytotoxicity
the firstto apply scanning electron microscopy of fish egg and mutagenicity to its chromosomes, be useful as a
membranes to any investigation of this sort. gauge of the physiological well-being of the embryo and
its development rate.
Significance of the Limited Argo Merchant Egg Study and The combination of this new method of studying the
Suggestions for the Future. Generally, little is known about chromosomes and mitotic-cytological development in
the natural factors in the environment that control the fish eggs with the common practice of sampling ichthyo-
mortality of fish eggs and larvae (NOAATechnical Report, plankton at sea may well provoke a surge of interest in
1976). This, combined with the sparse samples of fish development and cytogenetics of fish. This interest could
eggs lacking a good distribution of developmental generate opportunities for measuring and understanding
stages, with eggs of all stations oil-contaminated to some the impact of oil spills and petroleum hydrocarbons on
degree, makes it impossible to determine fully the signifi- commercially important and other fish.
cance of the cytological-cytogenetic findings on the cod
and pollock eggs from the Argo Merchant spill. Nonethe-
less, the potential of @uch a study applied to a field
catastrophe is demonstrated.
Pre-planned strategy for handling field sampling of Acknowledgments
ichthyoplankton in the event of future spills can result in The special assistance of D. Perry, A. Naplin, and R.
more adequate samples for cytological-cytogenetic Riccio in conducting the Argo Merchant study is
study in the event fish eggs are abundant enough. acknowledged. K. Sherman, NEFC, provided us with the
However, suitable contemporary field controls of same
species fish eggs may be insufficient or unavailable. It is fish eggs from the spill area, and helped focus the work
essential then, if field appraisals are to be made in the and discussion on fishery matters.
future, that adequate cytological-cytogenetic baseline
data be collected on important commercial species. Such
estimates would have to take into account the influence
of variable environmental factors like temperature and References
salinity and of synergisms between them, as well as the
ubiquitous background loads of contaminants. Essential AEC-TR-6940. lvanov, V. N. 1968. Effects of radioactive sub-
also is an increase in the currently inadequate base of stances on embryonic development of fish. p. 185.
research on the development of the earliest-stage eggs AEC-TR-7418. Migalovskaya, V. N. 1971. Chronic effects of
of commercial fish. Strontium-90 and Ytrium-90 on the frequency of chromo-
It is recognizably difficult to extrapolate from labora- somal aberrations in the embryonal cells of the Atlantic
tory data on the toxicity of oil to fish eggs and embryos to salmon. p. 89.
field conditions where the oil itself is affected by the AEC-TR-7299. Marine Radioecology. 1972. Distributed by: NTIS-
National Technical Information Service. U.S. Dept.
environment and the environment influences the expo- Commerce, 5285 Port Karsal Road, Springfield, Va.
sure the eggs receive. However, such laboratory data, as 22151.
well as field samples from uncontaminated areas, are Bateman, A. J., and S. S. Epstein. 1971. Dominant lethal muta-
required to obtain control values for the cytogenetic tions in mammals. p. 541-568. In: A. Hollaender (ed.),
indices, as well as for other biological parameters that Chemical mutagens: Principles and methods for their
might be measured. Both field and laboratory controls are detection, Vol. 2, Plenum Press.

124

Cave, M. S., and S. W. Brown. @ 957. The detection and nature of NOAA Technical Report. 1976. John R. Hunter (ed.). Report of a
dominant lethals in Lilium. 111. Rates of early embryogeny Colloquium on Larval Fish Mortality Studies and Their
in no'rmal and lethal ovules. Amer. J. Bot., 44:1-7. Relation to Fishery Research. NMFS Circ.-395. 5 p,
Donaldson, L. R., and R. F. Foster. 1957. Effects of radiation on Polikarpov, G. G. 1966. Radioecology of aquatic organisms.
aquatic organisms. p. 96-102. In: The effects of atomic Transi. by Scripta Technica, Ltd. Dist. by Reinhold, Book
radiation on oceanography and fisheries. Nat. Acad. Sci, Division, New York. V. Schultz and A. W. Clement, Jr.
Nat. Res. Counc. Publ. 551. (eds.). 314 p.
Duce, R. A., J. G. Quinn, C. E. Olney, S. R. Piotrowisz, B. J. Ray Purdom, C. E., and D. S. Woodhead. 1973. Radiation damage in
and T. L. Wade. 1972. Enrichment of heavy metals and fish. p. 67-73. In: H. E. Sutton and M. 1. Harris (eds.),
organic compounds in the surface microlayer of Narragan- Genetics and mutagenics of fish. Academic Press.
sett Bay, Rhode Island. Science, 176:161-163. Roberts, E. L. 1967. Chromosome cytology of the Osteichthyes.
Epstein, S. S., S. R. Joshi, E. Arnold, E. C. Page, and Y. Bishop. Progir. Fish-Cult., 29:75-83.
1970. Abnormal zygote development in mice after pater- Rosenthal, H., and D. F. Alderdice. 1976. Sublethal effects of
nal exposure to a chemical mutagen. Nature, 225:1260- environmental stressors, natural and pollutional, on
1261. marine fish eggs and larvae. J. Fish. Res. Bd. Canada,
Food and Agriculture Organization of the United Nations. 1977. 33:2047-2065.
Impact of oil on the marine environment. Reports and Roubal, W. T., and T. K. Collier. 1975. Spin-labeling technique for
Studies No. 6, IMCO/FAO/UNESCO/WMO/WHO/IAEA/ studying mode of action of petroleum hydrocarbons in
UN Joint Group of Experts on the Scientific Aspects of marine organisms. Fish. Bull., 78:299-305.
Marine Pollution (GESAMP). Scheier, A., and D. Gominger. 1976. A preliminary study of the
Hollaender, A. 1971a. Chemical mutagens: Principles and toxic effects of irradiated vs non-irradiated water soluble
methods for their detection, Vol. 1. Plenum Press. 310 p. fractions of #2 fuel oil. Bull. Environ. Contam. and Toxicol.,
Hollaender, A. 1971 b. Chemical mutagens: Principles and 16: 595-603.
methods for their detection, Vol. 2. Plenum Press. 300 p. Solberg, A. N. 1938. The susceptibility of Fundulus heteroctitus
Kuhnhold, W. W. 1972. Untersuchungen uber die Toxizitat von embryos to X-radiation. J. Exp. Zool., 78:441-470.
Roholextrakter und emulsionen auf Eier und Larven von Sonnenblick, B. P. 1940. Cytology and development of the
Dorsch und Hering. (investigations on the toxicity of crude embryos of X-rayed adult Drosophila melanogaster. Proc.
oil extracts and dispersions on eggs and larvae of cod and Natl. Acad. Science, 26:373-381.
herring). Ph.D. Thesis. University of Kiel, FRG. Sprague, J. B. 197 1. Measurement of pollutant toxicity to f ish, 111.
Kuhnhold, W. W. 1974. Investigations on the toxicity of seawater- Sublethal effects and "safe" concentrations. Water Res.,
extracts of three crude oils on eggs of cod (Gadus 5:245-266.
morhua L.). Ber. dt. wiss. Kommn. Meeresforsch, 23: Struhsaker, J. W., M. B. Eldridge, and T. Echeverria. 1974.
165-180. Effects of benzene (a water-soluble component of crude
Larson, R. A., L. L. Hunt, and D. W. Blankenship. 1977. Formation oil) on eggs and larvae of Pacific herring and northern
of toxic products from a #2 fuel oil by photooxidation. anchovy. p. 253-284. In: F. J. Vemberg and W. B. Vernberg
Environ. Science and Technol., 11:492-496. (eds.), Pollution and physiology of marine organisms.
Longwell, A. C. 1975. Mutagenicity of marine pollutants as it Academic Press.
could be affecting inshore and offshore fisheries. Informal von Borstel, R. C., 1960, On the nature of dominant lethality
Report No. 79, Middle Atlantic Coastal Fisheries Center, induced by radiation. Atti Associazione Genetica Italiana,
U.S. Dept. Commerce, NOAA, NMFS, N.E. Region. 72 1). 5:35-50.
Longwell, A. C. 1976. Chromosome mutagenesis in devel- von Borstel, R. C., and M. L. Rekemeyer. 1959. Radiation-
oping mackerel eggs sampled from the New York Bight. induced and genetically contrived dominant lethality in
NOAA Technical Memorandum ERL-MESA-7, U.S. Dept. Habrobracon and Drosophila. Genetics, 44:1053-1074.
Commerce, NOAA, ERL. 59 p. Whiting, A. R. 1945. Effects of X-rays on hatchability and on
Macintyre, F. 1974. The top millimeter of the ocean. Scientific chromosomes of Habrobracon eggstreated infirstmeiotic
American, 230:62-77. prophase and metaphase. Amer. Nat., 59-193-227.
McIntyre, A. D. 1976. Working Group on Pollution Baseline and
Monitoring Studies in the Oslo Commission and ICNAF
Areas - Report of the Sub-Group on the Feasibility of
Effects Monitoring, C.M. 1976/E: 44. 36 p.
Mironov, 0. G. 1968. Hydrocarbon pollution of the sea and its
influence on marine organisms. Helgol. wiss. Meeresun-
ters., 17:335-339.
Mironov, 0. G. 1969. The development of some Black Sea fishes
in sea water polluted by petroleum products. Probl.
Ichthyol., 9:919-922.
Mironov, 0. G. 1972. Biological resources of the sea and oil
pollution. Moscow, Peschchevaya Promyshlennost. 105 p.
(in Russian).
Muller, H. J. 1954a. The nature of genetic effects produced by
radiation. p. 351-474. In: A. Hollaender (ed.), Radiation
Biology, Vol. 1, Part 1. McGraw-Hill.
Muller, H. J. 1954b. The manner of production of mutation by
radiation, p. 475-626. In: A. Hollaender led.), Radiation
Biology, Vol. 1, Part 1. McGraw-Hill.
Muller, H. J. 1959. Advances in radiation mutagenesis through
studies in Drosophila. p. 146-160. In: Progress in nuclear
energy, Ser. VI, Vol. 2, Pergamon Press.
Murakami, A. 1971. Radiation-induced recessive visible muta-
tions of oocytes during meiosis in the silkworm. Genetics,
67:109-120.

125

Effects of the Water Soluble Fraction of a
Venezuelan Heavy Fuel Oil (No. 6) on Cod
Eggs and Larvae

W. W. Kuhnhold

Institut fur Meereskunde
Kiel, German Federal Republic

Abstract Fisheries Service during the time of the spill and was one
of the species spawning in the vicinity of the Argo
One-half, 3, and 7 day old eggs and 2, 4, and 8 day Merchant during monitoring cruises conducted by the
old larvae were exposed to the water soluble fraction of Northeast Fisheries Center.
Bunker C oil in static tests. The initial concentrations of Oil concentrations were chosen that were expected
the test medium were 10, 100, and 500 ppb of total to produce sublethal damage which could lower hatching
CC14-extractable hydrocarbons. The hatching success of rates. Heart development and heart beat rates were used
viable larvae and the difference in 50% survival time as indicators -of the physiological status of the test
between control and treated larvae were measured. In the organisms.
same test the heart beat rates of the treated embryos
were measured, and a comparison between observed
beat rate reduction and viable hatch was drawn. An attempt Materials and Methods
was made to extrapolate laboratory findings to field con-
ditions which existed at the time of the Argo Merchant The cod eggs used in the experiment were spawned
spill. and naturally fertilized at 5-60C in a tank at the Narra-
gansett Laboratorv of the Northeast Fisheries Center.

Introduction
To date, there have been no direct observations in 0.5
the field of effects of oil on embryogenesis and larval
a_
development of fish.
Estimates of oil damage to embryos and larvae have
been extrapolated from laboratory tests under simulated
field conditions. Although these types of experiments F- 0.4 -
U_
(e.g., determination of LC50's) are limited by the con- 0
straints of the laboratory, they remain an important tool Z -
for estimating adverse effects of oil until extensive field 0
work can be done. This study was conducted to measure 0.3 -
effects of an oil comparable to that carried by the
Argo Merchant.
Z
Earlier experiments examined the effects of the W
water soluble fraction (WSF) of several crude oils on 0Z
embryonic mortality (Kuhnhold, 1972); however a No. 6 oil 0 0.2 -
consists of a residual distillation fraction of crude oil so 0 2 4 6 8 110 .12. 14 16 1
it was not possible to extrapolate from crude to a No. 6 oil.
Since no samples of the Argo Merchant cargo was TIME (DAYS)
available to use in the studies, a Venezuelan Bunker C oil
Figure 1. Curve of hydrocarbon loss ofinifial (nomina0concentra-
0

(Texas A & M) was used throughout the experiment. Cod tion of 500 ppb total CC14-extractable hydrocarbons (THQ of the
was chosen as the test organism since it was available at water soluble fraction of Venezuelan fuel oil No. -6. All concen-
the Narragansett Laboratory of the National Marine trations were obtained by dilution of original WSF of 0. 72 ppm.

126

VV5F _NO6 FLIEL OIL Table 1. LC50's (initial concentration) ppb of total
hydrocarbons of the Water Soluble Fraction (WSF) of Venezuelan
TOTAL HATC VIABLE HATCH fuel oil No. 6 for cod embryos at 3 different ages.

PPB PPB
% HC HC
1001 0 Age of Eggs
10 0 at Exposure For For
80- Do 10 (days) Total Hatch Viable Hatch

60- 0.5 150-200 20-30
40- 100 3 200-250 30-40

20-
500 7 325-375 150-200
9 13 1%5 9 13 1@5 TIME OF EXPOSURE
DAYS DAYS BEFORE HATCH

Figure 2. Comparison of total and viable hatch of cod larvae
exposed to the water soluble fraction of Venezuelan fuel oil No. 6
at the given concentrations of total hydrocarbons (THQ.

so- No.6 Fuel Cil, 500/7 added to the surface of 11.5 liters of filtered seawater in a
Cod Eggs 15 liter carboy. The water was stirred slowly with a
magnetic stirrer for 12 hours without dispersing oil
droplets into the water.
The extract was kept under the oil film for another
40- 12 hours before the water phase containing the water
soluble fraction was drawn from the carboy. During the
test the dissolved compounds were subject to evapora-
tion. Two 1 -liter water samples were taken to determine
30- the initial total hydrocarbon concentration using CC14-
extraction and IR absorption. Standards were made
from dilutions of the oil in CC14. The loss of total hydro-
UJ
carbons (THQ from the water was approximately 25% in
4 days, and 32% in 10 days, which can be considered a
LO 20- r 100/015 low reduction in THC (Figure 1). Since no data were
0 1\
Z available about the actual concentrations of the WSF of
0
U the Argo Merchant oil at the spill site, initial concentra
Ui 100/7 -extractable
100/3 tions of 10, 100, and 500 ppb of total CC14
to- 10/015 hydrocarbons were used.
10/3
Control_.
10/7 Tests were carried out in open 1-liter jars, each
containing 100 eggs or larvae and.75 liter of extract. The
temperature was 70C. After hatching, the larvae were held
0 4 and observed in the jars for 4-5 days. Larvae which
8 9 10 11 12 13 14 15 16 17 survived to this point were considered "viable". At this
AGE (DAYS) time most of the yolk is resorbed and the larvae initiate
feeding. However, they were not fed in this experiment, as
Figure 3. Influence of the water soluble fraction (WSF) of no food source was available. To investigate effects of oil
Venezuelan No. 6 fuel oil on the heartbeat of cod embryos. on cod larvae after hatching, 2, 4, and 8 day old larvae
Exposure ages are 0.5, 3.0, and 7.0 days. WSF concentrations: from control tanks were transferred to the test jars. These
10, 100and500ppb. Vertical bars represent standard deviation ages were chosen to include the range of first feeding
of 10 measurements. larvae. Forcod, the initiation offeeding occurs 5 to 6 days
after hatching. If the first feeding does not occur by day
10, the cod larvae do not survive. Exposure of 8 day old
One-half, 3, and 7 day old eggs were exposed to an larvae would show the effect of delayed feeding in
extract of the oil for exposure times of 15.5, 13, and 9 addition to the oil effects.
days, respectively. The tests were static (i.e., there was In order to record heartbeat rates, a random sample
no replacement of the test medium throughout the of ten eggs was carefully transferred to a glass dish and
experiments). retransferred to the jar after inspection under a dissecting
The extraction of the No. 6 fuel oil was prepared microscope. The time that elapsed between 10 heart-
\100

according to Hyland (1973) to provide concentrations of beats was measured with a stopwatch. Precautions were
water soluble fractions (WSF) comparable to earlier taken to avoid temperature increase during heartbeat
studies. Thirty milliliters of No. 6 fuel oil were slowly counts.

127

100 - eggs. These data are interpolated graphically and show
WSF the difference between "survival until hatching" and
80 - "viable hatch".
ioo_@
GO - 500 Embryo Heartbeat Rates. During normal development
(at 70C) the embryo heartbeat rate was regular by day 9 at
40 - an average rate of 19.3 seconds/10 beats (= 31.1
bts/min). At hatching, the rate had increased to 9.1
20 - sec/10 beats (= 65.9 bts/min).
There was no detectable effect on heartbeat rate at
0 - T -1 10 ppb total hydrocarbons (Figure 3). One hundred parts
0 2 4 6 8 0 12 14 16 18 per billion had a retarding effect on the development of
the heart. The eggs exposed at 0.5 and 3 days did not
100- WSF' show a regular heartbeat until day 10 and the rate of 25.5
- - -;,C sec/10 beats (= 23.5 bts/min) was considerably slower
_J 80 - -I than the rates for the control embryos. Although the rates
< increased with time, they failed to reach the normal rate.
@:' 60 - 100- It is evident that the rate of increase of heartbeat during
>
cr development is correlated with the length of time the
500 '10
7D 40 - embryos were exposed to the oil. The 7 day embryos
showed no immediate effects, as the heartbeat rate was
20 - still within the normal range after three days. However,
the heartbeat range did not increase and never exceeded
the range for the 3 day eggs.
0 At 500 ppb, effects could be measured for the 7 day
0 2 4 6 8 10 12 14 16 Is eggs only because mortality was 100% for the 0.5 and 3
100- @c day eggs. After 2 days' exposure at day 9 the heartbeat
WSF had decreased (20.3 sec/10 beats = 29.6 bts/min)
80 - slightly compared to the heartbeat rate of the controls
(19.3 sec/1 0 beats = 31.1 bts/min). However, after one
60 - week, the rate had dropped to 50 sec/10 beats (-- 12
i0o bts/min) and the beats were extremely irregular so that
40 - 500), timing the sequence of beats became difficult. This is
shown in Figure 3 by the large standard deviation about
20 - 10 the mean beat rate.
It was also observed that as the exposure time
increased, more of the eggs slowly lost buoyancy and
0 sank to the bottom of the test jars. In general, the eggs
0 2 4 6 8 10 12 14 16 18 which sank to the bottom had slower heartbeat rates than
AGE OF LARVAE (DAYS) the remaining floating ones.
Figure 4-Survival of unfed young cod larvae exposed to the Larval Survival. Contrary to previous tests, viable larvae
water soluble fraction (WSF) of Venezuelan No. 6 fuel oil 2, 4, and hatched from eggs incubated in clean water were
8 days after hatching (Temp. 7. O'Q. selected for larval survival experiments. At all three ages
(2, 4, and 8 days) the survival time of the larvae generally
Results decreased with increasing hydrocarbon concentration
(Figure 4). The mean survival times were compared to
Hatching. Ninety-one to 97% of the control larvae were their corresponding controls only.
viable, so3-9% can be considered within the natural range Although it appears that the low concentration (10
of variability for larval mortality (Figure 2) in this experi- ppb) actually enhances survival for the 2 and 4 day old
ment. Of the eggs in the 10 ppb treatment, a maximum of larvae, the values at 10 ppb probably reflect the natural
7% were non-viable, which is within the range of variation variability in survival rates.
of the controls. The difference in survival times between control and
However, in the 10 ppb treatment, there was a sharp 4 day old larvae is less than for 2 day larvae (Table 2)
decrease in viability. In the 0.5 day eggs the "total hatch" indicating greater resistance of the 4 day old larvae. The
rate was normal, but none of the larvae survived more than survival time of the 8 day old larvae approached that of the
a few days after hatching. The 3 and 7 day old eggs controls with the 50% mortality point occurring only 1-2
showed generally no lethal effect until hatching. It is days earlier than for the controls.
evident that increasing viability is correlated with de- An additional morphological observation was made
creasing exposure time. The ratio of per cent "total on the moribund larvae. At closer inspection a substantial
hatch" to per cent "viable larvae" for 3 and 7 day old portion of the larvae had an abnormally developed apical
eggs was 78%/13% and 93%/70%, respectively. part of the primordial fin fold. The fin was swollen and
The 500 ppb treatment proved lethal for the 0.5 and showed, even at relatively low magnification (ca. 50 x), a
3 day eggs. Forty per cent of the 7 day eggs hatched, more irregular surface than in normal larvae (Figure 5).
7%

100-
5 @0@1,@ 1,0

0
5 0 0 0

but none were viable. Table 1 gives the LC50 values for The frequency and degree of this abnormality seemed to
"total hatch" and "viable hatch" of 0.5, 3 and 7 day old be correlated with the oil concentration. In some cases it

128

Table 2. Effect of the Water Soluble Fraction (WSF) of the Venezuelan fuel oil No. 6
on mean survival time of young unfed cod larvae.

Difference in
Initial 50% Time after 50% Survival Time
Age at Concentr. of Mortality Exposure between Control and
Exposure THC Ipjob) after fdays) Began (days) Treated Larvae (days)

2 0 13.7 11.7 - - -
10 14.1 12.0 +0.4
100 11.6 9.6 -2.1
500 9.2 7.2 -4.5

4 0 15.0 11.0 - - -
10 16.3 12.3 +1.3
100 14.0 10.0 -1.0
500 11.4 7.4 -3.6

8 0 13.4 5.4
10 13.2 5.2 -0.2
100 12.0 4.0 -1.4
500 11.5 3.5 -1.9

Table 3. Factors needed for assessment of damage to fish population
by spilled oil. normal primordial
/fin

Type of oil spilled Area of spawning ground

Area covered Time of spawning
(contaminated)

Concentration of dissolved Distribution of eggs and
hydrocarbons larvae (vert. and horiz.)

Chemical characteristics Specific sensitivity a
of WSF

Dilution gradient Duration of influence
(horiz. and vert.)
,1_7s_
15 n- @p ri rn ord i'a I
n-

was evident 3 to 4 days after exposure to the WSF, with
the larvae still alive but showing signs of reduced
viability. However, this phenomenon was not evaluated

made.
quantitatively, and histological examinations were not

Discussion

Naturally, it is difficult to extrapolate from laboratory
RIF"
findings to the field conditions which existed at the time
of the Argo Merchant spill. For a comprehensive assess-
ment of the damage to the fish population, several factors
must be considered (Table 3). Some of these are known,
or at least can be estimated. The largest problem is the
lack of adequate data on the concentration of dissolved Figure 5. Deformation of the primordial fin in cod larvae exposed
.hydrocarbons in the water at the spill site, the composi- to the water soluble fraction (WSF) of Venezuelan No. 6 fuel oil.,
tion of the water soluble fraction, and most important, the a) normalshapeof the apical region of thelin, b)deformedswollen
change in concentration gradients in the water through fin. The degree of this defect depends on exposure time and
time. WSF concentration but was not evaluated quantitatively.

129

UV-analyses of hydrocarbons in water samples is, no visible effects were observed for trout but negative
taken at various times after the spill and at different effects on larval survival were observed for winter
locations showed maximum values of 250 ppb THC flounder, one cannot exclude latent effects on offspring
(NOAA, 1977). The concentrations in the test series were in contaminated areas, The question remains: How much
determined by IR spectrometry, so comparison with UV does this affect a year class? This question can be
results is difficult. addressed with simultaneous laboratory, shipboard, and
As all water samples were taken at various times after in situ observations. Consideration should be given to the
the spill and at different locations they represent only a development of new technology to deal with the mainte-
few points out of a three dimensional pattern. Assuming nance and growth of embryos and larvae at sea, following
that time is the most important factor in this system, and the pioneering work of Lasker (1975) and Laurence
that dilution of the hydrocarbons occurs in horizontal and (1978).
vertical directions, initial concentrations close to the
slicks which were not recorded should have been some-
what higher. References
Considering the deleterious effects of oil on cod Anderson, J. W., D. B. Dixit, G. S. Ward and R. S. Foster, 1976.
eggs and larvae measured in the laboratory, it is certainly Effects of petroleum hydrocarbons on the rate of heart-
possible that there were immediate or delayed effects of beat and hatching success of estuarine fish embryos. In:
oil on pelagic fish eggs and larvae in the vicinity of the Pollution and Physiology of Marine Organisms, 11. F. J. and
spill. However, the organisms in the laboratory were W. B. Vernberg, Eds., Academic Press, 1976.
continuously exposed to the WSF, whereas the embryos Colton, J. B., Jr. and J. M. St. Onge. 1974. Distribution of Fish
and larvae in the field could have been exposed to varying Eggs and Larvae in Continental Shelf Waters, Nova
concentrations of WSF for varying periods of time. Scotia to Long Island. Serial Atlas of the Marine Environ-
Kuhnhold (1972) showed that short term influence on ment, Foiio 23, American Geographical Society.
Hodgins, H. 0., B. B. McCain andJ. W. Hawkes. 1977. Marinefish
eggs (short in regard to total incubation period: 1 to 4 and invertebrate diseases, host disease resistance, and
days) of WSF of crude oils can either immediately mani- pathological eff ects of petroleum. In: Effects of Petroleum
fest deleterious effects or result in gradual effects on Arctic and Subarctic Marine Environments and
depending on the type of oil. Physiological measure- Organisms. Vol. II. Biological Effects. D. C. Malins (Ed.)
ments had not been made parallel to the hatching counts New York, 1977, pp. 95-128.
by Kuhnhold. The measurement of heartbeat rate may be Hyland, J. L: 1973. Acute toxicity of No. 6 fuel oil to intertidal
a simple prehatch-parameter to judge viability of larvae; organisms in the lower York River, Virginia. M. S. Thesis.
however, short-term effects in larvae (after several hours College of William and Mary (VIMS). 75 pp.
to 2 days) can be reversible depending on the type of oil Kuhnhold, W. W. 1972. Investigations on the toxicity of crude oil
extracts and emulsions to eggs and larvae of cod and
to which the larvae are exposed. Different species also herring. Ph.D. Thesis, Univ. Kiel (in German).
vary in sensitivity to the WSF of oils. This means that the Kuhnhold, W. W. 1974. Investigations on thetoxicityof seawater-
time beyond which adverse effects are irreversible in extracts of three crude oils on eggs of cod (Gadus morhua
larvae varies from one species to another. This was L.). Ber. dt. wiss. Kommn. Meeresforschung 23:165-180.
demonstrated for young larvae of plaice, cod and Nor- Kuhnhold, W. W. 1978. Ingestion of oil particles by zooplankters.
wegian herring, increasing in sensitivity with the order (in prep.)
mentioned (Kuhnhold, 1972). These findings suggest the Lasker, R. 1975. Field criteria for survival of anchovy larvae: the
effects of oil on cod in the field should have been relation between inshore chlorophyll maximum layers and
generally less severe than under the laboratory condi- successful first-feeding. Fish. Bull. 73:453-462.
Laurence, G. 1978. An Environmental Chamber for Monitoring
tions. "in �itu" Growth and Survival of Larval Fishes. Manuscript,
The extent of damage to the brood population also Northeast Fisheries Center.
depends on the horizontal and vertical distribution of NOAA. 1977. TheArgo Merchant Oil Spill, a Preliminary Scientific
eggs and larvae in the spill area. NEFC plankton surveys Report. NOAA Special Report. P. L. Grose and J. S.
after the Argo Merchant spill showed cod and pollock Mattson (Eds.), U.S. Department of Commerce, National
eggs present (NOAA, 1977). However, the spill did not Oceanic and Atmospheric Administration, U.S.A.
occur during the peak spawning period of either species
and it covered only a minor part of the spawning area
(Colton and St. Onge, 1974), so it is difficult to assess the
extent of acute damage to the total population.
On the other hand a quite important question arises:
What is the secondary effect of the oil on reproduction if
it accumulates in the food web and enters adult fish? It has
been observed that oil particles are ingested by zoo-
plankters; many individuals of the most abundant species
were found with oil in their digestive tracts (Kuhnhold,
1978). As most of the oil seems to pass unaltered through
the animals with the feces and is sedimented, it is made
available to another section of the food web, which is the
food basis for clemersal fish species. Recent investiga-
tions focused on the effects of incorporated crude oil on
the offspring of trout (Hodgins et al., 1977) and of water
soluble components of No. 2 fuel oil on offspring of
winter flounder (Kuhnhold, 1978). Although the results
of these two investigations are contradictory, that

130

Interactions Between Petroleum and
Benthic Fauna at the Argo Merchant
Spill Site

Sheldon D. Pratt

Graduate School of Oceanography
University of Rhode Island
Kingston, Rhode Island

Abstract the samples, review of the geology of the spill area, and
Funding was provided for collection and archiving of examination of benthic organisms in a small number of
grab samples. These preliminary studies have identified
quantitative benthic grab samplesfrom theArgo Merchant the form of oil in sediments at the spill site, the organisms
spill site. Oil was found in sediments within 3-4 km of the present at the site, and ways in which these organisms
wreck in February and July 1977. The basis for the con- may interact with spilled oil. Additional information on oil
clusions of this report are visual inspection of samples analyses and the histopathology of larger organisms
during collection and partial identification of benthos obtained on University of Rhode island cruises is presented
in 14 samples. byHoff manand Quinn (1978) and Brown and Cooper (1978).
The spill occurred on a sand ridge bordered by
gravel-bottomed channels. Both sand and gravel environ-
ments were exposed to oil. On the ridge continuous Methods
movement of sand waves may bury oil, break it into
smaller particles, or release it into the water. Sand col- Only a few grab samples were obtained on URI
lected at the bow of the wreck contained 2-0.03 mm cruise EN-002 and EN-003 (December 28-30, 1976;
diameter oil particles which only weakly adhered to sand January 26-29, 1977). These included samples from a
grains. Few aggregated or coated grains were seen. reference site 30 miles southwest of the spill area and
The channel-bottom fauna had a high standing crop qualitative samples from gravel east of the spill area.
of both sessile and motile species in all samples
obtained, but could not be described quantitatively. The
ridge sand supported few macrobenthos and a relatively
homogeneous interstitial community. At the bow of the
wreck 4-122 ppm of oil was found in February and 0.2-0.6
ppm in July. There was a slight increase in density and 0 6n -1
.41
diversity of interstitial benthos at the later date. In the SDI LITTLE GEORGES
February samples oil was observed in the guts of int
er 39 '34 W
41-N
stitial harpacticoids and a polychaete (Ophyrotrocha sp.) 3
< - 24 .22
40. 0 *31 "291 9
and adhering to the appendages of a burrowing P111 '42 25. 10.
amphipod. 0 fS '28
;02 0 11 - .26 .21 .20 .11 .6 -,2-
30, , I ; -
!a1 .12
'I W 111/ .27 -
17 '15
Introduction .13
6
1 participated in five University of Rhode Island "-40ft-
cruises to the area of the Argo Merchant grounding. A
large number of grab samples were taken for chemical
analysis and to archive for possible studies of impact of 70-W 68.
oil on individual benthic invertebrates and on benthic Figure 1. Station locations, EN-064, February 9-13, 1977. Argo
communities. This report is based on visual inspection of Merchant wreck is adjacent to station 43.

131

69*3d W25'

69040 69020'W
5,'
15

613. 064
4
Fishing Rip O'j
I Shoals 25
5 7(B 5 8. 9 70 62
0 B 0
76 61 0',
@ OP6
41*00'N -55 72 e28
e53
48* 49* 50 5 0@;2 054
20 15 20
101 25 1 15
47 //Iod
41' 20
046 00, 20

45 5
40050' 0

.44

Figure 2. Station locations, EN-005, February 22-2 7, 19 77. Argo
Merchant bow and stem are shown by initials. Stations in which 25 2
oil was observed are circled. 10 fathom contour shown. 110 5
CONTOURS IN FATHOMS

10.
On EN-004 (February 9-13, 1977) 40 stations were IOM
sampled from areas which had been in the path of the AIOOOm A@ 8 1000 In B
floating oil (Figure 1). This area extended 65 miles east- Figure 3. Topography of Fishing Rip tidal ridge. Location of Argo
southeast of Fishing Rip, the site of the spill. Oil was Merchant wreckage shown by rectangle; dashed lines are the
found in sediments only in the immediate vicinity of positions of sand wave crests inferred from 1 fathom contours of
the wreck (Hoffman and Quinn, 1978). G&GS map 0708N-53; cross sections AA'and BB'are shown
Sampling on EN-005 (February 22-27) was designed below.
to describe the distribution and form of oil in sediments
and its impact on organisms in the immediate vicinity of to a 500 ml graduated cylinder containing a dense
the wreck (Figure 2). colloidal silica solution (Ludox; sp. wt. 1.39). A film of tap
On July 22-24, 1977 a 65' vessel was used to water is floated on the Ludox (de Jonge and Bouwman,
resample oil impacted areas to establish the time course 1977). After one hour animals and particles are pipetted
of dispersion and weathering of oil and the long term from the Ludox-water interface. This material is washed
impact on organisms. and preserved. Particles still in suspension in the Ludox
Three 0.1m' Smith-McIntyre grab samples were are removed on 0.039 mm netting. The sediment is
taken at most stations. With the grab weighted to 135 kg washed and examined under a binocular microscope for
full samples were always obtained in coarse sands" heavy bodied animals and for animals or oil particles
however penetration was poor in fine compact sand and adhering to sand grains.
sample retention was poor in gravel. Large and small After the sediment had been examined an aliquot
scallop dredges were used to qualitatively sample was removed, dried, seived at 1/2 @ intervals, and grain
rocky areas. size tractions weighed and retained.
Grab samples were divided with a sheet metal plate At this time 14 samples have been at least partially
with half allocated for biological analysis. From this a
subsample (35.2 CM2 X 10 cm) was removed with a length sorted. Identification to species will not be possible in
of core tube and preserved separately. Both biological soft-bodied meiofauna because of the type of
samples were preserved in 10% buffered seawater- preservation.
formalin with rosebengal dye added on the last two
cruises. Results and Discussion
I normally separate macrobenthos with a 0.75 mm
mesh seive. This was not practical for these samples Geology. The topography of the wreck site is shown in
since many had mean grain sizes of 0.6-1.55 mm. The Figure 3. This map suggests how complex and dynamic
following flotation technique was developed to separate the area is. Fishing Rip is near the outer edge of a system
larger microfauna and macrofauna from the small sub- of sand ridges extending to Nantucket Island. These
sample. Preservative and stain is washed out of the ridges consist of debris from a former peninsula. Swift
sample on a 0.074 mm seive. The sample is slowly added (1975) has described the form and possible dynamics of

132

100 The extreme mobility of the sands on Fishing Rip
excl6des most macrobenthic animals with the exception
of the powerful burrowing amphipods. A single surf clam
160 recovered on the ridge appeared to have been swept
01
VI// along the sediment surface. On the other hand, the
Z porous and well oxygenated sand can probably support
UJ interstitial mejofauna to a great depth. The stable gravel
U 60 and rock bottoms of the channels can support dense
UJ populations of attached, clinging, and cryptic species.
CL
Within the range of grain sizes present, sorting is
probably taking place continuously and varies with the
X 40
0 position of samples on sand waves. This is suggested by
the unimodal and bimodal grain size distributions found at
station 1-70 (Figure 4). This variation is important in sedi-
ment porosity, permeability, and surface area, and in animal
20
A>- distributions. It is unfortunate that a recording fathometer
was not available on any URI cruises so that grain size and
0 oil concentration could be correlated with position of
2.0 1.4 1,0 0.71 0.5 0'3 5 0.2 5 017 7 samples on sand waves.
DIAMETER (MM) Oil in Sediments. The form of oil in sediments will deter-
Figure 4. Cumulative grain size distribution of sediment samples mine whether it is released during sand movement or is
and I/ indicate February and July samples respectively). available for ingestion by benthic organisms.
Many oil droplets did not adhere to sand grains, but
floated free during sampling, storage washing, and animal
flotation. The upper size limit of most of the floating
sand ridges in various shelf locations. Figure 3 is drawn in droplets was 1-2 mm. Droplets of 0. 1 -1.0 mm reached the
a style similar to Swift's map of the inner Nantucket Ludox surface during animal flotation while droplets of
Shoals. 0.03-0.6mm had a much slower upward velocity and were
Major features of this system are northeast- filtered from the Ludox column.
southwest sand ridges parallel to the dominant tidal Microscopic examination of sediments showed that
current flow which are separated by gravel bottom there was little tendency of oil to coat either mineral
channels. The tendency for the channels to break into e b grains or fragments of shell or echinoderm skeleton. In a
and flood sinus couplets is seen south of the wreck sit . 200 cc sample of heavily contaminated sand (1-70-4) only
Large sand waves, shown by dashed lines, III e 118 quartz grains and one shell fragment had visible oil on
parallel to the dominant flow on the ridge flanks. Swift them. Oil totally coated two particles made up of fine
(1975) notes that they become parallel to ridge crests as sediment, rust, and weathered oil which probably origi-
they get shallower. Sand waves near Fishing Rip are up nated on a vessel. There were six aggregates in which 2-3
to 6 meters high and have wavelengths of 150 and 400 sand grains were stuck together to form 1 mm particles.
meters along the lines shown in the figure. The relatively Rings of loosely adhering droplets (0.1 mm) were
flattop of Fishing Rip is presumably formed bywind wave observed on several sand grains. This pattern resulted
dispersion. Smaller sand waves and ripples from tidal when attempts were made to spread oil on wet sand
currents and surface waves are probably found through-, grains and large droplets broke into smaller ones. The
out the sandy area. continuous movement of sand on Fishing Rip would have
The sand on the Nantucket Shoals ridges must be a similar tendency to break the oil into smaller and
almost always in motion as bed load and in near-bed smaller droplets while releasing them into the water
suspension. Smith (1969) has described rapid erosion column.
and rebuilding of sand waves on a ridge in Vineyard The use of dry sediments or sediments coated with
Sound. Waves up to 6 meters high were significantly wetting agents to sink spilled oil and the observation of oil
modified in a half tidal cycle. Smith found that the ridge he coated sediments on the upper levels of beaches may
was studying was in dynamic equilibrium over long give the impression that heavy oils will adhere to wet
periods. Swift (1975) discussed various mechanisms sand, The fact that they do not may explain why beaches
which could remove sand from channels and deposit it on "clean themselves" so well.
ridges and conserve the ridge form during marine The presence of loose droplets of oil rather than
transgression. coatings on sand grains would make it available to larger
The special nature of this ridge and channel system selective deposit feeders but less available to very small
has a number of consequences on the variables of con- forms and to those feeding on sand grain coatings.
cern in this study. Once oil reached the ridge surface by
vertical transport of small particles or by.leakage from Oil-Fauna Interaction. The term "interaction" is used
sunken portions of the vessel, it was probably buried and because at the present time there is little evidence of
mixed with sand to a depth of 2-6 meters. If the oil adhered "impact" in the form of dead or dying animals or of
to sediments there would be a tendency for it to remain on changes in population densities. Physiological and
the ridge and not be swept into channels or be carried histological abnormalities have been found in some larger
great distances. If the oil did not adhere to sand and animals, but it is difficult to project these effects in space
remained buoyant, it could be released by sediment or time. Three sediment types with characteristic benthic
motion and transported away by currents. fauna were sampled during this program. The following

133

Table 1. Number of organisms for selected taxa recovered from sediment at the bow of the Argo Merchant
(Core sample 35 CM2 x 10cm, seive size 0.074 mm. See Hoffman and Quinn, 1978, for details of hydrocarbon analysis.)

February 22-27, 1977 July 22-24, 7976
70-1 70-2 70-3 70-4 70-1 70-2 70-3

Crustaceans
Harpacticoid copepods 35 24 60 62 109 600 78
Ostracods 6 1 2 - 4 53 20

Polychaetes
Ophryotrocha sp. - 22 3 4 1 5 2
Polychaete A 1 11 - - - 2 2

Nemetodes 7 7 10 10 8 72 80

Soft bodied forms (turbellarians and small
nemerteans) 30 45 25 19 28 51 50

Minimum number of species 9 8 6 4 11 12 13

Total individuals 88 ill 107 95 156 794 248

Mean grain diameter (mm) 1.55 0.80 0.93 1.3 0.74 0.60 0.66

Hydrocarbons (Mg/gm dry wt.) 11.5- - 4.0- 5.1- 0.4- 0.2 <0.2
19.7 10.2, 122 0.6
34.7

sections describe them and interactions with spilled oil production of large numbers of oil particles following the
which were observed or hypothesized. Arrow spill suggest that at times the gravel bottom fauna
1) At reference station 1 (44 m) and at the deep was exposed to high concentrations of suspended oil.
eastern stations (80-90 m), fine compact sand is found Brown and Cooper (1978) report on the pathology of
with abundant macrobenthic species. Free burrowing a hermit crab from the channel bottom adjacent to the
amphipods and isopods are numerically dominant. Sand Argo Merchant bow. This individual had apparently been
dollars are found in shallower areas and ocean quahogs exposed to a large mass of oil which it had consumed.
in deeper areas. Oil was never found in sediments of this Some horse mussels examined may also have been
type and so samples from these areas will be archived. affected by oil. Thurberg and Gould (1978) observed
2) Gravelly sediments in Great South Channel and depressed gill tissue oxygen consumption in sea
in the channels between sand ridges near the wreck site scallops and horse mussels from the spill site soon after
support diverse faunal assemblages. Biomass may be the spill. This may be a result of filteration of suspended
very high, particularly where the horse mussel (Modiolus oil particles.
modiolus) is present. A 0. 1 M2 sample in loose gravel from I know of no analysis in which oil was found in the
station 2-2 contained over 37 species and 598 indivi- gills, guts, or pseudofeces of suspension feeders. It
duals. A 1000 cc mass of sponge and colonial tunicate would be necessary to take samples for such analyses
from dredge station 64 contained over 27 macrobenthic before oil was cleared from the digestive system.
species and 224 individuals. Although sessile suspen- In summary, the channel areas around the Fishing
sion feeders make up the greatest biomass in these Rip ridge are difficult to sample quantitatively; they are
areas, species of all purchase and feeding types are physically heterogeneous and probably received doses
found. A large polychaete (Nereis pelagica) and a bivalve of particulate oil and water soluble fractions which were
(Hiatella arctica) occupy burrows in sponge. Brittle stars, variable in concentration and duration. Examination of
scale worms, nudibranchs, and turbellarians crawl on individual animals is the only way to detect impact on the
surfaces. Deposit feeding polychaetes are present, hard bottom. A knowledge of the community makeup and
possibly taking advantage of biodeposits of horse- the life history of the fauna can be used to guide choice of
mussels. Crabs and starfish are present but not usually species for analysis.
sampled by grabs. These areas are important in pro- 3) The final sediment type sampled was the motile,
viding food for ground fish (Wigley, 1968). Shelly sand making up Fishing Rip ridge and areas to the
Gravel bottom fauna was exposed to oil in the southwest. It would be inaccurate to call the macro-
channel east of the wreck site. Hoffman and Quinn (1978) benthos of this area a "community" since the term
found relatively high concentrations at Station 56 and trace suggests biological interaction and constancy. This fauna
amounts at Stations 61 and 50, north and south of Station could be considered a remnant of a community in a less
56. The presence of small particles of oil in surrounding active bottom.
sands and the observations of Forrester (1971) on the The populations of interstitial meiofauna appear to

134

cultured as bioassay organisms by Akesson (1975). The
species studied had generation times of about one month
and are deposit feeders resistant to many stresses. While
the species found on Nantucket Shoals may not be
resistant to pollutant stresses, the fact that it feeds on oil
does suggest an opportunistic life style. The counts of
0-7MM Ophryotrocha sp. made so far show no change between
February and July.
Oil ingestion by calanoid copepods has been
observed at oil spill sites by Conover (1971) and Maurer
(1977). Spooner and Corkett (1974) found that feeding
1.2 MM slowed in calanoid copepods in a suspension of equal
numbers of oil and phytoplankton particles and a concen-
tration of 10 ppm of oil. It was thus not surprising to find
harpacticoid copecods also ingesting oil. Harpacticoids
as a group may be considered stress resistant. Some
species are known to have high tolerance to oil (Dalla
Venezia and Fossato, 1977). It is not known whether the
slight increase in harpacticoid numbers found at Station 70
represents natural variation, recovery after decrease
caused by oil, or even failure to reach a higher natural
density because of the oil.
I do not have enough data to make conclusions
Imm about variation in population numbers or diversity, or
about secondary effects. Sanders (1978) examined three
grabs from two stations 10 krn west of the wreck in
December 1976, when no oil was present, and again in
February 1977, when oil was visible in the sediments. He
Figure 5. Examples of interactions between benthic organisms found a consistent decline in densities of many small
and oil from February samples. A polychaete (Ophryotrocha sp.) polychaetes including the dominant syllid species which
and a harpacticoid copepod contain ingested oil (sta. 70-2). Oil he interprets as due to transported particulate oil.
adheres to an appendage of the amphipod Psammanyx nobilis Drastic reduction in all meiofaunal groups took
(Sta. 50-1); there is no tissue distal to the oil. place over five months in enclosed marine ecosystems in
which light fuel oil was held at around 150 ppb (Elmgren,
1978). Such a non-specific response might not be the
be more consistent. At this time I have no counts from a case where oil ingestion is significant. Percy (11976) con-
coarse sand reference station. Counts from an impacted trasts a scavenging amphipod which would not eat oil
station in February and July (Table 1) show that har- tainted fish and an isopod which was "oblivious to
pacticoid copepods and turbellarians are numerically presence of oils". Percy found that the isopod was also
dominant while nematodes are relatively unimportant. It resistant to the toxic effects of oil. If such a relationship
is difficult to compare these counts with published was generally true, oil ingestorswould not be diff erentially
densities because a large seive size was used (0.071 vs. eliminated.
the usual 0.044 mm). Furthermore, only the surface 10 cm In conclusion I recommend that analyses be com-
was sampled and it is not known to what depth the fauna pleted on the small number of oil-containing samples
penetrate. Penetration to one meter would give densities taken by various investigators and the data integrated. In
per m' comparable to those found in a sandy beach by addition, more attention should be given to particulate oil
McIntyre and Murison (1973): harpacticaids 70,000- within permeable sediments. When oil particles are
considered, rather than water accommodated fractions,
1,700,000 vs. 1,000,000 and total meiofauna 250,000- oil concentration is less important than the fraction of an
2,300,000 vs. 500,000-6,700,000. individual organism's food which is oil. The Nantucket
Interstitial species could interact with oil within Shoals area directly affected by oil is comparable to
sediments by contact or ingestion of droplets or by ocean beaches in degree of sediment movement and
exposure to toxic soluble fractions released in inter- dominance of interstitial fauna. Observations on the size,
stitial water. Ingestion is possible because most of the weathering, release, and ingestion of droplets discussed
individuals are deposit feeders. in this report are important in predicting the fate and
Examples of interaction of animals with oil are effects of oil on beaches which are sites of both chronic
shown in Figure 5. Oil was found adhering to the uropod and catastrophic grounding.
of a large burrowing amphipod. Tissue appears to have
disappeared distal to the rings of oil. Other investigators
have found oil adhering to mouthparts and posterior
appendages of amphipods (Sanders, 1978).
In two heavily contaminated samples half the Acknowledgments
Ophryotrocha sp. and a few percent of the larger
harpacticoids had ingested oil. Ophryotrocha is of particu- Sample collection was carried out under NOAA
NX

lar interest because members of this genus have been contract 03-7-022-351231.

135

References
Akesson, B. 1975. Bioassay studies with polychaetes of the
genus Ophryotrocha as test animals. In J, H. Koeman and
J, J. T. W. A. Strik (eds.), Sublethal eff ects of toxic
chemicals on aquatic animals. Proc. Swedish-Netherlands
Symp., Wageningen, The Netherlands, Sept. 2-5, 1975:
121-13&
Brown, R. S. and K. R. Cooper. 1978. Histopathologic analyses
of zooplankton and benthic organisms from the vicinity of
the Argo Merchant Proceedings of the Argo Merchant
Symposium, Jan. 1978, Univ. of Rhode Island,
Conover, R. J. 1971. Some relations between zooplankton and
Bunker C oil in Chedabucto Bay following the wreck of the
tanker Arrow. J. Fish. Res. Bd. Canada 28: 1327-1330.
Dalla Venezia, L. and V. U. Fossato. 1977. Characteristics of
suspensions of Kuwait oil and Corexit 7664 and their short
and long-term effects on Tisbe bulbisetosa (Copepoda:
Harpacticoida). Mar. Biol. 42: 233-237.
de Jonge, V. N., and L. A. Bouwman. 1977. A simple density
separation technique for quantitative isolation of meio-
benthos using the colloidal silica Ludox-TM. Mar. Biol. 42:
143-148,
Elmgren, R. 1978. Personal communication. Marine Ecosystem
Research Laboratory, Graduate School of Oceanography,
University of Rhode Island,
Forrester, W. D. 1971. Distribution of suspended oil particles
following the grounding of the tanker Arrow J. Mar. Res.
29:151-170.
Hoffman, E. J., and J. G. Quinn. 1978. A comparison of Argo
Merchant oil and sediment hydrocarbons from Nantucket
Shoals. Proceedings of the Argo Merchant Symposium,
Jan. 1978, Univ. of Rhode Island.
Maurer, R. 0. 1977. Zooplankton studies. In P. L. Grose and
J. S. Mattson (eds.), The Argo Merchant Oil Spill, A
Preliminary Scientific Report. NOAA Special Report,
March 1977.
McIntyre, A. D. and D. J. Murison. 1973. The meiofauna of a flat-
fish nursery ground. J. Mar. Biol. Ass. U.K. 53: 93-118.
Percy, J. A. 1976. Responses of arctic marine crustaceans to crude
oil and oil-tainted food. Environ. Pollution 10: 155-162.
Sanders, H. L. 1978. Personal communication. Woods Hole
Oceanographic Institution, Woods Hole, Massachusetts.
Smith, J. D. 1969. Geomorphology of a sand ridge. J, Geol. 77.
39-55.
Spooner, M. F. and C. J. Corkett. 1974. A method for testing the
toxicity of suspended oil droplets on planktonic copepods
used at Plymouth. In L. R. Beynon and E. B. Cowell (eds),
Ecological Aspects of Toxicity Testing of Oils and Dis-
persants, John Wiley and Sons.
Swift, 0. J. P. 1975. Tidal sand ridges and shoal-retreat massifs..
Mar. Geol. 18: 105-134.
Thurberg, F. P. and E. Gould. 1978. Some physiological effects
of the Argo Merchant oil spill on several marine teleosts
and bivalve molluscs. Proceedings of the Argo Merchant
Symposium, Jan. 1978, Univ. of Rhode Island.
Wigley, R. L. 1968. Benthic invertebrates of the New England
fishing banks. Underwater Naturalist 5: 8-13.

136

Fish Predation on Oil-Contaminated Prey
from the Region of the Argo Merchant
Oil spill

Ray E. Bowman and Richard W. Langton

Woods Hole Laboratory
Northeast Fisheries Center
National Marine Fisheries Service
Woods Hole, Massachusetts

Abstract effect of oil on the fish stocks. Contamination of the fish,
orfish eggs, and their invertebrate preycould havealong-
The stomach contents of 21 species of fish and term effect on the Georges Bank fishery.
squid were analyzed to determine the potential impact of The purpose of this paper is to summarize the
Argo Merchant oil on the fish stocks in the Northwest findings of the investigation of the impact of the Argo
Atlantic. Important prey groups found in the stomachs of Merchant oil spill on the food habits of demersal marine
predators sampled in the region of the oil spill included fish and squid, and to demonstrate the potential pathways
amphipods, polychaete worms, rock crabs, and American for the transfer of oil residues through the food web.
sand lance. The quantities and types of foods eaten by
each predator were similar to data previously collected.
Amphipods covered with oil were found in the stomachs
of Atlantic cod and little skate. Although no oil was found
in their stomachs, American sand lance were found to
----------
feed on the same genera of copepods previously noted to
be contaminated with Argo Merchant oil. Predator-prey
relationships showed that 81 percent of the predators
that were represented ate amphipods and 43 percent ate
American sand lance, thus establishing two potential
pathways for the oil to have been passed on to the 7 0-
higher trophic levels. 16. - - 0
0 41-

0 3
29 113
Introduction
--- 100
The oil tankerArgo Merchant ran aground on Fishing ------
Rip, 29 nautical miles southeast of Nantucket Island, 71* 70* 69* 68' 67* 6%.
6
Massachusetts, on December 15, 1976. At the time, she I
was carrying 7,700,000 gallons of No. 6 fuel oil, most of 0 DelU 76-13, 12122-23,76 0 DOE ?7-07, 7/4 10,'77
which was released into the environment on December 21 AlbN 77_07, 811-911,'77 Argo Merchant Shipwreck
when the ship broke in half. This resulted in one of the Figure 1. Location of stations in the vicinity o hant
largest and most extensively studied oil spills in U.S. f the Argo Marc
history (Grose and Mattson, 1977). shipwreck where fish and squid were collected for stomach-
contents analysis. The three cruises are indicated in the figure.
Following the Argo Merchant shipwreck a number of Stations 29 and 36, where oil was found in the stomachs of
biological studies were initiated to assess the impact of Atlantic cod and little skate, respectively, are indicated by the
the oil on the ecosystem. Although winds and currents appropriate station numbers. Stylized oil-slick borders, shown in
carried the oil offshore, eliminating the direct threat of oil the figure as dashed lines, are for December 22, 1976, and
on the New England coast, there was concern about the January 3, 1977.

137

Table 1. Stomach contents offish and squid collected in the vicinity of the Argo Merchant oil spill. The data are expressed as a percent of the
total quantity of prey consumed by each predator. The number of stomachs examined, the number found empty, the average weight of
prey per stomach, and the length range of the fish examined is given for each predator at the bottom of the table.

Predator Species
stomach-contents Smoo th Spiny L ittle Winter Thorny Alewife A t1an tic Haddock Silver
group dogfish dogfish skate skate Ska te cod hake

Cnidaria 3.3 0.1 21-1 + + 59.5

Polychaeta 0.2 0.1 8.7 28.6 29.5 1.7 24.2

Crustacea 98.9 18.5 73.5 L.0- @L- 99.1 32.4 1.4 97.8
Amphipoda 0.7 43.7 + + 96.6 11.3 0.8 83.8
Isopoda 1.3 1.3 0.2 0.2 0.3
Cancriclae 88.5 15.2 15A 0.2 9.5
Other Decapoda 10.4 2.6 12.5 5.5 4.1 9.5 0.4 11.0
Other Crustacea + 0.6 + 2.5 1.9 + 2.7
Mollusca 31.9 3.5 6.6 0.1 1.3 .2.0
Bivalvia 12.2 0.7 6.6 0.5 1.9
Gastropoda 0.8 0.1
Cephalopoda 19.7 2.8 0.1

Echinodermata + 1.3
Echinoidea + 0.1
Ophiuroidea 1.2

Urochorclata

Pisces 0.8 42.3 6.8 43.6 4-3-4 62.1 1.6 0.1
Rajidae 5.9
Clupeidae 2.5
Gadidae 5.9 2.8
Ammodytidae 0.2 1.4 2.7 36.4 44.4 1.6
Other pisces 0.6 26.6 1.3 7.2 43.4 17.7 0.1

Miscellaneous 0.1 3.7 6.5 13-9 22.9 0.9 2.2 0.4 2.1

Sand and Rocks 0.2 0.9 0.2 0.3 9.6

Number examined 8 68 89 23 4 10 81 21 14
Number empty 0 18 7 4 0 0 3 2 6
Mean weight per
stomach (g) 91.8 7.0 2.0 6.3 17.3 2.5 18.1 11.3 0.4
Length range (cm) 93-101 30-100 30-51 52-104 73-85 22-49 33-100 34-72 20-38

Materials and Methods total weight of all stomach contents for each predator.
Stomachs were collected from fish and squid The actual weight constituting each prey group may be
caught with an otter trawl during three cruises conducted calculated by multiplying the mean weight per stomach by
by the Northeast Fisheries Center (NEFC). The first two both the number of fish examined and the appropriate
cruises were carried out immediately after the spill by percent weight. Stomachs were considered empty when
the R/V Delaware /I during December 22-23,1976 (Dela- the amount of any debris found in the stomach weighed
ware 1176-13), and January 4-10,1977 (Delaware 1177-01 less than 0.001 g. All common names of fish mentioned in
the text and tables are those recommended by the
(Figure 1). A follow-up cruise, 8 months later, was carried American Fisheries Society (Bailey, 1970).
out by the RIVAlbatross IV during August 1-September 1,
1977 (Albatross IV 77-07) (Figure 1). A total of 514
stomachs from 21 species of fish and squid was collected Results and Discussion
during the three cruises. The stomachs were excised
aboard ship, individually labeled according to species, Contaminated Prey. Three hundred and five stomachs from
length, sex, station, and cruise, and then preserved in 16 species of fish were collected on the first two cruises.
10% formalin. At the NEFC's Woods Hole Laboratory the Oil, which was subsequently identified as the same type of
stomachs were opened and the contents washed onto a oil carried by theArgo Merchant (see MacLeod et al., 1978),
fine-mesh screen. The stomach contents were then was fcund on prey in the stomachs of two species of fish.
transferred to trays, manually sorted, identified to the In two Atlantic cod stomachs, gammaridean amphipods
lowest possible taxon, blotted dry, weighed to the nearest (Gammarus annulatus andAnonyx sarsi) were fouled with
0.001 g, and the weight recorded. Results are presented oil. These fish were caught at Station 29, which was
as the percent that each prey group contributed to the approximately 25 nautical miles southwest of the wreck.

138

Table 1. (cont.)

Predator Species
Stomach-contents Pollock Red hake Ocean American Butter- Sea Longhorn Window- American
group pout sand fish raven sculpin pane plaice
lance

Cnidaria 3.9 .0.1 + +
Polychaeta 5.1 10.1 0.5 1.9 93.5

Crustacea 29.6 76.3 12.6 50.0 1.0 82.5 97.2 74.9
Amphipoda 0.2 16.3 1.9 + 6.6 54.0
lsopoda 1.1 0.1 0.2 +
Cancridae 14.0 9.2 82.5 2.3
Other Decapoda 7.5 2.8 0.2 + 64.2 20.5
Other Crustacea 21.9 42.1 1.2 50.0 1.0 + 23.9 0.4

Mallusca + 0.6 + + 0.2
Bivalvia 0.6 + 0.2
Gastropoda
Cephalopoda +
Echinodermata 70.7 6.5
Echinoidea 70.7
Ophiuroidea 6.5

Urochordata 99.0

Pisces 69.8 0.6 9.0 1.1 13.3
Rajidae
Clupeidae
Gadidae 0.1
Ammodytidae 68.0 0.9 11.4
Other pisces 1.8 0.6 8.9 0.2 1.9

Miscellaneous 0.2 2.3 1.5 50.0 0.1 1.0 9.9
Sand and Rocks 0.4 11.2 .5.0 8.4

Number examined 10 19 17 8 2 9 22 33 5
Number empty 0 1 1 3 0 2 3 17 3
Mean weight per
stomach (g) 23.0 0.9 12.4 0.001 0.3 17.3 1.1 0.8 0.4
Length range (cm) 28-87 26-39 46-81 13-19 18 18-32 23-31 27-38 28-48

Oil was also found on a caprellid amphipod in the stomach previous data was shown by the haddock, which at one
of one little skate which was collected at Station 36 station fed heavily on cerianthid anemones. This prey is
(Figure 1). not usual in the haddock's diet, and yet accounted for
59.5% of the stomach contents weight. Haddock in this
Food Habits. The food habits of fish and squid from the region have previously been noted to feed mainly on
area of the Argo Merchant shipwreck have been sum- crustaceans, mollusks, echinoderms, worms, and fish
marized in Table 1 and differ little from previous published (Wigley, 1956; Wigley and Theroux, 1965; Bow-
and unpublished data on the food habits of fish from the man, 1975).
Northwest Atlantic (Bigelow and Schroeder, 1953; Leim
and Scott, 1966; Tyler, 1971, 1972; Bowman, 1975; Major Prey. The analysis of fish and squid stomach
Maurer and Bowman, 1975; Bowman et al., 1976). contents showed that the majority of the predators from
However, several noteworthy observations were made the area of theArgo Merchant shipwreck utilized four main
during this study. Apparent local abundances of prey groups: amphipods, polychaete worms, rock crabs
certain prey items have influenced the food habits of two (Cancer), and American sand lance. The importance of
species of fish. First, spiny dogfish were found to have these foods was established by determining the number
fed on sea scallop (Placopecten magellanicus) viscera of predators which had eaten the same prey group and
which constituted 12.2% (see Bivalvia, Table 1) of their the quantity of that prey consumed by each predator. If a
diet. Very little shell or adductor muscle was found in their given prey group was eaten by more than 40% of the
stomachs. Researchers in the past noted that haddock predators and made up 5% or more of the total stomach-
and Atlantic cod also fed on sea scallop viscera (Wigley, contents weight for all of the 21 predators, it was con-
1956; Bowman, 1975). This was due to scallop fishermen sidered major prey.
cleaning the scallops and discarding the Viscera while at Calculations made directly from Table 1 show that
sea. Secondly, the most noticeable deviation in diet from 60% of the total stomach-contents weight of the 21

139

Table 1. (cont.)

Predator Species Marine
Stomach -con tents Yellowtail Win ter Illex Mammals
group flounder flounder illecebrosus

Cnidaria + 32.8

Polychaeta 38.9 5.1
Man
Crustacea 44.4 54.3 5.1
Amphipoda 29.7 51.3 +
lsopoda 0.3 +
Cancridae 0.2
Other Decapoda 14.1 2.7 0.3
Other Crustaces 0.3 0.1 4.8
Red 1Hoke
Mollusca 4.0 0.4
Bivalvia 3.7 0.3 Silver Hake
Hadalb@ck@
Gastropoda 0.3 0.1
Cephalopoda Marine Fish
Echinodermata + (including Commerctol Spec4@s)
Echinoidea
Ophiuroidea +

Urochorclata 4.4
Pisces + + 66.9 Sandlonce
Rajidae
Clupeidae
Gadidae
Ammodytidae
Other pisces + + 66.9
Miscellaneous 7.8 1.5 28.0 Copepods 4mphipods
Sand and rocks 4.9 1.5 1
Number examined 40 19 10 Figure 2. Two potential pathways for the transfer of oil residues
through the food web.
Number empty 10 7 3
Mean weight per
stomach (g) 0.9 1.0 0.9 tail flounder (38.9%), thorny skate (29.5%), winter skate
Length range (cm) 23-42 18-46 21-24 (28.6%), haddock (24.2%), and ocean pout (10. 1 %).
Rock crabs accounted for 25.6% of the total
stomach-contents weight and occurred in 10, or 48%, of
the 21 predators. Large quantities of these crabs were
found in the stomachs of smooth dogfish (88.5%), sea
raven (82.5%), little -skate (15.4%), spiny dogfish (15.2%),
predators was composed of the four main prey groups and red hake (14.0%).
previously mentioned. Each of these groups is discussed The last major prey, American sand lance, repre-
below from the standpoint of: (1) the percent each group sented 21.6% of the total weight of the combined stomach
constituted of the total stomach-contents weight for all contents. Nine, or 43%, of the 21 predators utilized sand
predators combined, (2) the number of predators found lance as prey. They were found in large amounts in the
with each particular prey in their stomachs, and (3) the stomachs of pollock (68.0%), Atlantic cod (44.4%), winter
fish and squid whose diet contained more than 10% by skate (36.4%), and windowpane (11.4%).
weight of each prey group. Food Web Transfer of OiL Reports on the accumulation
Amphipods, primarily the gammaridean amphipod and transfer of oil through the food web have been
Gammarus, made up 7.9% of the total stomach-contents published by a number of different authors (see reviews
weight and were found in the stomachs of 17, or 81 %, of in Wolfe, 1977; Malins, 1977). Zooplankters, for example,
the 21 predator species. The largest quantities of have been observed to ingest oil droplets suspended in
amphipods occurred in the stomachs of the alewife the water column following several oil spills (Conover,
(96.6%), silver hake (83.8%), windowpane (54.0%), winter 1971; Maurer, 1977). In the laboratory it has been shown
flounder (51.3%), little skate (43.7%), yellowtail flounder that zooplankton, in particular Calanus helgolandicus, will
(29.7%), red hake (16.3%), and Atlantic cod (11.3%). accumulate hydrocarbons from their diet and that this is a
The second prey group, polychaete worms, consti- more important route for the biological transfer of petro-
tuted 4.9% of the weight of the stomach contents and leum residues than uptake directly from solution (Corner
were eaten by 14, or 67%, of the predators. They were et al., 1976). Brown shrimp (Crangon crangon) were also
heavily preyed upon by American plaice (93.5%), yellow- reported to eat sunken crude oil during laboratory studies

140

on the toxicity of oil that had been deposited on the Bowman, R. E. 1975. Food habits of Atlantic cod, haddock, and
bottom, after treatment with a sand-slurry and wetting silver hake in the Northwest Atlantic, 1969-1972. U.S. Nat.
agent (Blackman, 1972). In a later study, Blackman (1974) Mar. Fish. Serv., Northeast Fish. Ctr., Woods Hole Lab.
found that oil-contaminated shrimp were eaten by plaice Ref. 75-01. 53 pp.
(Pleuronectes platessa) more frequently than the uncon- Bowman, R.E., R. 0. Maurer, and J. A. Murphy. 1976. Stomach
contents of 29 fish species from five regions in the North-
taminated controls. He concluded that the oil affected the west Atlantic -- data report. U.S. Nat. Mar. Fish. Serv.,
shrimps' behavior and made them more available for Northeast Fish. Ctr., Woods Hole Lab. Ref. 76-10. 37 pp.
predation.- In the same study it was found that plaice Conover, R. J. 1971. Some relations between zooplankton and
would intentionally eat sunken crude oil. Horn et al. (1970) Bunker C oil in Chedabucto Bay following the wreck of the
showed that prey associated with tar balls in the tanker Arrow J. Fish. Res. Bd. Can. 28:1327-1330.
Mediterranean Sea and eastern North Atlantic Ocean Corner, E. D. S., R. P. Harris, C. C. Kilvington, and S. C. M. O'Hara.
were readily consumed by the saury, an epipelagic fish 1976. Petroleum compounds in the marine food web:
and important component of the ocean food web. short-term experiments on the fate of naphthalene in
Following theArgo Merchant shipwreck, oil was not Calanus. J. Mar. Biol. Ass. U.K. 56:121-133.
Grose, P. L., and J. S. Mattson, eds. 1977. TheArgo Merchant oil
found in large quantities or as isolated clumps in the fish spill, a preliminary scientific report. NOAA Spec. Rep.,
stomachs examined. It is clear, however, that prey March 1977. 133 pp. + appendices.
contaminated by oil was not avoided. The significance of Horn, M. H., J. M. Teal, and R. H. Backus. 1970. Petroleum lumps
this observation is that it establishes several pathways by on the surface of the sea. Science 168:245-246.
which oil could be transferred to higher trophic levels Leim, A. H., and W. B. Scott. 1966. Fishes of the Atlantic coast of
(Figure 2). The identification of oil in the gut of several Canada. Fish. Res. Bd. Can. Bull. 155. 485 pp.
species of copepods (Maurer, 1977) forms the basis of MacLeod, W. D., M. Uyeda, L. C. Thomas, and D. W. Brown. 1978.
one of two pathways. Three genera of these copepods, Hydrocarbon patterns in some marine biota and sedi-
Calanus, Centropages, and Pseudocalanus, were identi- ments following the Argo Merchant spill -- a preliminary
report. In Argo Merchant Symposium, Center for Manage-
fied as prey of the American sand lance. The sand lance ment Studies, Univ. of Rhode Island.
itself is a very important prey of many commercial fish Malins, D. C., ed. 1977. Effects of petroleum on arctic and sub-
species, i.e., Atlantic cod, haddock, and yellowtail arctic marine environments and organisms. Vol. 11. Biologi-
flounder (Scott, 1968). The sand lance is also a potential cal Effects. Academic Press, Inc., New York. 500 pp.
prey of whales (Bigelow and Schroeder, 1953; Nemoto, Maurer, R. 0. 1977. Zooplankton studies. Sect. 41.1. In P. L.
1959; Overholtz and Nicolas, 1978), thus extending the Grose andJ. S. Mattson, eds. The Argo Merchant oil spill, a
pathway for the transfer of oil residues to higher marine preliminary scientific report. NOAA Spec. Rep., March
trophic levels, 1977. 133 pp. + appendices.
The second pathway begins with amphipods, such Maurer, R. 0., and R. E. Bowman. 1975. Food habits of marine
as those that were contaminated with oil, which are an fishes of the Northwest Atlantic -- data report. U.S. Nat.
Mar. Fish. Serv., Northeast Fish. Ctr., Woods Hole Lab.
extremely important prey of marine fish. Over 40 different Ref. 75-03. 90 pp.
species of fish in the Northwest Atlantic are known to Nemoto, T. 1959. Food of baleen whales with reference to whale
prey on amphipods (Maurer and Bowman, 1975). Further- movements. Whale Res. Inst. Sci. Rep. 14:149-290.
more, juvenile fish of such species as Atlantic cod, Overholtz, W. J., and J. R. Nicolas. 1978. Northwest Atlantic fin-
haddock, silver hake, red hake, yellowtail flounder, back (Balaenoptera physalus) and humpback (Megaptera
American plaice, and winter flounder depend heavily novaeangliae) whale feeding activity on the American
upon amphipods as a food source (Bigelow and sand lance (Ammodytes americanus). (in preparation).
Schroeder, 1953; Leim and Scott, 1966; unpublished Scott, J. S. 1968. Morphometrics, distribution, growth, and
observations). Since many of these species are exploited maturity of offshore sand lance (Ammodytes dubius) on
the Nova Scotia banks. J. Fish. Res. Bd. Can.
commercially, this extends the pathway of oil directly 25:1775-1785.
to man. Tyler, A. V. 1971. Monthly changes in stomach contents of
demersal fishes in Passamaquoddy Bay, N.B. Fish. Res.
Bd. Can. Tech. Rep. 288. 9 pp. + appendices.
Acknowledgments Tyler, A. V. 1972. Food resource division among northern marine
dernersal fishes. J. Fish. Res. Bd. Can. 29:997-1003.
We thank Dr. Roland Wigley for his guidance in the Wigley, R. L. 1956. Food habits of Georges Bank haddock. U.S.
preparation of this manuscript, and Eric George, Nancy I Fish Wildl. Serv: Spec. Sci. Rep. -- Fish. 165. 26 pp.
Kohler, Sidney Worthen, and Richard Brodeur for their Wigley, R. L., and R. B. Theroux. 1965. Seasonal food habits of
assistance with the stomach-contents analysis. Highlands Ground haddock. Trans. Am. Fish. Soc.
94:243-251.
Wolfe, D. A., ed. 1977. Fate and effects of petroleum hydro-
carbons in marine ecosystems and organisms. Pergamon
References Press, Inc., New York. 478 pp.

Bailey, R. M., chairman. 1970. A list of common and scientific
names of fishes from the United States and Canada, 3rd
ed. Amer. Fish. Soc. Spec. Publ. No. 6. 150 pp.
Bigelow, H. B., and W. C. Schroeder. 1953. Fishes of the Gulf of
Maine. U.S. Fish Wildl. Serv. Fish. Bull. 53. 577 pp.
Blackman, R. A. A. 1972. Effects of sunken crude oil on the
feeding and survival of the brown shrimp, Crangon
crangon. Int. Counc. Explor. Sea. CM 1972/K:13. 8 pp.
Blackman, R. A. A. 1974. Effects of sunken oil on the feeding of
plaice on brown shrimps and other benthos. Int. Counc.
Explor. Sea. CM 1974/E:24. 7 pp.

141

Effect of the Argo Merchant Oil Spill on
Bird Populations off the New England
Coast, 15 December 1976 - January 1977

Kevin D. Powers
Manomet Bird Observatory
Manomet, Massachusetts

William Timothy Rumage
Biology Department
Boston University
Boston, Massachusetts

Abstract spills by determining the species and numbers of oiled
birds found on beaches (Bourne et al., 1967; Greenwood
Bird observations were conducted near the site of et al., 1971; Hope-Jones et al., 1970; Brown et al., 1973).
Argo Merchant from 15 to 24 December 1976. During Shore counts of live and dead oiled birds, however, give a
that period 1120 birds of 13 species were recorded. distorted and inadequate picture of the actual mortality
Almost 92 percent of the birds sighted were gulls; Great suffered from a spill (Taning, 1952; Hawkes, 1961; Erick-
Black-backed (Larus marinus), Herring (L. argentatus), son, 1962; Hope-Jones et al., 1970). Therefore, this study
and Black-legged Kittiwake (Rissa tridactyla). Approxi- presents information gathered at sea on the distribution
mately 59 and 41 percent of the total number of Herring of oiled and non-oiled birds during and after the Argo
and Great Black-backed Gulls, respectively, were visibly Merchant spill.
oiled. Observations made at sea during January 1977 Although the physiological effects and toxicity of oil
indicated that more birds were present outside the oil on waterfowl have been examined (Hartung, 1963,1964,
slick area but a larger proportion of visibly oiled birds 1965, 1967; Hartung and Hunt, 1966), no comparable
were within the affected area. information has been documented for marine birds. In this
From 20 December 1976 to 24 January 1977,69 live study, necropsies were performed on a sample of birds
and 112 dead birds of 16 species were collected from the killed by the Argo Merchant spill to determine which
beaches of Nantucket Island and Martha's Vineyard. organs were affected, and the probable causes of death.
Alcids (49%), gulls (27%), and loons (119%) were the most
numerous species found. Fifteen specimens of 5 species
of beached birds were examined internally, and it was Methods
found that the lungs and kidneys were the vital organs Pelagic Observations. From 15 to 24 December 1976, bird
most seriously affected. observations were made during daylight hours from the
Although bird abundances are difficult to estimate USCGC Vigilant, which was positioned near the grounded
at sea, and shore counts of beached birds are not repre- tanker, Argo Merchant Species, abundance, and
sentative or accurate of actual mortality from an oil spill, behavior of each bird sighting were recorded as well as
data indicate that the Argo Merchant spill probably had the occurrence of oiled plumage.
minimal effects on coastal and marine bird populations In January, 1977, quantitative bird observations
off the New England coast. were made during three cruises to Nantucket Shoals and
Georges Bank. The cruises were on RIVDelaware // from
Introduction 4-10 January, USCGC Decisive from 21 January to 02
February, and R/V Endeavor from 26-30 January. Obser-
Oil spills and oil pollution present hazards to coastal vations of coriimon to abundant species were quantified
and marine birds. Previous authors have based their using "acceptable 10-minute watches." An acceptable
estimates of total mortality from coastal and offshore 10-minute watch is a count of the total number of each

142

A ARGO IERCHANY I ARCO MERCHANT

AVERAGE NO. BIRDS
A v MEN PER A) MINUTES
OF ACC PTABLE 10-MINUTE 1 0 HOME SEE.
WATCHES FOR EACH BLOCK I).S

T1@
IS SO ..o
........... 31 60 ......... 4D tSo.o 1OG-o
III B IDID'o
T T Y
Figure 1. Effort, January 1977. Figure 2c. Distribution of Herring Gull, January 1977.

A IRGO M-T X WERCKPINT

---------------

A
A SAVERAGE NO. BIRDS AVERAGE NO. BIROS
EEN PER 10 MINUTES
A MEN PER 10 MINUTES
0 IS e NONE B EN NONE BE"
o.3
....... o3 .1.45
0 a ED - --------- 411 ko - I.o
IGO E) a (b
o B.0 10.0
IS A 10.0 W.0
@ . 30.0 100.0
.. . ....... A 30.0 - tw.O
A HOO.0
V Y - III- --------- Y
Figure 2a. Distribution of Northern Fulmar, January 1977. Figure 2d. Distribution of Black-legged Kittiwake, January 19 7 7.

AND. ENCHANT I ARGO MERCHANT

AVERACE NO. BIRDS
SEEN PER 10 MINUTES
o NONE SEEN
03
GS
'0 3.0 --- -----
A ". I... - ----- INDICATES A IO'N O'W BLOCK
AID.. W.0 IN WHICH T.E.E IS AT 1E.S1
A30.0 1100.0 ONE RECORD OF THE SPECIES
A

A ERCHANI
"n rv A

A AN- ME-1

A
-r@w"

X AND. MERCHANT . @ARWME..C ANT

T Y
Figu re 2 b - Distribution of Great Black-backed Gull, January 19 7 7. Figure 3a. Sightings of Large Auks, January 1977.

143

A 4

x ARGO.- A ARGO MEREHANT.

o o e o,_,i'

0
v REMENT OILED .114GS
MEN PER 10 MINUTES
0 NONE SEE.
5.D
.. . ... a 5.0 -15.0
@Tts A LOCK E) G--Q--e 15.0 W.
p -IN THERE IS AT LEAST 50.0
ONE OF THE SREEIE3 ......... (D .50.0 10.0

Figure 3b. Sightings of Dovekies, January 1977. Figure 4c. Distribution of Oiled Herring Gulls, January 1977.

X ARGO MERCHANT
9 ARGO MERENANT

o cc) cc) (o) 0-

0 0 o
e 0 E)

0 lp 9f ED 9 0 0
x
PERCENT OILED BIRDS KRCENT OILED BIRDS
0 0 0 MEN REA 10 MINUTES 0 0 0 9 0 91" 11 MEN PER tO MINUTES
0 NONE MEN 0 NONE SEEN
0 o4 :'. _' . () 0,0 5D
0 0 0 C) 0 5.o mo 0 9 10 0 0 9 ---------- Q) 5 0 15.o
........... SO o
0 0 .,..o ...... 0 "o'. Mo

T Y V T 11, 1
Figure 4a. Distribution of Oiled Northern Fulmars, January 1977. Figure 4d. Distribution of Oiled Black-legged Kittiwakes, January
1977.

p,
X ARM MERCHANT P

X ARGO IERE T
0 0 E) 9_

9""'

-1ZA 0 a 0 0 PERCENT ft. SRGDS
0 0 ID 0 9 yv,' MEN PER W MINUTES
0
--- ....... 9
DD 5.
5.o DID
11 A---.
j
A ARGO MEREHANT -
'@n ,r..

X ARM MXRC-T
A R-G0- _ERC...,T
0

.... .. 5o.o To.o
I 1 0 1. Y 7 ...... ..........

Figure 4b. Distribution of Oiled Great Black-backed Gulls,
Januar@ 1977. Figure Sa. Map of Oil Slick, 21 December 1976.

144

(3) Lungs, neys, brain and liver were examined
0 for congestion.
(4) The digestive tract was examined for food and
parasites.
(5) The remainder of the respiratory tract was
examined for blockage.
X ARGO MERCHANT
(6) Lung, kidney, and liver sections from all
specimens were preserved in 10% buffered formalin.
Other tissue samples (e.g. heart, intestine, gizzard) were
taken from some specimens. All tissue samples collected
were placed on file at the Biological Science Center at
Boston University.
a A

Results
Pelagic Observations. From the USCGC Vigilant, 1120
birds of 13 species were recorded (U.S. Dept. of Comm./
T T .1. NOAA, 1977) with daily total counts ranging from 28 to
261 birds. Almost 92% of the total number of birds
Figure 5b. Map of Oil Slick, 27 December 1976. counted were gulls, Great Black-backed (L. marinus),
Herring (L. argentatus), and Black-legged Kittiwake
(R. tridactyla); 6% were Gannets (Morus bassanus);
1 % were fulmars (Fulmarus glacialis) and alcids (Alcidae).
The incidence of oiling was highest for Great Black-
backed and Herring Gulls, at 41 % and 59%, respectively, of
the total for each species, but was lower for Gannets (12%)
and kittiwakes (9%).
In January a total of 146 acceptable 10-minute
atches were made during three cruises to Nantucket
X ARGO MERCHANTI w
Shoals and Georges Bank. Effort, i.e. the number of
acceptable 10-minute watches made per 30'N x 30'W
--------- latitude/longitude block, is indicated in Figure 1. Maps
indicating the average number of individuals counted per
acceptable 10-minute watch for fulmars, Great Black-
A backed Gulls, Herring Gulls, and Kittiwakes are compiled in
Figures 2a-d. Maps indicating 1 O'N x 1 OW blocks in which
there was at least one sighting of large auks (Alca torda or
Uria spp.) and Dovekies (Plautus al/e) are illustrated in
Figures 3a-b. Using all acceptable 1 0-minute watches, per-
------ centages of the number of visibly oiled birds to total
number of each species inspected for oiling were
mapped for fulmars and the three gull species in each
Figure 5c. Map of Oil Slick, 2 January 1977. 30'N x 30'W latitude/longitude block surveyed (Figures
4a-d). The results indicated a higher number of fulmars and
gulls outside of the oil slick area but a larger proportion of
species within sight of the ship with the aid of binoculars, visibly oiled birds near the grounded tanker and oil slick
when the ship is moving on a fixed course and at a con- track. The movement and area covered by the oil on 21
stant speed, 4 kts or faster, and when visibility is greater and 27 December 1,976 and 02 January 1977 are indi-
than 1 km. (Brown et al., 1975). This system of counting cated in Figures 5a-c (U.S. Dept. of Comm./NOAA, 1977).
birds often results in inflated estimates of certain species A total of 47 large auks was recorded in 16 sightings
because of repeated counts of the same ship-following (Figure 3a). All sightings, except one, were made within 40
individuals. All sightings of rare or infrequently occurring nautical miles of the wrecked tanker. Identification of the
species were recorded. The species, number, age, large auks to species was difficult because they were
behavior, and, when possible, the presence and degree usually seen sitting on the water and observation was
of oiling were recorded. often hampered by rough seas and strong winds. No
From 20 December 1976 to 24 January 1977, the visibly oiled large auks were seen on any of the January
beaches of Nantucket Island and Martha's Vineyard were cruises, but most were not observed at a range that would
patrolled for stranded birds. Each bird found, whether permit detection of oil on the underparts. A total of 91
dead or alive, was numbered, and the species, sex, age, Dovekies was recorded in 19 sightings (Figure 3b). Most
date, location and degree of oiling were recorded. Dovekie sightings were made on the Northern Edge and
Necropsies were conducted as follows: Northeast Peak of Georges Bank, away from the spill. In a
X MERCHANT

(1) An external search for lesions and the degree of flock of 54 Dovekies seen at 40' 42'N, 66' 51'W on
oiling was made. 6 January 1977, 10 were visibly oiled. Two additional
(2) A mid-ventral incision was made from the oiled Dovekies were sighted at 41'21'N, 66' 57'W, on the
sternum to the vent. same day.

145

Table 1. Numbers of Birds, by Species, Collected from Beaches probably due to blockage of the capillary spaces by
at Nantucket Island and Martha's Vineyard, 20 December 1976 ingested oil, and 2 specimens had lipid pneumonia. These
to 24 January 1977, findings are similar to those made by Hartung and Hunt
(11966) on waterfowl. Two types of renal pathology were
observed. The first consisted of cellular debris
Percent apparently causing a blockage of Bowman's capsule, and
Species Live Dead Total of Total the second involved a precipitation of urates in the
Common Loon 11 21 32 18 tubules, which again resulted in blockage of urine flow.
Red-throated Loon 1 1 1 The result in either case was the same; the bird was
Gannet 1 1 .2 1 unable to filter waste products from the blood, which
Grebe sp. 1 1 1 resulted in uremic poisoning. No oil was found in the
Double-crested Cormorant 1 1 1 kidneys, but the digestive tract would have broken down
Cormorant sp. 1 1 1 the oil into other by-products, thereby making the oil
Common Eider 2 3 5 3 macroscopically undetectable. Since these findings
White-winged Scoter 1 1 1 match the descriptions of oil-induced kidney changes
Common Scoter 1 1 1 studied by Hartung and Hunt (11966), we assume that the
Great Black-backed Gull 2 29 31 17 pathologies observed are oilrrelated. It is possible, how-
Herring Gull 15 15 8 ever, that similar results could have been produced by
Bonaparte's Gull 1 1 1 dehydration. There were three cases in which both the
Black-legged kittivvake 1 1 2 1 lung and kidney were congested, but neither organ alone
Razorbill 16 10 26 14 appeared to be sufficiently affected to cause death.
Common Murre 32 17 49 27 One Common Murre (Uria aalge), which had only
Thick-billed Murre 3 7 10 6 traces of oil on its plumage, was not a victim of an oil-
Murre sp. 1 1 1 related death, but had a chronic infestation of parasitic
Dovekie 1 1 1 flukes in the kidney and digestive tract. No taxonomic
69 112 181 identification of the, parasites was made.

Beached Bird Survey. Birds started to wash ashore on Discussion
Nantucket Island 5 days after the tanker went aground. To best determine the effect of an oil spill on pelagic
From 20 December 1976 to 24 January 1977, 173 birds birds, sea observations are essential. Beached bird
were collected from beaches at Nantucket Island surveys give a low estimate of the actual harm done, but
(Cardoza, 1977a, 1977b) and 8 oiled birds washed ashore more importantly, beach surveys do not provide any data
on Martha's Vineyard (Gus B. David, personal communica- to determine the potential threat presented by an oil spill.
tion). A total of 69 live and 112 dead birds of 16 species Purely marine species, such as auks and fulmars, do not
were handled (Table 1). From the total number of birds come ashore unless they are severely oiled (Bourne,
found during the beached bird surveys, it appears that 1968). Many partially oiled birds remain at sea and
alcids (49%), gulls (27%), and loons (19%) sustained the eventually die there, and because the water-repellent
greatest impact. properties of the plumage are destroyed, the birds be-
Of the 181 birds retrieved from the beaches, 15 come waterlogged and sink (Clark, 1968). After the
specimens were examined internally. Results from the Argo Merchant went aground, several thousand surface
necropsies are summarized in Table 2. Pathology and seabed drifters were released at various positions
findings indicated three conditions common to all near the tanker (U.S. Dept. of Comm./NOAA, 1977). No
specimens: surface drift cards were found and only 1 seabed drifter
(1) all were underweight as determined by expo- was recovered on shore. Because of adverse tidal cur-
sure of the keel rents and westerly prevailing winds at that time, the
(2) all lacked a layer of body fat probability of birds drifting ashore to Cape Cod and the
(3) none had food in the digestive tract. islands was remote. We suspect that the small number of
The lungs and kidneys were the vital organs most birds that went ashore had made an active effort to get
seriously affected. The lungs in 7 birds had hemorrhaged, there after being oiled.

Table 2. Summary of Pathology Findings.

Number of Specimens with:
Sample Hemmorrhagic Kidney KidneylLung Parasites
Species Size L ungs Pneumonia Blockage Congestion (Chronic)

Common Loon 5 2 1 2 0 0
Great Black-backed Gull 2 2 0 0 0 0
Herring Gull 3 2 0 0 1 0
Common Murre 4 1 0 1 1 1
Thick-billed Murre 1 0 1 0 0 0

146

Of the 181 birds that did come ashore, alcids (49%) dehydration would give similar results. By destroying the
were the most numerous species. Sea observations, water-repellent properties, oil on the plumage could have
however, indicate that gulls were the most affected impaired the birds' ability to swim and/or catch food.
species. There are two reasons for the variation between Since all of the birds were underweight and had no food
the pelagic and beached bird observations. The first is in the digestive tract, it is possible that dehydration could
that few alcids were observed during the cruises in have occurred from a lack of dietary derived water.
January, while unprecedented sightings of alcids, such
as 800 Razorbills, 4000 Thick-billed Murres, and an
additional 1200 unidentified alcids at Race Point,
Provincetown (E. Mass. Bird Observer, 1977), were made
along the Massachusetts coast. Since the alcids were Acknowledgments
close to shore, those affected by the spill had a higher We thank the National Marine Fisheries Service,
probability of coming ashore than did more distant birds. U.S. Coast Guard and University of Rhode Island for
Second, the larger numbers of fulmars and gulls were permitting bird observations aboard their vessels, and to
associated with trawlers fishing on the Northeast Peak J. M. Loughlin, N. Houghton, and C. S. Sharf for their
and Northeast Channel (Figures 2a,b,d). During the counts, assistance with the field work. We gratefully acknow-
concentrations of several thousand individuals were ledge the cooperation provided by R. Andrews, USFWS;
estimated for each of the following species; fulmar, Great H. C. Boyar, NMFS; Lt. H. G. Ketchen, USCG; and Dr. F.
Black-backed Gull, and kittiwake. Counts made in the Heppner, URI, for arranging passage aboard the vessels,
vicinity of foreign fishing activities on the continental and to B. Blodgett, Massachusetts Division of Fisheries
slope south of Massachusetts and Rhode Island and Wildlife, for supplying beached birds for examination.
averaged 30-100+ Great Black-backed and Herring We are indebted to the Manomet Bird Observatory, U.S.
Gulls but 0-10 fulmars per watch. In contrast, near the Fish and Wildlife Service, Office of Biological Services
grounded tanker and oil'slick track, 0-30 fulmars or gulls (USFWS Contract No. 14-16-0005-6057), and Boston
were recorded per watch. University Biology Department for support. We also thank
The percentage of visibly oiled birds was often Raymond A. Paynter and Edythe L. P. Anthony for their
greater near the tanker or oil slick track than in sur- comments on the manuscript.
rounding waters. From 3 to 30 percent of fulmars
observed on Nantucket Shoals, Great South Channel,
and western Georges Bank were visibly oiled. No oiled
fulmars were recorded in the other areas surveyed. References
Percentages of oiled gulls were greatest on Nantucket
Shoals, Great South Channel, and western Georges Bird Observer of Eastern Massachusetts. 1977. Bird observer
Bank, but oiled gulls were observed in other areas. The summary. 5(l): 27 and 5(2): 52-53.
tendency of large gulls and kittiwakes to follow ships Bourne, W. R. P. 1968. Observation of an encounter between
birds and floating oil. Nature 219(154):632.
probably caused the wide dispersal of oiled individuals, Bourne, W. R. P., J. D. Parrack, and G. R. Potts. 1967. Birds killed
thus reducing the probability of their coming ashore when in the Torrey Canyon disaster. Nature (London) 215:
sick or dying. 1123-1125.
Beached bird surveys provide the only specimens Brown, R. G. B., D. 1. Gillespie, A. R. Lock, P. A. Pearce, and G. H.
which can be examined to determine the manner in which Watson. 1973. Bird mortality from oil slicks off eastern
the birds came in contact with the oil as well as the eff ects Canada, February-April 1970. Canadian Field-Nat., 87:
of oil on the birds. Loons and alcids were generally oiled 225-234.
on the belly, sides, and lower back (Cardoza, 1977a) Brown, R. G. B., D. N. Nettleship, P. Germain, C. E. Tull, and T.
indicating that the birds came in contact with the oil whe@ Davis. 1975. Atlas of eastern Canadian seabirds.
Canadian Wildl. Serv., Information Canada, Ottawa, KIA
resting on the water. Gulls were also oiled on the breast OS9. 220p.
(Cardoza, 1977a) indicating that they may become oiled Cardoza, J. E. 1977a. Oiled bird recovery program. Memo dated
while feeding. Diving birds, such as loons and auks, are 3 January 1977 to Director. Mass. Div. of Fish. and Wildl.
the species most vulnerable to oil spills because of the files, Boston, Mass.
amount of time they spend resting on the water (Bourne Cardoza, J. E. 1977b. Update on oiled bird collection. Memo
et al., 1967; Hope-Jones et al., 1970; Greenwood et al., dated 26 January 1977 to Director. Mass. Div. of Fish. and
1971; Brown et al., 1975). Wildl. files, Boston, Mass.
As stated previously, the lungs and the kidneyswere Clark, R. B. 1968. Oil pollution and the conservation of seabirds.
the organs most seriously affected. The most likely route Internat. Conf. Oil Pollut. Sea. 5:76-112.
Erickson, R. C. 1962. Effects of oil pollution of migratory birds.
for oil entering the lungs is through the glottis, either by Trans. Seminar Biol. Problems Water Pollut., 5:177-181.
leakage or through inhalation. If, for example, the bird Greenwood, J. J. D., R. J. Donally, C. J. Feare, N. D. Gordon, and
chokes or coughs while preening oiled feathers, the oil G. Waterston. 1971. A massive wreck of oiled birds;
could enter the trachea. Once oil is in the lungs, it northeast Britain, winter 1970. Scot. Birds. 6:235-250.
becomes lodged in the parabronchi where it is sur- Hartung, R. 1963. Ingestion of oil by waterfowl. Pap. Mich. Acad.
rounded by erythrocytes, lymphocytes, and neutrophils. Sci., Arts, Letters. 48:49-55.
Blockage of some parabronchi probably causes an Hartung, R. 1964. Some eff ects of oil on waterfowl. (Ph.D. Thesis,
increase in blood pressure in other parabronchi and the Univ.. Michigan, Ann Arbor.) Diss. Abs., 25:6866.
combination of restricted circulatory flow and increased Hartung, R. 1965. Some effects of oiling on reproduction in
ducks. J. Wildl. Manage., 29:872-874.
blood pressure in the lungs results in hemorrhaging Hartung, R. 1967. Energy metabolism in oil-covered ducks. J.
which can kilt the bird. Wildl. Manage., 31:798-804.
The kidney pathology matches the description of oil- Hartung, R., and G. S. Hunt. 1966. Toxicity of some oils to water-
induced changes from Hartung and Hart (1966), but fowl. J. Wildl. Manage., 30:564-569.

147

awkes, A. L. 1961. A review of the nature and extent of damage
Caused by oil pollution at sea. Trans. N.A, Wildl and Nat.
Res, Conf, 26:343-355.
Hope-Jones, P., G, Howells, E.I.S. Rees, and J. Wilson. 1970.
Effect of Hamilton Trader oil on birds in the Irish Sea in
May 1969. Brit Birds. 63-97-110.
Taning, A. V. 1952. Oliedoden. Sveriges Nat_ 5: 114-122.
U.S. Dept. of Commerce/NOAA. 1977. The Argo Merchant oil
spill - a preliminary scientific report. Ed. by PL. Grose and
J.S. Mattson. NOAA, Environmental Res. Lab., Boulder,
Colo. NOAA Spec. Rept., 133p. + tables.

148

The Argo Merchant Oil Spill and
the Fisheries

Kenneth Sherman and Donna Busch

National Marine Fisheries Service
Northeast Fisheries Center
Narragansett Laboratory
Narragansett, Rhode Island

Abstract shore and in a general southeasterly direction off the
The impact of oil spilled from the Argo Merchant on Continental Shelf into slope and oceanic water. The
impact of theArgo oil on the fisheries during its relatively
fish stocks has not been catastrophic. No evidence of limited residence time is the subject of the present report.
large scale mortalities of juvenile or adult fish has been To date no comprehensive study has been carried
observed in the 12 months (Jan.-Dec. '77) following the out on the effects of oil on the productivity of fish popula-
spill. There has been, however, evidence of oil contamina- tions on the northeast Continental Shelf. In fact, with the
tion in fish, shellfish, and zooplankton populations in the exception of the Torrey Canyon Spill, the Santa Barbara,
area of the spill. Mortalities of developing cod and pollock and more recently, the EKOFISK spills, most studies on
embryos contaminated with oil were observed. A few the effects of oil on fish and shellfish have been con-
adult fish (<5% of those examined) had apparently cerned with the onshore or near-shore impacts on littoral
ingested Argo-like oil. The dominant zooplankton species organisms (Sanders, 1977). Sublethal effects of crude oil
in the area of the spill was the copepod Centropages have been described for saithe, Pollachius virens
typicus, which is an important food of larval and adult fish. (pollock), in the immediate vicinity of a grounded tanker
Large numbers of this species were contaminated with off the Norwegian coast (Grahl-Nielsen et al., 1976). In
petroleum hydrocarbons similar to No. 6 fuel oil, indi- experiments where salmon and saithe were exposed to a
cating that an important pathway in the food web of the maximum of 50 p gms EKOFISK crude oil/liter seawater,
Nantucket Shoals-Georges Bank ecosystems was im- both species showed residues of petroleum hydro-
pacted. The long-term effects of the Argo oil on the carbons within 7 hrs of dosing. Following termination of
marine ecosystem will be difficult to assess. A more dosing after 68 days, both species showed naphthaline
significant problem concerns the chronic background levels comparable to those existing prior to dosing,
levels of petroleum hydrocarbons present in the surface indicating that at moderate dosage levels, effects are
waters inhabited by fish eggs and larvae. sublethal and reversible (Brandal et al., 1976).
Laboratory studies have shown that crude oil can
Introduction damage developing fish eggs and cause high mortalities
in cod, herring and capelin embryos (Kuhnhold, 1969,
The tankerArgo Merchant ran aground on Nantucket 1974; Johannessen, 1976). Also, the zooplankton food of
Shoals 15 December 1976. By 8 February 1977 approxi- fish larvae suffer high mortalities from exposures to crude
mately 7.7 million gallons of No. 6 fuel oil had been oil in laboratory experiments (Mironov, 1969). Recent
released into the waters of the Continental Shelf. In the studies in large scale microcosms at the University of
immediate vicinity of the wreck, concentrations of petro- Rhode Island have also revealed that No. 2 fuel oil at
leum hydrocarbons up to 250 ppb were detected. Heavy levels of 100-150 ppb has dramatic negative effects on
winter winds and seas contributed to the fragmentation population numbers and physiological responses of
of large "pancakes" viewed in the immediate vicinity of copepods (S. Vargo, personal communication). In con-
the Argo following the breakup of the vessel on 21 trast, observations from collections made at sea have
December. Available evidence from Coast Guard over- shown that zooplankters, particularly copepods, can
flights of the spill zone indicated that within 40 days the oil ingest particles of oil and pass them through the gut with-
was carried in small "pancakes" and "streaks" under the out any observable negative effects (Conover, 1971,
influence of prevailing winds and currents away from Parker, 1970). Some species of adult fish have been

149

observed to avoid areas contaminated with oil. However, Methods
the more sensitive egg and larval stages are carried by
the tides and currents and lack the ability to avoid oil spill Several data bases were used to evaluate the
areas. Bivalve shellfish (quahogs, scallops, mussels) are impact of Argo oil on the fisheries in the region of the spill
sedentary and have only limited capability to remove including fish catches, zooplankton, ichthyoplankton, and
large amounts of petroleum hydrocarbons. They suffer neuston. Monthly catch data from commercial landings
significant mortalities in areas contaminated with oil were compiled for the area in the vicinity of the wreck for
(Blumeret al., 1970; Thomas, 1973; Jeffries and Johnson, 1975, 1976, and 1977. To allow for changes in the distri-
1975). Proper assessments of the impact of a major spill bution of the stocks, catches from a wider area encom-
on the Continental Shelf require the combined effort of passing both Nantucket Shoals and Georges Bank were
extensive sea sampling and experimental support analyzed (Figure 1). Fisheries independent abundance
studies. information was obtained from the results of the spring
and autumn bottom trawl surveys conducted as part of the
Marine Resources Monitoring Assessment and Predic-
tion Program (MARMAP) of the National Marine Fisheries
Service. The surveys have been made continuously off
the northeast coast for the past 15 years by the Northeast
Fisheries Center (Grosslein, 1976) and represent a
valuable time-series of comparable data (Figure 2).
72* 711* Zooplankton and ichthyoplankton samples were
collected with paired 60-cm bongo nets towed obliquely
through the water column from just off bottom to the
surface at 1.5 kts. Also, surface plankton was collected
with a 0.5 rn x 1.0 m neuston net towed at the surface at
1.5 kts. The dates and areas of collection are given in
40' Table 1.
Figure 1. Statistical areas examined for U.S. commercial catch Information on the presence of Argo oil over the
data 1975, 1976, 1977. The small rectangle represents fishing fishing grounds was obtained directly from the fishermen
grounds adjacent to the Argo v@reck. The large rectangle repre- in two activities. The first was from an initial survey of
sents ICNAF statistical subareas Ve and 5Zw The Argo wreck fishermen in each of the principal ports of the northeast
site is marked with an X coast. Fishermen were interviewed at dockside by NMFS

Offshm Sx%qs ( 1973/ 74 13mbried
CUMLLATmEAMA
40F ffo-IT701V FLOTI vdm sur&Y 1974 oriy

0*0 % 10 OV 46
6%
0 %
%
Dow,
Ir J6
%
0

% %
SO %
dp 0. r
s

400 0
SFRM

Figure 2. Bottom trawl stations occupied in spring 19 73 and 19 74 offshore surveys and spring 1974 inshore survey (from Grosslein, 1976).

150

Table 1. Dates and Areas of Sample Collections.

Vessel Date Location of Sample Area No. Samples Examined

Delaware 22-24 Dec. 1976 Argo area, Nantucket Shoals* 9
NMFS, NOAA

Delaware //, 4-10 Jan. 1977 Nantucket Shoals, Georges Bank, Race Point* 43
NMFS, NOAA

Mt. Mitchell, 12-26 Feb. 1977 Nantucket Shoals, Georges Bank 38
NOAA

Endeavor, URI 22-27 Feb. 1977 Nantucket Shoals- 20

Wieczno, 27 Feb.-6 Mar. 1977 Nantucket Shoals, Georges Bank* 41
Poland

Gorlitz, GDR 3-15 Mar.; 15 Mar.- Southern New England, Nantucket Shoals, 22
3 Apr. 1977 Georges Bank

Albatross IV, 13 Apr.-14 May 1977 Nantucket Shoals, Georges Bank* 18
NMFS, NOAA

Nogliki, USSR 22 May-6 Jun, 1977 Southern New England, Nantucket Shoals, 34
Georges Bank

Yubileiny, 31 Jul.-15 Aug.; Southern New England, Nantucket Shoals, 33
USSR 17 Aug.-3 Sept. 1977 Georges Bank

Wieczno, 4-24 Oct. 1977 Nantucket Shoals, Georges Bank 19
Poland

Argus, USSR 15 Oct.-10 Nov. 1977 Southern New England, Nantucket Shoals, 32
Georges Bank

* Cruise on which fish samples were collected.
* Petroleum hydrocarbon analysis of zooplankton samples from this cruise.

personnel. In addition, observations of 5 groups repre- buoys caused a deterioration of air valves resulting in a
senting several hundred vessels, captains, and deck loss of these buoys and consequently the gear they
hands were collated and analyzed by employees of marked. The crew's clothing became fouled during hand-
Development Sciences Inc. of Barnstable, Mass. The ling the gear. One lobster fishing vessel had its gear in the
results of this survey were used in the analysis. immediate area of the oil drift and, as a result, had to
change over the water circulation system from a con-
The Fisheries of Nantucket Shoals and Adjacent Waters. tinuous to a closed one, i.e., instead of taking in water
NEFC Port Agents attempted to assess the impact of the from the area of the oil drift and contaminating the catch,
Argo Merchant oil spill on the daily activities of the com- water from a clean area was used and circulated within
mercial fishermen. In New England approximately 900 the vessel's holding system. No reports of adverse
direct interviews were made from a total of 4,000 fishing effects of Argo oil were made by foreign fishing vessels.
trips between 21 December and 30 January, at the ports In a draft report (Development Sciences, 1977), a
of Portland, Rockland, Gloucester, Boston, and NOW Bed- consortium of 5 non-profit fisheries organizations initi-
ford, Massachusetts; and Newport and Pt. Judith, Rhode ated and carried out a program aimed at determining the
Island. Of 26 interviews conducted where evidence of the impact of the oil on the fish of the Nantucket area. The 5
oil spill was noted during that period, 5 (20%) indicated groups represent several hundred vessels, captains and
direct loss of catch, fouling or loss of gear. The other 21 deck hands. The activity was supported by the Office of
(80%) reported "oily" birds. All ofthese incidents occurred Technology Assessment and coordinated by Develop-
in the area to the southeast of the site of the Argo Mer- ment Sciences of Barnstable, Mass. The fishing fleet
chant at locations associated with fishing operations. reported seeing oil slicks associated with Argo Merchant
Specific problems were loss of gear, fouling, or loss until March 30 with major concentrations observed before
of catch. A scalloper fishing very near the wreck area had 10 February. The most significant impact reported by
his catch and gear fouled by an oil slick; the catch from fishermen was the hundreds of oiled birds observed
that towwas discarded as unmarketable. Captains oftwo during the period from January through April. They found
vessels, fishing American lobster on the edge of the no evidence of oil on the bottom. The only oil damage
Continental Shelf, believe that oil fouling of inflatable reported was limited to the observation of oil in the

151

Table 2. Abundant Fish Species in the Nantucket Shoals - Georges Bank Area.a

Spill Zone Total Area
Metric Tons Metric Tons
Apr- Jan- Jan- Jan-
Dec. Dec. Dec. Dec.
1975 1976 1975 1976

Sea scallops 2086 5753 Yellowtail flounder 16,986 14,515

Cod 5260 5742 Cod 15,227 14,213

Winter flounder 1932 1547 Sea scallops 7,556 14,725

Yellowtall flounder 1203 1333 Menhaden 7,168 5,718

Pollock 333 641 Silver hake 6,584 6,407

Haddock 369 416 Winter flounder 5R55 4,601

Silver hake 282 175 Sea herring 4,041 541

Windowpane 239 358 Haddock 3986 2,894

Redfish 209 119 Redfish 3,113 2,143

Summer flounder 175 314 Pollock 2,979 3,843

American plaice 29 58 Summer flounder 1,943 3,418

aThese species represent 94-99% of the total catch in each of the two Vears.

stomach of two codfish shortly after the spill. tions. In combination, these events have reduced the fish
We examined commercial catch statistics from biomass substantially from former abundance levels in
statistical areas adjacent to the wreck site. The catch the 1950's and early 1960's (Clark and Brown, 1977).
data from the spill zone came from an area of approxi- Against this dramatic change in biomass, we are attempt-
mately 3,600 square miles. Recognizing that fish undergo ing to sort out the impact of Argo oil on the stocks in the
migration for feeding and spawning, we examined catch vicinity of the spill.
statistics of a larger area. The broader area examined We have seen no dramatic changes in the stocks of
covered approximately 54,000 square miles (Figure 1). yellowtail flounder, cod, sea scallops, or winter flounder
We examined the catches for two years (1975, from 1975 through 1977. The catches of the four species
1976) prior to the spill to identify the most abundant for spring and fall of 75,76, and'77 are plotted in Figure
species in the spill area. The three most frequently 5a-h. It should be noted that this is preliminary data and
caught species are: cod, sea scallops and winter flounder the variability inherent in the data has not been analyzed,
(Table 2). A summary of monthly commercial catch data but no downward trends in abundance were observed.
for the abundant species is given for both the spill area Although these tour species were abundant around the
and the Nantucket Shoals-Georges Bank area in Figures wreck site, they are widely distributed over the survey
3 and 4. Catches of cod and scallops are highest in 1977. area, with concentrations on the northeast part of
Winter flounder catches are similar in each of the three Georges Bank; the wide distribution of the species
years. No downward trends in abundance were evident enhances the probability of survival.
(Figure 3). In the total area examined, cod, scallops, and lchthyoplankton. At the time of the spill, six species of fish
yellowtail flounder were the most frequently caught larvae were in the collections: sand launce, cod, pollock,
species. Cod and scallop catches were higher in 1977 and rockling, hake, and herring. Of these species, only sand
yellowtail flounder landings were similar among the three launce was abundant (Tables 3a and b). Other larvae
years (Figure 4). These data are unweighted catches, and were rare.
are used here as only an indication of catch trends in the The abundance of sand launce larvae decreased
area. No allowance has been made for changes among sharply at the two sampling stations within the spill area
the three years in fishing effort. and showed an increase at the periphery of the oiled
Bottom Trawl Survey. The semi-annual bottom trawl sur- waters in the bongo tows, but decreased in the neuston
veys conducted by NEFC monitor changes in abundance samples (Figure 6). However, the reasons for the de-
of the principal fish stocks in the area. Major changes in crease are not clear. It may have been associated with the
the abundance of the stocks prior to the Argo Merchant negative impact of the oil on the viability of larvae, or it
spill were the result of the interaction between intensive may have been a reflection of the "patchiness" charac-
fishing and naturally occurring environmental fluctua- teristic of larval fish distributions. The sand launce, while

152

1500 3429 3608
SEA SCALLOPS - SEA SCALLOPS

3200 -

1000
2400

500 1600

Boo
0 7-7j
1500- COD" 0
3200
- COD
U)
z
0 1000 z 2400
0

2
500 1600
ix
W
LLJ 800
0 2
600 .0 T I I I I I T-T-I I I
WINTER FLOUNDER 3200
- YELLOWTAIL

400 -
01975 2400
- 61976
200 - 01977 1600

0 101975
1 1 800 61976
J F M A M j J A S 0 N D 101977
MONTH
0-
J F M A M J J A S 0 N D
MONTH
Figure 3. Monthly U.S. commercial catch of winter flounder, cod, Figure 4. Monthly U.S. commercial catch of yellowtail flounder,
and sea scallops in the Argo Merchant spill area for 1,975, 1976, cod, andsea scallops in the NantucketShoals-Georges Bankarea
1977, for 1975, 1976, 1977.

not important in the commercial fishery, is a key species had a negative impact on new recruits to cod, pollock, and
in the ecosystem. It is the basic food of predatory fish scallop populations of the Nantucket Shoals, area
including cod, haddock, silver hake, as well as marine remains an open question. Additional ichthyoplankton,
mammals including porpoises and whales. Data from prior bottom trawl and shellfish surveys scheduled for 1978
years ('73, '74, '75, '76) has shown abundances of sand and 1979 by NMFS should provide data sufficient for
launce in the southern New England-Mid Atlantic Bight making a more complete evaluation of any long-term
areas increasing every year since 1973. Preliminary effects.
analysis of the 1977 data suggests this trend is
continuing. Zooplankton. The communities of zooplankton in the
There is evidence that Argo oil caused. cytogenetic vicinity of the oil spill were described in the NOAA (1977)
mortalities to fish eggs. Mortality among pollock eggs report "The Argo Merchant Oil Spill." Significant con-
increased significantly in the areas of the spill and some tamination was found in copepods subsequent to the
@OD

mortality of cod eggs was evident (Longwell, 1977). Also, spill. The dominant copepod in the area of the spill was
laboratory experiments have shown that Argo-type oil is Centropages typicus, a food of both larval and adult fish.
toxic to developing cod eggs and can cause high levels of Large numbers of C typicus were contaminated with
embryo mortality (Kuhnhold, 1977). WhetherArgo oil has hydrocarbons similar to No. 6 fuel oil indicating that an

153

YELLOWTAIL FLOUNDER YELLOWTAIL FLOUNDER

14 - 14 11 W, - Iq @ 14 1% 1
SPRING '75 FALL '75

XG n

ILI,

lit
SPRING '76 FALL '76

30
0 *

no 4E

0'v7- FALL '77
SPRING ITT
<_ @_

33
lei "a

,\1s.)atross 11V 77-02
71 0- Delaware 11 77-12
ZI,

Figure 5a. NEFC bottom trawl catches of yellowtail flounder for Figure 5b. NEFC bottom trawl catches of yellowtail flounder for
spring 1975,1976,1977. Large circle, 10-20 kilos, medium circle, fall 1975, 1976, 1977. Large circle, 10-20 kilos, medium circle,
5-10 kilos, small circle, <5 kilos. 5-10 kilos; small circle, <5 kilos.

154

C 0 D C 0 D

j;
SPRING T5 FALL '75

F1 A 14 14

FALL 'T6
SPRING '76

0 'M

M 14 1; 14 11 4

FALL 'TT

SPRING '77

Zi@

-bS

:402 Delware 11 77-12
A4 1@-
77-02
Albatross IV.
-7

Figure 5c. NEFC bottom trawl catches of cod for spring 1975, Figure 5d. NEFCbottom trawl catches ofcod forfall 1975,1976,
1976, 1977. Large circle, 50-400 kilos, medium circle, 20-50 1977. Large circle, 50-300 kilos, mediumcircle, 20-50 Mos,-small
kilos, small circle, <20 kilos. circle, <20 kilos.

155

SEA SCALLOP SEA SCALLOP

FALL 'T
SPRING '76 5

30 30

Mo so

FALL '76
SPRING '77
M
105 Z-0 Is, 0-
56! n

41,

Albatross IV 77-02 14 14 it 11 1;
14 1; I't 1; 1!-

LL '77
Figure 5e. NEFC bottom trawl catches of sea scallops for spring we FA
40
1976, 1977. Large circle, 100-300 scallops, medium circle,
20-100 scallops; small circle, <20 scallops.

important component of the food web was impacted.
When oil droplets removed from the alimentary tracts of 11113
C. typicus were examined for petroleum hydrocarbons Delaware 11 77-12
using gas chromatography, the resulting chromatograms
were similar to chromatograms of oil from the Argo Mer-
chant (Kuhnhold, personal communication).
Four zooplankton samples from a time-series col-
lected from December 1976 through November 1977
were sent to the NOAA Analytical Laboratory, Seattle, Figure 5f. NEFC bottom trawl catches of sea scallops for fall
Wash., for hydrocarbon analysis. The samples analyzed 1975, 1976, 1977. Large circle, 50-700 scallops, medium circle,
were from three cruises conducted in.January through 20-50 scallops; small circle, <20 scallops.

156

WINTER FLOUNDER WINTER FLOUNDER
(BLACKBACK) (BLACKBACK)

FALL 75
SPRING'76

60
4e

FALL '76
SP ING'@7

:SS

41'

F1
Alhatros
s IV 77-02

FA L L! 7 7
Figure 5g. NEFC bottom trawl catches of winter flounder for
spring 1976,1977. Largecircle, 10-125 kilos; mediumcircle, 5-10
kilos, small circle, <5 kilos.

(4.j>
006

March 1977. No other samples from the time series were
e
sent. Of the samples analyzed in Seattle, one sampl
taken on Endeavor Cruise 005 in February (410 01'N, Wware It 77-12
69031'W), west of the Argo bow section, contained satu- A4 44
rated hydrocarbons which closely resembled theArgo Mer-
chant cargo hydrocarbons, suggesting that the Argo could
have been the source of contamination. Polak (personal
communication) using a U-V f luorescence technique also
found Argo-type oil in zooplankton from the spill site in Figure Sh. NEFC bottom trawl catches of winter flounder for fall
February. The sample from the January cruise taken 1975,1976,1977. Large circle, 10-125 kilos; medium circle, 5-10
southwest (40050'N, 69035'W) of the wreck did not give a kilos; small circle, <5 kilos.

157

Table 3a. Numbers of Fish Eggs and Larvae per 1000 m3 Collected A total of 9% of the fish had oi I in their stomachs, but
in the Area of the Argo Merchant Oil Spill, Dec. 1976. only 4.5% of that resembled Argo Merchant cargo. Four
and a half percent of the samples had detectable amounts
of fuel oil in the flesh, but this did not match well with
Fish Larvael Fish Eggsl Argo Merchant oil, indicating that these fish had been
1000 m3 1000 m3 exposed to other sources of petroleum contamination. A
61 cm 1x5m 61cm. 1x5m summary of the incidence of fuel oil in the fish and shell-
Station Bongo Neuston Bongo Neuston fish samples is given in Table 4.
Number Species .505 .505 .505 .505 MacLeod reported that his results are not consis-
4 Clupea harengus 14 tent since some of the oil residues were found in fish from
potentially contaminated areas and others are from sta-
Ammodytes tions outside of the area overrun by the oil slick. For
americanus 4,062 example, winter and windowpane flounders from St 'atiop,
Gadus morhua 612 182 3, Delaware It Cruise 77-01, both showed oil in stomach
Pollachius and flesh even tho 'ugh they were taken close to shore off
virens 86 5 Nauset Light, Cape Cod, well away from the oil spill. On
the other hand, the only fish specimen observed to have
5 A. americanus 21,197 2,172 oil on it, a winter flounder from Station 4, showed no
traces of oil in the stomach or in the flesh (Figure 7).
G. morhua 1 72 23 To determine if Argo Merchant oil was assimilated
P. virens 95 20 into the flesh over a period of time an additional 24
samples of fish and shellfish were sent to Dr. MacLeod on
Enchelyopus 30 June 1977. The seven f ish species array was the same
cimbrius 1 except no windowpane flounder were included and a
sample of little skate was sent. The invertebrates in-
6 C. harengus 3 cluded hermit crabs, Cancer sp., ocean quahogs and sea
A. americanus 6,475 1,752 scallops. These samples were collected on Wieczno
G. morhua 8 14 77-01, 18 February-6 March 1977, and are p Irese .ntly
being analyzed.
A virens 20 1 8 9
Neuston Series. We examined neuston samples through
Urophycis sp. 1 an annual cycle in the spill area from November 1976
through November 1977 (11 cruises) to determine the
7 A. american.us 2,967 extent of oil contamination. We qualitatively estimated the
G. morhua 7 11 amount of particulate oil and tar in 332 samples.'Of this
P. virens. 72 number, 36% were contaminated (Table 5). This is not sur-
prising in view of previous work conducted by NEFC in the
.8 A. americanus 7,857 41 area. The level of oil and tar contamination in neuston sam-
ples taken between Virginia and Cape Cod ranged be-
G. morhua 30 18 tween,0.05 and 1.04 Mg/M2 in winter of 1973 and between
P. virens 1,439 826 0. 18 and 0.77 Mg/M2 in summer of 1972 (Sherman et al.,
1974). Of an additional 382 neuston samples examined
9 A. americanus 41,184 222 from the MARMAP/U.S. Coast Guard cooperative monthly
neuston sampling program (March 1975-Nov. 1977),48%
G. m.orhua, 31.7 217 showed contamination. The samples were collected from
P. virens 36 188 the area bounded by 680-73OW and 39'30'-40030'N. The
locations of samples containing oil and/or tar from the
Argo spill area are given in Figure 8. The effect of oil
residues on neuston is not clear. However, it has been
established that oil can be toxic to clupeid larvae
high correlation with Argo oil. In the third and fourth' (Kuhnhold, 1977). The larvae of Ammodytes and other
samples from a subsequent cruise located east and east- clupeids continue to be exposed to petroleum hydro-
southeast of the wreck (40054'N, 69'14'W and 41003'N, carbons as important constituents of the neuston.
68'59'W) in later February-March, no oil was found.
Hydrocarbon Analysis of Fish and. Invertebrates. On Long-Term Assessment The grounding and breaking of
February 1, 1977, a total of 66 specimens in 22 samples the Argo Merchant and subsequent groundings of other
including. 7 fish species (Cod, haddock, red and silver oil tankers on the Continental Shelf are dramatically
hake, windowpane, winter and yellowtail flounder) and 3 illustrative of specific events that are not predictable. For
invertebrate. species (sea scallop, lobster and sand example, the Northeast Fisheries Center is frequently
dollar) which were collected on Delaware 11 76-13 and requested by responsible officials-- local, state, and
77-01 cruises were sent to Dr. William MacLeod at the Federal--to assess the impact of major environmental
NOAA National Analytical Facility, Seattle, Washington, incidents on the fishery resources of the northeast Con-
for hydrocarbon analysis. Three individual fish or inverte- tinental Shelf. To deal with these incidents special
brates of the same species were pooled to constitute a studies are initiated to assess the impact on the environ-
single sample. Stomachs suspected of having oil content ment and living resources. These efforts, however, are of
were analyzed separately. limited duration, and conducted with little information on

158

Table 3b. Abundance of Fish Larvae per Sample Collected in the Area
of the Argo Merchant Oil Spill, 4-10 January 1977.

Species
ca U
Z 42
ri
M CZ
Z ri 4,--- -r- .I I
q)
Station co
r U
4 Lu !@

1 4,088 1 1 0 0 0 0 1 0
2 736 0 9 0 0 0 0 0 0
3 236 0 0 0 2 0 0 4 0
4 116 0 18 0 0 0 0 1 0
5 409 0 30 0 0 2 0 42 0
6 1,372 0 10 1 1 1 0 a 0
7 4,316 0 14 0 0 2 0 2 0
8 434 0 3 0 0 0 0 0 0
9 59 0 0 0 0 0 0 0 0
10 8 0 0 0 0 0 0 0 0
11 11 0 0 0 0 0 0 0 0
12 314 0 4 1 0 0 1 0 0
13 302 0 0 4 0 0 0 0 0
*14 1,333 0 0 0 0 0 0 3 0
15 900 0 1 1 0 2 0 0 0
16 306 0 6 1 0 3 0 0 0
17 348 0 4 0 0 0 0 0 0
18 4 0 0 0 0 0 0 0 0
*19 165 0 0 1 0 0 0 0 0
*20 3 0 0 1 0 0 0 0 0
*21 3 0 0 0 0 0 0 0 0
*22 0 0 0 0 0 0 0 0 0
*23 1 0 0 0 0 0 0 0 0
*24 52 0 0 0 0 0 0 0 0
*25 49 0 0 0 0 0 0 0 0
*26 550 0 1 0 0 0 0 0 0
*27 1,499 0 3 3 0 0 0 4 0
*28 854 0 8 6 0 0 0 0 0
29 6,093 0 1 0 0 0 0 0 0
*30 3,720 0 0 0 0 0 0 8 0
*31 37,952 0 0 0 0 0 0 0 0
*32 16,720 0 5 1 0 0 0 0 0
33 2,165 0 8 0 0 0 0 5 1
*34 1,420 .0 23 0 0 0 0 0 0
35 286 0 2 0 0 0 0 0 0
*36 5 0 0 0 0 0 0 0 0
*37 0 0 0 0 0 0 0 0 0
*38 159 0 0 0 0 0 0 0 0
*39 62 0 0 0 0 0 0 0 0

NOTE: Values in the table represent combined totals of larvae from surface and water column samples. They are not standardized for
volume of water strained.
*Station was within oil slick border as of 3 January 1977.

the initial physiological condition of the living resources. monitoring of ocean environments and populations, and
Only limited information on the baseline conditions or (3) provides suitable information for interim or near-term
health of the stocks is available. We are dealing with a policy guidance and decision making. The integrated
complex system that requires a combination of s *hort-term approach shbuld couple in-depth "process oriented"
tactical observations that can be evaluated against a studies at specific sites with long-term monitoring of the
background of long-term baseline information on the productivity of fish stocks. The Northeast Fisheries
condition and health of fish and shellfish stocks. Center is attempting to bring together these investigative
In order to effectively deal with these problems an elements in a program called Ocean Pulse. Ongoing sur-
approach is needed that: (1) encompasses the coordina- veys of fish, plankton, benthos, and hydrography off the
tion of the short-term assessment studies of various northeast coast of the U.S. will be augmented by deve-
groups and agencies conducted in response to a pollu- loping a physiological baseline at 20 sites on the Con-
tion "event," (2) provides for regionally oriented long-term tinental Shelf to be systematically monitored during the

159

61 cm. 8ONVO .505 mm. next several years (Figure 9). The Argo spill area repre-
1112 X I M. NEVSWN.505mm
EE OIL SPILL AREA sents one of these sites. The principal scientists respon-
sible for developing the diagnostic tests have been
SAND LAUNCE designated. During the next six months they will continue
AMMODYTES AMERICANUS to examine physiological conditions among selected
samples of key populations collected from the spill area.
MARMAP surveys will continue to provide information on
..........................
4000- changes in the abundance and distribution of stocks and
their environments. The Ocean Pulse Program will moni-
tor physiological conditions to detect sublethal or chronic
impacts of contaminants on the growth, survival, and
productivity of important marine populations. Special
attention will be focused on documenting effects that are
ro 3000-
not readily apparent without biochemical, genetic,
and pathological examination at the tissue level.
0
0 Data collected from clean and contaminated areas
will represent baselines against which the impact of oil
U) and other contaminants can be evaluated.

.............. ....... ..
U) 2000-
Z Summary
From the evidence examined, we can conclude that:
0 (1) The fishery in the vicinity of Nantucket Shoals for
the more abundant species (cod, yellowtail flounder,
................... ....
C;
Z 1000- winter flounder and scallops) after the movement of the oil
offshore was not adversely affected by the spill. Both the
commercial catch statistics and the fisheries indepen-
dent bottom trawl survey results revealed no downward
trends in abundance. Over the broader area encompas-
sing Georges Bank, the results are similar.
01 (2) Of the fish and shellfish analyzed for petroleum
5 6 hydrocarbon content, <5% showed any suggestion of
4 7 8 9 Argo oil. The suspected oil contamination was limited to
STATION NUMBER the stomachs of 2 cod and 1 windowpane flounder sug-
gesting the oil was ingested.
(3) No adverse effects on fisheries were reported by
69*30 fishermen. However, it is important to recognize that
these observations were limited largely to the less sensi-
41*30' tive adult fish, and to a lesser degree tojuvenile segments
DELAWARE 11 - CRUISE DE 76-13 of the populations that are targeted by the fishery.
ARGO MERCHANT OIL SPILL (4) The sensitive early developmental stages of fish
DEC. 22-24, 1976 were impacted, as evidenced by the observed mortalities
of cod and pollock embryos. However, important environ-
mental factors mitigated the adverse effects on these
species: (a) the residence time of the oil over the spawn-
ing grounds was minimized by prevailing offshore winds
and currents, and (b) the lack of significant amounts of oil
detected on the bottom. While some change in the abun-
dance of Ammodytes larvae was observed, the species is
IV wide-spread, occurring naturally in "patches." Additional
analysis is required to evaluate the level of population
41*00' change against the expected natural variability in
9 distribution.
(5) Evidence was found of oil entering the food web
10 of Nantucket Shoals. The dominant zooplankter was the
4 7 copepod C. typicus. This species ingested oil particles
5 although no harmful effects on zooplankton have been
reported in the literature.
6 (6) Although the short-term effects of the Argo spill
were not catastrophic, a proper evaluation of the impact
of petroleum hydrocarbons on Continental Shelf popula-
69*130' 69*00' tions has yet to be carried out. The Argo spill, while
1@6.
spectacular, represents < 1 % of the total annual input of
Figure 6. Numbers of sand launce per 100 m' collected on petroleum hydrocarbons in the World Ocean (NAS 1975).
Delaware 11 cruise 76-13 at each of the sampling stations. The long-term chronic problem requires far more atten-
Stippling designates stations with oil visible in surface waters. tion than is presently being directed to its solution.

160

Table 4. Summary of Hydrocarbon Analyses.

No. Samplesb, No. Samples
Species Cruise Stationa Analyzed Contaminated Remarks

Fish
Cod (Gadus morhua) DE 77-01 3 1 0
38 1 1 High degree of correspondence
with Argo Merchant arkenes in
stomach but two orders of
magnitude less oil than in spec-
ial sample from Station 29.
Haddock (Melanogrammus 8 1 0
aeglefinus)
27 1 0
Silver hake (Merluccius 3 0
bilinearis)
24 1 Fairly heavy concentration of
heavy fuel oil in stomach.
Red hake (Urophycis chuss) 10 0
24 0
Yellowtail flounder 3 0
(Limanda ferruginea)
31 0
Winter flounder 3 1, Moderate concentration of oil
(Pseudopleuronectes in stomach. None in the flesh.
americanus)
31 1 Heavy concentration of oil in
stomach. Lesser concentration
in flesh.
Windowpane flounder 3 1 Light concentration of oil in
(Scophthalmus the stomach; similar correspon-
aquosus) dence with Argo Merchant or
other similar type of fuel oil;
no oil in flesh.
DE 76-13 4 1 0
Invertebrates
Sea scallop DE 77-01 3 1 0
(Placopecten
magellanicus) 39 1 0
Lobster (Homarus DE 76-13 6 1 0
americanus)
DE 77-01 6 1 1 Slight concentration of oil in
flesh.
Sand dollar DE 77-01 19 1 0
(Echinarachinius
parma) 34 1 0
Special Samples
Contents of cod stomach 29 1 1 Large amount of oil in stom-
containing oil (individual) ach; fairly high degree of cor-
respondence with Argo Mer-
chant hydrocarbons.
Winter flounder with DE 76-13 4 1 1 No traces of oil in stomach or
external smudge of oil (individual) f lesh.

aFigure 7.
bThree individuals were pooled to make up one sample.

161

70* 68* 66*

DELAWARE 11 CRUISES ...............
42*
42 a 76-13, DEC. 22-23,1976
1.3
. TT-01, JAN. 4 -10,1977
--- OIL SLICK BORDER

10 12 16. 17 0
43 42
41 .0 31
27 - ----- 20
'00 29 100 - @ @ W'
I1 '07 24.
2 0 '21
34@

40* 40.

70* 68 66*
Figure 7. Station locations of Delaware 76-13 and 77-01.

Table 5. Presence of Oil and Tar in Neuston Samples.

No. Samples No. % No. Heavily Location of Heavy
Cruise Date Examined Contaminated Contaminated- Contaminated Contamination

Researcher 26 Nov.-12 Dec. 32 2 6 0
76-01 1976

Delaware 22-24 5 5 100 5 South and Southeast of Argo
76-13 December, 1976 Merchant wreck

Delaware /1 4-10 38 11 29 4 South and Southeast of Argo
77-01 January, 1977 Merchant wreck

Mt. M 'itchell 12-26 38 21 55 2 Southwest of Argo Merchant
77-02 February, 1977 wreck

Endeavor 22-27 Feb. 20 7 35 2 South and Northwest of
005 Argo Merchant wreck

Wieczno 27 Feb.-6 Mar. 41 18 44 4 South and Southeast of Argo
77-01 Merchant wreck

Gorlitz 3-15 Mar.; 22 2 9 0
77-01 15 Mar.-3 Apr.

Albatross V 13 Apr, 18 8 44 3 West and Southwest of Argo
77-02 14 May Merchant wreck

Nogliki 22 May-6 June 34 15 44 2 Northwest of Argo Merchant
77-02 wreck

Yubileiny 31 Jul.-15 Aug.; 33 15 46 5 Three Southwest of the
77-02 17 Aug.-3 Sept. wreck - One East of the
wreck and One Northeast of
the wreck

Wieczno 4-24 Oct. 19 11 58 2 Southeast of the wreck
77-06

Argus 15 Oct.-10 Nov. 32 5 16 0
77-01

Total 332 120 36 29

162

70' 68* 700 680
a I

DELAWARE 11 76-13 MT MITCHELL 77-02
22-24 DEC 1976
.. .....- 12-26 FES 1977 ...............
42* 420

................ .......... . .....
. .. ..... .... A0o
40' 40*

DELAWARE 31 ENDEAVOR EN-005
4-10 JAN 1977 22-27 FES' 1977

42*
42*

........ ..
............
....................... 100
40* 40*

70* 68* 70* 68*

Figure 8a. Locations ofneuston stations in the Argo area where samples contained oil or tar, December
1976-February 1977.
TO* 68. 70* 68*
6
WIECZNO 77-01 ALBATROSS IV 77-0 42*
27 FEB -6 MAR 1977 13 APR - 14.,MAY-- -197.7
...................

..........
40'
.100"", .............. ......... . ........... ..............
....................... ...............

GORLITZ 77 NOGLIKI 77-02
3-15 MAR; 15 22 MAY - 6 JUN ..... !A7T
3 APR 197-:r-,
42*

................. ...................

...........
............ .................
40*
2

-01
77-01

...............

Figure 8b. Locations of neuston stations in the Argo area where samples contained oil or tar, February
1977-June 1977.

163

YUBILEINY 77-02 ARGUS 77 - Of
31 JUL - 15 AUG
15 OCT - 10 NOV 197 7
..................
..... ...........
17 AUG.7---3"'9ff0"-"�' 42', 42* 42*

..............
....100 ...... .................0 ...100- .. ................. ..........
40' 40* 40*

70' 66*

-06
T7
WltczNO
4-24 OCT 19T7
................
42* 42*

.................

40* - 40*

Figure 8c. Locations of neuston stations in the Argo area where samples contained oil or tar, July 19 77-November 1977.

INTEGRATED PULSE/MARMAP SAMPLING

.1C .*
xz

4;
.0.

MARMAP SURVEYS 1 11
S
P
RING & FALL

STRESSED OCEAN PULSE STATIONS
UNSTRESSED OCEAN PULSE STATIONS
@100

Figure 9. Integrated PulselMARMAP sampling areas off the northeast coast of U.S.

164

References

Blumer, M., G. Souza, and J. Sass. 1970. Hydrocarbon pollution
of edible shellfish by an oil spill. Mar. Biol. 5:195-202.
Brandal, P. 0., 0. Grahl-Nielsen, T. Neppelberg, K. H. Palmork,
K. Westrheirn and S. Wilhelmsen. 1976. Oil-tainting of fish,
a laboratory test on salmon and saithe. ICES, C.M.1 976/E,
33:1-6.
Clark, S.H. and B. E. Brown. 1977. Changes of biomass of fin-
fishes and squiclsfromthe Gulf of Maineto Cape Hatteras,
1963-74, as determined from researchvessel surveyclata.
Fish. Bull. 75(l):1-21.
Conover, R. J. 1971. Some relationships between zooplankton
and Bunker C oil in Chedabucto Bay following the wreck of
the ship Arrow J. Fish Res. Bd. Canada 28:1327-1330.
Development Sciences, Inc. 1977. The contribution of fishing
industry knowledge toward assessing the effects of the
Argo Merchant oil spill. Draft report submitted to Office of
Tech. Assess., Wash. D.C. 12 p.
Grahl-Nielsen, 0., T. Neppelberg, K. H. Palmork, K. Westrheim
and S. Wilhelmsen. 1976. The Drupa oil spill, investigation
concerning oil, water and fish. ICES, C.M. 1 976/E:34:1-18.
Grose, P. L. and J. S. Mattson (eds.). 1977. TheArgo Merchant Oil
Spill: A Preliminary Scientific Report. NOAA Spec. Report.
Boulder, Colo. 133 p. App. V11-58.
Grosslein, M. D. 1976. Some results of fish surveys in the Mid-
Atlantic important for assessing environmental impacts.
In: Middle Atlantic Continental Shelf and the New York
Bight. Proc. of the Symp. 1975 Vol. 11. Amer. Soc. Limnol.
Oceanogr. p. 312-328.
Jeffries, H. P. and W. C. Johnson, 11. 1975. Petroleum, tempera-
ture, and toxicants: Examples of suspected responses by
plankton and benthos on the Continental Shelf. In: Effects
of Energy-Related Activities on the Atlantic Continental
Shelf, Bernard Manowitz, ed., p.96-108.
Johannessen, K. 1. 1976. Effects of seawater extract of EKOFISK
oil on hatching success of Barents Sea capelin. ICES,
C.M.1976/E:29:1-12.
Kuhnhold, W. W. 1969. The influence of water-soluble consti-
tuents of crude oils and crude oil fractions on the onto-
genetic development of herring fry. Ber. der Deut. Wiss
Kommn. Meeresforsch. 20:165-171.
Kuhnhold, W. W. 1974. Investigations on the toxicity of seawater-
extracts of three crude oils on eggs of cod (Gadus morhua
L.) Ber. der Deut. Wiss. Kommn. Meeresforsch. 23:
165-180.
Kuhnhold, W. W. 1977. Eff ects of the water soluble fraction of a
Venezuelan heavy fuel oil (No. 6) on cod eggs and larvae.
ICES, C.M. 1977.
Longwell, A. C. 1977. A genetic look at fish eggs and oil. Oceanus
20(4):46-58.
Mironov, 0. G. 1969. The effect of oil pollution upon some repre-
sentatives of the Black Sea zooplankton. Zoologicheskii
Zhurnal 48(7):980-984 (English Translation).
National Academy of Sciences. 1975. Petroleum in the marine
environment. Washington, D.C. Natl. Acad. Sci. 107 p.
Parker, C. A. 1970. The ultimate fate of crude oil at sea--uptake of
oil by zooplankton. AML Report No. B. 198(M).
Sanders, H. L. 1977. The West Falmouth Spill-Florida, 1969.
Oceanus 20(4):15-24.
Sherman, K., J. B. Colton, R. L. Dryfoos, K. D. Knapp, and B. S.
Kinnear. 1974. Distribution of tar balls and neuston
sampling in the Gulf Stream System. NBS Spec. Publ. 409,
Proc. Mar. Poll. Mon. Symp. Gaithersburg, Md. 13-17
May 1974.
Thomas, M. L. H. 1973. Effects of Bunker C oil on intertidal and
lagoonal biota in Chedabucto Bay, Nova Scotia. J. Fish.
Res. Bd. Canada 30:83-90.

165

Socio-Economic Study

Peter Fricke and John Maiolo

Public Knowledge and Perceptions of the
Effects of the Argo Merchant Oil Spill

Peter Fricke and John Maiolo

East Carolina University
Greenville, North Carolina

Thomas and Znaniecki have suggested that if tions. At the time of the Bouchard 65 stranding, Buzzards
people define a situation as real, it will be real in its con- Bay was extensively covered by ice and the channels to
sequences (1927). People act on the basis ofperceptions the Cape Cod Canal were being kept open by a U.S. Coast
of situations, even if those perceptions do not square Guard ice-breaking vessel. Initially the Bouchard 65 oil
with those of others, or some objdctive indicator of a spill was reported by the regional news media, but soon
situation. As simple as this notion is it has been a power- reports could only be found in the local press.
ful tool for social science. Such a notion has assisted in Both spills were reported by the press as causing
the explanation of conflict, even when the objective widespread ecological damage, which in turn would, it
conditions would lead one to a prediction of harmony, and was believed, harm the fishing and tourist industries of
vice versa, the analysis of religious behavior, the adoption the region. Thus an article in the Christian Science
of new practices and so on. Monitor reported that the loss of income to fishermen in
The research reported in this paper explores the five Massachusetts towns caused by pollution from the
perceptions that residents of Cape Cod, Martha's Vine- Argo Merchant could amount to $27.6 million, with a
yard and Nantucket have of the nature and effects of the further possible loss of $83 million to fish processing and
oil spills that occurred in the area in December 1976, and marketing companies (Christian Science Monitor, Decem-
January 1977. It is felt that knowledge of those percep- ber 23, 1976). Fortunately, none of the oil spilled by the
tions is crucial in understanding the nature of the public Argo Merchant appears to have drifted ashore, and the
response to perceived disasters, as public policies are cost of the response to the accident is estimated at $2.7
developed to avert and manage such potential disasters million to date (Adams, 1977). This does not include the
of an environmental nature. costs of any possible damage to the fishery, since this
The Argo Merchant went aground on the Nantucket cannot be determined until statistics for the 1977 year-
Shoals on December 15, 1976. In the ensuing week the class of fish are gathered by the National Marine
tanker broke up and released approximately 27,000 tons Fisheries Service in the next two or three years. It is pos-
of No. 6 residual fuel oil into the North Atlantic. This pollu- sible that even when fishery statistics are available, it will
tion incident became a major news media event, and radio, be diff icult to isolate precisely the impact of theArgo Mer-
television and Press coverage was intensive until the New chant and Bouchard 65 oil spills from. the effects of the
Year (Fricke, ed., 1977). The Bouchard 65, a barge extremely cold winter and spring of 1976-1977 (Vailey
carrying a cargo of No. 2 heating oil, grounded on the Advocate, January 5, 1977; Cape Codder, January 6,
Cleveland Ledge in Buzzards Bay on January 28, 1977. 1977; Falmouth Enterprise, January 7, 1977; Grose and
Approximately 80,000 gallons of oilwerereleased intothe Mattson, 1977).
Bay at the site of the grounding and off Wings Neck, to The Bouchard 65 oil spill had an immediate eff ect,
which point the barge was taken for cargo transfer opera- which was widely reported in the local press, in that the

169

Table 1. Passengers Carried by the Woods Hole, Martha's Vineyard and Nantucket
Steamship Authority Between Cape Cod and the Islands*

1972 1973 7974 1975 1976 1977
No. Y. t No. %t No. %t No. %t No. %t No. %t

January 23,328 100 27,063 116 25,879 111 28,231 121 28,594 123 22,956 98
February 22,205 100 27,803 125 24,343 110 29,574 133 35,044 158 26,299 118
March 30,360 100 36,878 122 36,919 122 40,947 135 37,591 124 36,011 119
April 52,655 100 63,483 121 61,300 116 57,553 109 73,139 139 72,666 138
May 77,999 100 82,267 106 90,428 116 100,039 128 104,220 134 117,790 151
June 125,414 100 133,479 106 133,759 107 154,731 123 156,803 125 160,371 128
July 220,638 100 213,085 97 226,608 103 263,809 120 258,988 117 298,893 135
August 260,421 100 265,326 102 279,059 107 312,548 120 292,025 112 329,170 126
September 128,997 100 145,969 113 132,064 102 144,896 112 157,570 122 169,185 131
Totals 942,017 1 100 995,353 1 106 1 1,010,359 1 107 1 1,132,328 1 120 1 1,143,974 1 121 1 1,233,341 1 131
*Statistics supplied by the Woods Hole, Martha's Vineyard and Nantucket Steamship Authority.
t% of traffic carried in 1972.

Table 2. Expenditure by Travellers to Cape Cod, Martha's Vineyard and Nantucket in Dollarsl

Total for
1975 Cape Cod* Martha's Vineyard- Nantucket region

Winter 37090 840 811 38741
Spring 77023 3543 3136 83702
Summer 147262 10155 6483 163900
Fall 50811 1928 1791 54530

1975 Total 312186 16466 12221 340873

1976
Winter 29323 724 918 30965
Spring 65479 3888 3864 73231
Summer 143739 9830 8105 161674
Fall 32212 1746 1582 35540

1976 Total 270753 16188 14469 301410

*13arnstable County **Dukes County
Information from N.G. Cournoyer and J.K. Kindahl, Travel and Tourism in Massachusetts, 1976.
University of Massachusetts, Amherst. 1977.

shellfish beds in the north-eastern section of Buzzards third phase of reporting consisted of other items which
Bay were closed to the public. Buzzards Bay has suffered also referred to the spills or brief items concerning
from previous spills,and the area has a recent history of continuing research, legislation or action related to the
closed shellfish beds (Blumer et al., 1970). In addition, the spills. Throughout the period covered by the survey of
Bouchard 65 oil impinged more dieectly on the property of local newspapers carried out for this study, from January
local residents, since the clean-up operations were con- to September 1977, third-phase news stories were
ducted from various points along the shoreline (Schwob, published. The proportion declined from 48% of all
1977). marine-related news items in January to 12% in Septem-
After the initial "crisis" coverage of the oil spill ber in the Falmouth Enterprise, for example. It should be
incidents by the press, news items commonly moved to a noted that editorial phase articles also appeared when
second phase of reporting. This phase, which we term new legislation was prepared (May), congressional
.'editorial," discussed, in comment columns and leaders, hearings were held (August), and when books and reports
the response to the spill and measures that were neces- were published (April and August) which bore on the
sary to either improve the response or avert spills. In the spills.
case of the Argo Merchant this phase lasted for most of From the above it would appear that the reading
January, i.e. for approximately four weeks after the crisis public, at least, had a variety of information being pre-
phase of reporting had ended. The Bouchard 65 incident sented throughout the period up to the study of attitudes
attracted less attention, and the crisis phase was over by carried out in October and November. Television and
10 February and the editorial phase by 20 February. The radio, by nature of their medium, tend to rely on the

170

dramatic image and immediate coverage of events. Thus travel and tourism is not, yet available, it would appear
the grounding of the Argo Merchant provided dramatic from press reports of beach use and participation in other
viewing while the vessel was still above water, and little recreational activities, such as the recreational fishing
drama afterward. The Bouchard 65 spill was not a drama- contests, that 1977 has been a good year for the tourist
tic incident; a huddle of tugs, barges, Coast Guard buoy and travel industry. A word of caution should be injected
tenders and a lot of ice were all there was to see on the here, however; nearly all the businessmen interviewed
television screen for the two days the barge remained off agreed that if either the oil from the Argo Merchant spill
Wings Neck. had come ashore or there had been a spill in the late
There were two fears articulated in conversations spring or early summer, the impact upon tourism would
with Cape and Island residents and reported in the press probably have been great.
at the time of the two spills. First, that the oil spilled would The fishing industry on Cape Cod and the Islands is
damage the ecology and environment that the residents a complex one, but characterized by coastal and near-
enjoyed. Second, there was the fear that the publicity shore fishing efforts. The vessels used are relatively
given by the news media would harm the fishing and small and the longest trips made are of the order of a
tourist industries. An expression of this concern was week. The fishing effort is directed toward both shellfish
shown in the reports in the Boston Globe and Christian and fin-fish, and the fish caught are landed at the large
Science Monitor on 23 December that the possible losses centers, Nantucket, Edgartown, Chatham, Provincetown
computed by Massachusetts' off icials for the fishing and and New Bedford. Since fishermen frequently land fish at
tourist industries were of the orderof $158 million in 1977. ports other than their homeport, it is difficuitto distinguish
On the same day a report appeared in the Daily Hampshire clearly the impact an oil spill occurring offshore would
Gazette of a news conference by Dr. Evelyn Murphy, have on a particular community. The fishermen can, if a
Massachusetts Secretary for Environmental Affairs, in particular fishing area is closed becauseof oil slicks,shift
which the setting up of a system of inspection for possible his efforts elsewhere unless he is tied to the exploitation
oil contamination of fish was announced. At the same of a certain species in a specific area.
press briefing spokesmen for the fishing industry assured
consumers that fish available in the shops were not Massachusetts' Commercial Fisheries by Landings
contaminated by oil. and Value*
In addition to the press coverage, information was
received by the residents through other channels. The 1975 1976
Audubon Society, for example, mailed a circular letter to Landings (in'000 lbs) 269,952 288,518
many residents of the Cape and Islands expressing Value (in '000 $) 78,470 97,605
concern about the impacts of spilled oil on the environ- *Information from National Marine Fisheries Service,
ment. The Sierra Club also contacted residents for the Fisheries of the United States: 1976. U.S. Department of
same purpose. Congressman Gerry Studds sends a Commerce/National Oceanic and Atmospheric Adminis-
quarterly Report to the People to all postal patrons in the tration, Washington, D.C. 1977.
12th Congressional District in Massachusetts. In 1977,
the issues of this document emphasized his concern
about flags of convenience, U.S. Coast Guard capabilities Although the press had reported the possibility of a
at times of oil spills, and the need for compensation for major loss of income to fishermen due to the Argo
damage caused by oil pollution. We concluded, then, that Merchant oil spill, the fishermen interviewed in the course
information concerning oil spills was readily available to of this study reported that the 1977 season was as good
any resident who wished to understand their nature, or better than the 1976 season for both catch and
effects and impacts. earnings. Of the three cases in which it was reported that
. Tourism, and the service industries associated with the spill had affected earnings, these had been temporary
it, is big business on the Cape and Islands (see Table 1). It losses and were recouped later in the season. The loss of
is estimated that 75 to 80% of the overall economy of the income in each of the three cases was due to the closure
region is directly attributable to travellers - seasonal home- of the area around the slick by the U.S. Coast Guard and
owners, businessmen, and tourists -as they move about the consequent inability of the fishermen to haul lobster
and stay in the Cape Cod region (Cournoyer and Kindahl, pots.
1977). As can be seen in Table 2, visitors to the region spent Recreational fishing is also an important use of the
nearly $341 million in 1975, and over $301 million in 1976, seas around Cape Cod and the Islands. It is estimated by
on lodgings and services. The drop in expenditures by the National Marine Fisheries Service, in their annual
travellers in 1976 is attributed by businessmen and officials review, Fisheries of the United States: 1976, that some
of their organizations to bicentennial activities elsewhere 626,000 households in Massachusetts engage in
in the United States which attracted both seasonal resi- recreational marine fishing each year. This is the equiva-
dents and tourists. This is reflected in Table 2, which shows lent of 1,300,000 persons approximately. As can be seen
the number of passengers carried by the Woods Hole, in Table 3, many others from nearby states also fish
Martha's Vineyard and Nantucket Steamship Authority in Massachusetts waters. Again, interviews with
ferries. In 1977, however, the number of passengers the 102 recreational fishermen in our sample indicated
carried is significantly higher than in 1975 and 1976. that there was no drop in catch which could be attributed
Hoteliers and other businessmen reported in interviews to theArgo Merchant accident. The 1977 period was seen
that the 1977 season was as good as or better than that of as good as or better than previous years by three-
1975 in terms of both volume of trade and in income quarters of the recreational fishermen. Also interviewed
earned. in the study were seven owners of bait and tackle shops,
Although the annual summary prepared for the State who also reported that their businesses did well i,n 1977.
of Massachusetts by the University of Massachusetts on Their previous best year had been 1975, and all charac-

171

Table 3. Estimated Number of People Participating in Marine United States and had come to live in the Cape Cod
Recreational Fishing in Massachusetts by Northeastern State of region where they had established themselves in small
Residence, June, 1973 - June, 1974* businesses or were working part-time for others. A
retired person in our sample was one who was not
Connecticut 94,000 engaged in any form of work or occupation.
Delaware 5,000 Again the characteristics of the region exerted
Washington, D.C. 2,000 themselves in the proportion of persons who defined
Maine 7,000 their occupation as marine-related, i.e. their job existed
Maryland 16,000 because of the proximity of, and access to, the ocean.
Massachusetts 1,300,000 Many of these respondents were in the tourist industry,
New Hampshire 36,000 and in all they constituted about one-fifth of the sample.
New Jersey 98,000
New York 271,000
Pennsylvania, 83,000
Rhode Island 61,000
Vermont 10,000 Percentage of Respondents Who Define Their Occupa-
Virginia 14,000 tions As Marine-related By Community
West Virginia 3,000
TOTAL 1,998,000 Chatham Edgartown Falmouth Nantucket All
21% 15% 19% 29% 1861.
Information from Fisheries of the U.S., 1976. U.S. Dept of
Commerce, NOAA/NMFS, Washington, D.C. 1977.

terized 1977 as surpassing 1975 for volume of tackle The proximityto the ocean is reflected also in recreational
sales and rental. activities, such as swimming, boating and fishing. During
As can be seen from the review of the tourist and the summer months two-thirds of the sample engaged in
fishing industries above itwould appearthatthe effectsof these pastimes, and 11 % continued to do so in the winter
the Argo Merchant oil spill on the economy of Cape Cod as well.
and the Islands were negligible. The published scientific The respondents also utilized the mass media to a
studies of the Argo Merchant incident carried out to date large extent. Thus 87% of the sample reported that they
also appear to show that impact of the spill upon the watched television for at least one hour every day, 89%
environment was negligible. In this study we were read a newspaper for at least an hour a day, and 73%
interested in the knowledge that the residents of the listened to the radio for more than an hour a day. The
region had of the impact of the spill and their perceptions, indications were, therefore, that the sample would be
if any, of the effects upon their communities and their own well-informed and concerned about the nature and
activities. To this end 260 residents of Chatham, Edgar- effects of oil spills. The interviewers sought to verify this.
town, Falmouth and Nantucket were interviewed during In order to assess the quality of the knowledge of
October and November 1977. In addition, a special sub- respondents a scale was constructed on which the depth
sample of 48 business people were also interviewed in of information could be measured. The researchers
depth to obtain information about economic activities in labelled the respondents well informed if they were able
the region. Both samples were selected at random and to identify two major polluters, and the location of the
can, we believe, be considered representative of the incidents, in the area during the winter 1976-77. Thus it
communities from which they were drawn. was expected that a well-informed respondent would
The sample population of 260 selected for the per- know of the Argo Merchant and the Bouchard 65 acci-
ception study had a modal age of 39 years and a mean dents, and possibly would be able to name other incidents
age of 51 years. Their period of residence in the area as well. Persons knowing of one of the two vessels and its
ranged from one year to more than 50 years, with a mode location were labelled informed, persons knowing either
of 8 years and a mean of 15 years. This reflects the of one of the two ships or of one of the two locations were
population growth, largely through in-migration, in the labelled poorly informed. Persons not able to recall any
past two decades. The population of Barnstable County, information were labelled uninformed. As it turned out this
126,481 in the state census of 1975, has doubled in scale had to be collapsed to the categories of informed
twenty years. These migrants are well educated (modal and poorly-informed/uninformed, because only 14% of
number of years of education is 12 while the mean is 13 the sample could be considered well-informed, and 17%
years) and in "white-collar" and professional occupa- met our criteria for informed respondents. Fully 69% of
tions. In fact 25% of the sample were the owners of small the sample were badly informed or had no knowledge (or
businesses or in business for themselves. Again, this recall) of the two major spills that occurred in the area
reflects the service sector of an economy dominated by eleven months prior to the interviews. (See Table 4.)
travel and tourism. Skilled "blue-collar" workers formed The respondents were asked to name the organiza-
10% of the sample, and another 10% were unskilled tion or agency which had the responsibility for ensuring
workers. Approximately 14% of the sample were retired that spilled oil was cleaned up. For coastal oil spills, such
persons, but again, this brought into focus another as those of theArgo Merchant and Bouchard 65, the Coast
anomaly of this region; many respondents who were gain- Guard is responsible for clean up and it was expected
fully employed had retired from jobs in other parts of the that well-informed and informed respondents would know

172

Table 4. Knowledge of the Argo Merchant and Bouchard 65 Oil Spills by Community (%)

Level of Knowledge Chatham Edgartown Falmouth Nantucket To tal

Informed 29 25 45 23 31
Poorly informed/ 71 75 55 77 69
uninformed

Table 5. Comparison of Informed and Poorly I nformed/U n informed differences were found by community. Ninety-five per-
Respondents' Knowledge Regarding Responsibility for Oil Spill cent of the residents of Falmouth who were interviewed
Clean Up W said they "knew" of effects. In Edgartown, 82% of the
sample "knew" of effects, in Nantucket, 59%, and in
Responsible Organization Informed group Poorly informedl Chatham, 52%. Newspapers were found to be the most
uninformedgroup frequently cited source of such knowledge (53%),
U.S. Coast Guard 54 47 followed by television (47%), and radio (33%). (See
Table 6.)
Coast Guard and 20 8 The respondents were asked if they had actually
other agencies seen or experienced effects of the Argo Merchant oil
spill, and 45% of all the sample said they had seen effects.
Other 15 23 Thirty-four percent of the whole sample reported that their
lives had been affected in some way, but only 5% reported
Don't Know 11 38 that specific activities had been affected by the oil spill.
Of those who reported experiencing effects (34% of the
whole sample) it was found that the reported effects were
11 presumed" to be effects, but at best were indirect. For
Table 6. K 'nowledge of Argo Merchant Spill Effects Reported by example, the increased price of shellfish was attributed to
Respondents by Source of Information and Community the Argo Merchant spill, as was the rising cost of fuel oil.
(N.B. multiple sources of information could be, and were, given) Some respondents reported that the knowledge that
birds and fish were being affected by the spill distressed
Source Community* them, and as such this was an effect of the incident. (See
Table 7.)
Chatham Edgartown Falmouth Nantucket When comparison was made of the perceptions of
the informed group with those of the poorly-informed/
T.V. 56 31 61 39 uninformed group concerning the effects of the wreck-of
the Argo Merchant it was somewhat surprising to the
Radio 49 28 37 26 researchers to discover that the informed group indicated
that they had perceived greater effects than the other
Newspapers 49 68 74 12 group (see Table 8). However, of those who reported
that a specific activity had been affected, there
Friends 7 37 23 38 was no difference between groups. It should be noted
here that of those who reported that theArgo Merchant Oil
*Magazines, relatives and other excluded because of small spill affected a'specific activity, two of the respondents
number of responses. were 'fishermen unable to reach the lobster fishing
grounds because the oil slick lay in their path, and a third
respondent was a fish merchant whose supply of lobsters
had been curtailed because of restrictions placed on
this. The results are shown in Table 5, and indicate vessels fishing near the slick. The other respondents
that nearly three-quarters (74%) of the informed either had to undertake work directly connected with the
group believed that the Coast Guard had some or all of spill or had benefited from increased demand for travel
the responsibility for oil spill clean up. Fifty-five percent of and tourist services created by the influx of personnel
the poorly informed/uninformed group of respondents associated in some way with the spill.
believed that the Coast Guard had some or all of the When the sources of information utilized by the
responsibility, while 23% named other agencies and 38% "knowledge" groupings of respondents were analyzed
did not know who was responsible. few differences were found, with one exception. Fewer
Information about the Argo Merchant oil spill was informed respondents (20%) indicated that friends and/or
obtained from television programs by 65% of the sample. relatives were a source of information than the other
In addition, 53% of the respondents reported that they group (32%). (See Table 9.)
obtained information from newspapers, and 42% also Included in this random sample of 260 residents of
heard news stories about the spill on the radio. Relatives Trie Cape and Islands were 65 owners of businesses. Of
were an additional source of information for 5% of the these, 35 (54%) reported that their trade was seasonal,
respondents, and friends for 18%. Over three-quarters i.e. was tourist oriented. As was expected the percentage
(76%) of the persons interviewed said that they "knew" differed by community, viz. 82% in Edgartown, 64% in
of some of the effects of theArgo Merchant spill, and here Chatham, 50% in Nantucket, and 38% in Falmouth. All had

173

Table 7. Perceived Effects of the Argo Merchant Oil Spill by Community (%)

Chatham Edgartown Falmouth Nantucket Sample

Reported
seeing effects 23 74 74 53 38 45

Reported
affecting respondent 16 43 49 25 34

Reported specific
activities affected 5%

*Number of responses too small to compare by community.

Table 8. Comparison of Perceived Effects of the Argo Merchant Oil as (60%,) or better (32%) than previous years.
Spill Reported by Informed and Poorly Informed/Uninformed Attention was also directed at those respondents
Respondents N of group) who enjoyed fishing as a pastime. Of this sub-sample of
102 respondents, 58% indicated that they owned a boat
Poorly informedl (alone or with a partner). Nearly three-quarters of the
Perception Informed Group uninformed Group recreational fishermen perceived the 1977 fishing
season as being the same or better than previous years.
Reported seeing an Of those who indicated that fishing was worse, many
effect 49 39 commented that "natural" cyclical factors were
operating. Oil spills were not blamed for poor catches,
Reported affecting with the exception of those respondents seeking shellfish
respondent 38 33 in Vineyard Sound, and spills prior to the Argo Merchant
were said to have damaged this fishery.
Reported specific To summarize, of the sample of 260 residents of
activity affected 5 5 Cape Cod, Martha's Vineyard and Nantucket, interviewed
about their knowledge and perceptions of the effects of
the Argo Merchant oil spill, 69% were either poorly
Table 9. Comparison of Sources of Information About the Argo informed or uninformed (or had no recall) of the events of
Merchant Oil Spill Between Informed and Poorly Informed/ the winter of 1976-77. An operating assumption of the
Uninformed Respondents study was that if basic information was unknown, more
(%of group) detailed knowledge of the extent of oil spills, environ-
mental damage and pollution clean-up would also be
Poorly informedl unknown. It can be fairly said that an uninformed public is
Source Informed group uninformedgroup a susceptible one in that impressions, incomplete reports
and informal sources of information can appreciably
T.V. 46 46 affect public opinion to the extent that a large gap exists
between perceptions and reality.
Radio 33 31 The three major sources of information utilized by
the respondents were television, newspapers and radio.
Newspapers 53 52 If our index of knowledge is accepted, these media
functioned to inform and keep informed less than a third
Magazines 12 7 of the sample. Yet three-quarters of the sample said they
"knew" of effects of the Argo Merchant oil spill and 45%
Friends/Relatives 20 32 reported that they had "seen" effects, most of which were
environmental. An inspection of the data obtained by in-
depth questioning by interviewers reveals that these
felt concerned that the Argo Merchant's mishap would "observed" effects were taken seriously by the respon-
adversely affect their summer season because of the dents. The perceptions of the situation for those
weight of publicity in December 1976. However, 60% of respondents, whether or not those perceptions square
the business owners felt that the summer of 1977 had with reality, are that the Argo Merchant oil spill was
brought the normal number of tourists to their region. damaging to the environment and/or to the social and
Forty percent of the owners felt that a change had economic life of the Cape Cod region. It is believed by the
occurred, and of these, three-quarters said that their authors that knowledge of these perceptions needs to be
trade and the number of tourists had increased. Of all the incorporated into any effects to inform the public and into
proprietors, 92% felt their business in 1977 was the same any planning for the prevention of or response to oil spills.

174

Acknowledgments
The research on which this paper is based was
carried out by the Institute for Coastal and Marine
Resources and the Department of Sociology and Anthro-
pology at East Carolina University for the National
Oceanic and Atmospheric Administration under Contract
No. 03-7-022-35194.

References

Adams, G., Landmark bill on oil spills, in Vineyard Gazette,
September 20, 1977.
Blumer, M. J., G. Souza, H. L. Sanders, J. F. Grassle, and G. C.
Hampson, The West Falmouth oil spill: Persistence of the
pollution eight months after the accident. Woods Hole
Oceanographic Institute Technical Report 70-44, Woods
Hole. 1970.
Cournoyer, N. G., and J. K. Kindahl, Travel and Tourism in
Massachusetts, 1976. University of Massachusetts,
Amherst. 1947.
Fricke, P. H. (ed.), A catalog of press clippings relating to the
Argo Merchant oil spill. Woods Hole Oceanographic Insti-
tution Technical Report to the National Oceanic and
Atmospheric Administration, Woods Hole. 1977.
Grose, P. L., and J. S. Mattson, The Argo Merchant oil spill: A
preliminary report. U.S. Department of Commerce/NOAA,
Washington, D.C. March, 1977, pp. iii-vi.
Schwob, Rear Admiral William S., USCG, Testimony at the
Hearing of the U.S. House of Representatives Committee
on Merchant Marine and Fisheries, Subcommittee on
Coast Guard and Navigation, Hyannis, August 15, 1977.
Thomas, W. I., and F. Znaniecki, The Polish Peasantin Europe and
America. Alfred Knopf, New York. 1927.

175

Future Response Plans

Paul Lefcourt

Ecological Damage Assessment of Oil
Spills-the Federal Government's
Response

Paul Lefcourt

U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C.

The authority for the federal government to respond *Establishes a system of enforcement and provides for
to spills of, oil and hazardous substances stems from the penalties for violation of any part of Section 311;
Federal Water Pollution Control Act (FWPCA) Amend- *Assigns responsibility for various aspects of oil spill
ments of 1972 (PL 92-500, Section 311) and the 1977 response among federal agencies;
Amendments to the Act (33 U.S.C. 466 et seq.). The *Establishes a national response team to oversee the
applicable regulations are given in the National Oil and federal government's involvement in oil spills, composed
Hazardous Substances Pollution Contingency Plan of five primary federal agencies (DOT, DOI, DOD, EPA,
(40 CFR 1510). DOC) and a number of advisory agencies;
The highlights of Section 311 of the FWPCA are *Establishes mechanisms to coordinate all oil soill
as follows: response through a single federal agent - the "On-Scene
*Prohibits the discharge of harmful quantities of oil and Coordinator;"
other hazardous substances to the waters of the U.S.; *Establishes a schedule to regulate the use of chemi-
*Requires notification to the federal government when- cals and other additives to remove oil and hazardous
ever a prohibited discharge occurs; substances discharge.
*Authorizes the President to remove or arrange for the
removal of an oil spill unless such removal is properly
carried out by the responsible party; The National Contingency Plan (NCP) identifies five
*Requires the promulgation of a national contingency federal agencies for primary responsibility in dealing with
plan to provide for efficient, coordinated and effective oil and hazardous substances spills. The primary agen-
action to minimize damage from oil or hazardous cies and their responsibilities are as follows:
substances spills; 1. Department of Transportation (DOT)
*Establishes liability limits for vessels, onshore facili- *Provide expertise in port safety and security, marine
ties and offshore facilities for the costs of removing spills law enforcement, navigation and construction, manning,
caused by these facilities; operat *ion and safety of vessels;
*Authorizes the President to promulgate regulations: *Furnish or provide for the On-Scene Coordinator
-establishing methods for removal of oil and (OSC) for coastal waters and the Great Lakes;
hazardous substances spills, *Chair RRT in areas where it provides OSC;
-governing the inspection of vessels carrying oil and *Devel 'op, implement and revise regional contingency
hazardous substances cargos, plans in areas where it provides OSC.
-establishing criteria for the development and imple- 2. Department of the Interior (DOI)
mentation of local and regional contingency plans *Provide expertise in oil drilling; producing, handling
for spill removal, and and pipeline transportation;
-establishing procedures and methods for preven- *Through its regional coordinators, provide technical
ting discharges of oil and hazardous substances expertise to the OSC and RRT with respect to land, fish
from vessels, onshore facilities and offshore and wildlife, and other resources for which it is
facilities; responsible;
*Authorizes the establishment of a revolving fund to *In conjunction with state liaison to RRT arrange for
pay for the cost of oil and hazardous substances spill and coordinate actions of groups to establish bird
removal; collection, cleaning and recovery centers.

179

3. Department of Defense (DOD) The objectives of the task force were to provide a plan of
*Provide assistance in critical pollution discharges, action which would enable the federal government to:
consistent with its operational requirements; -provide highly qualified and coordinated scientific
*Provide assistance in the maintenance of navigation support to regional response teams and on-scene
channels, salvage, and removal of navigation obstruc- coordinators during major spill incidents;
tions. -upgrade our capability to assess environmental
4. Department of Commerce (DOC) damage associated with these spills; and
*Through NOAA, provide support to the NRT, RRT and -capitalize on the unique research opportunities
OSC with respect to marine environmental data, living which are often afforded by major spills and thus
marine resources, current and predicted meteorological, improve our ability to support future clean-up and
hydrologic and oceanographic conditions for the high damage assessment activities.
seas, coastal and inland waters; The task force concluded:
*Provide maps and charts, including tides and -although a great deal of scientific capability exists
currents for coastal and territorial waters, and the at present to provide assistance to those involved in
Great Lakes; clean-up, damage assessment and public informa-
*Through MARAD, provide advice on the design, tion activities, we are not now prepared to provide
construction and operation of merchant ships. that support on a routine basis;
5. Environmental Protection Agency (EPA) -future efforts must be organized in a manner which
*Provide expertise regarding: will allow @i prompt match of scientific capability to
-environmental effects of pollution discharges the issues involved in operations and assessment;
-environmental pollution control techniques and
-assessment of damages; -the research base is currently inadequate to re-
*Provide advice to the RRT and OSC of the degree of solve many of the questions raised by such spills.
hazard a particular discharge poses to the public health The task force arrived at the following recommendations
or welfare; to the National Response Team:
*Furnish or provide for the OSC for inland waters; -establishment of a national scientific support
*Chair RRT for areas in which it provides the OSC; coordinator as part of the NRT;
*Develop, implement and revise regional contingency -establishment of scientific support coordinators
plans for areas where it provides the OSC; (SSC) on either a regional or an area basis, the
*Provide guidance to and coordinate with DOT re- coordinator to serve as the single contact with the
garding pollution control and protection of the environ- OSC for all environmental inputs/issues;
ment in the preparation of regional plans. -development of scientific support teams to function
under the SSC as an operational arm of the regional
It is important to note that the U.S. Coast Guard response team; and
provides the OSC for spills occurring in coastal waters -development of regional contingency plans for
(extending inland to the extent of tidal influence) and the ecological assessment through a series of work-
Great Lakes. The U.S. Environmental Protection Agency shops to be held in each EPA coastal region.
(EPA) provides the OSC for all other waters of the United
States and its possessions, i.e., essentially all fresh
water spills. In carrying out his function the OSC has The first regional workshop was held for New
broad authority to take whatever steps are necessary to England (EPA, Region 1) August 28-31, 1977, in Hartford,
ensure a timely and effective clean-up operation. The Connecticut. Based on the workshop, specific plans for
OSC uses as guidance existing regional and local spill performing ecological damage assessments of major oil
contingency plans. He can receive advice and assistance spills are being formulated.
from both the Regional Response Team (RRT) and the
The second workshop was held in Anchorage,
National Response Team (NRT). The OSC has at his Alaska on November 28-30,1977, to assist in developing
disposal a spill contingency fund which he can use a plan specific to the needs of Alaska.
at his discretion to pay the costs of the clean-up The third workshop, for the entire Gulf of Mexico,
operation. was held April 3-5,1978, in Tampa, Florida.
Under the terms of the NCP, the EPA is assigned
the responsibility for determining the extent of damage The objectives of the workshops are:
resulting from major spills. Funding for damage *To identify experts p!esently working in oil pollution
assessment studies can not be obtained from the research and operations in the region;
present spill contingency fund. Nevertheless, 19 *To compile research needs and projects applicable to
studies of spills have been performed using EPA the environmental characteristics of the region which
funding sources. could comprise elements of damage assessment studies
or research programs;
The Argo Merchant spill revealed that in carrying out *To catalogue identified experts and their areas of
a large damage assessment study, many areas, such as expertise into a regional directory of oil spill workers who
coordination of studies, planning for studies, and the could be called on for scientific advice to the On-Scene
state of the art in performing studies, were less than Coordinator during spill incidents;
desirable. In the spring of 1977 the National Response *To identify presently available resources, facilities,
Team moved to improve its capabilities for performing and support services for scientific responses to oil spills.
assessments of damage, especially ecological damage,
by establishing an ad hoc inter-agency task force to The emergency response for damage assessment
examine the problem of damage assessment and report studies focuses on using local scientific resources
back to the NRT their recommendations. supplemented by national expertise and capability in

180

specialized areas such as analytical chemistry and spill
trajectory analysis. The planned field programs will be
under the direction of the Scientific Support Coordinator
(SSC). The national plan identifies two SSC's for each
coastal area - one from EPA for spills originating inside
the baseline from which the territorial sea is measured
(essentially spills in bays, estuaries, rivers, etc.) and a
SSC from NOAA responsible for spills originating outside
the baseline (such as the Argo Merchant spill).
Pending legislation dealing with the so-called
"superfund" would provide, if passed, a $200,000,000
liability fund. The unique feature of this legislation from an
ecological science perspective is the provision allowing
for compensation for damages to "natural resources".
Clearly, if this legislation passes, the marine science
community will be tasked with defining "damage" to
"natural resources" and perfecting the methodology to
allow the use of scientific information in economic/legal
evaluations. The federal program briefly described above
is an attempt to come to grips with this challenging
problem and in the process develop a framework for most
effectively using scarce resources to learn what is the
true extent of damage resulting from spills of oil and
hazardous substances.

181

COASTAL ZONE
INFORMATION CENTER

I I BA7E DUE

I I I
I I I
GAYLORDINO. 2333 1PRINTED IN U.S.A.I

1111
3 6668-14105 8208@